foresight user’s manual table of contents - farin user’s manual table of contents 5.2.7.6 1....

FORESIGHT User’s Manual

Table of Contents 5.2.7.6

1. Establishing a Framework for Making Effective Asset-Liability Management Decisions 1.a. Overview 1.b. Definitions and Concepts 1.c. Establishing Financial Goals 1.d. Establishing Risk Tolerance Levels 1.e. Interest Rate Risk Measurement Tools and Techniques

1.e.i. Gap Analysis 1.e.ii. Duration Analysis 1.e.iii. Dealing with Option Risk

1.f. Using an Asset-Liability Model to Perform Income at Risk Testing 1.g. Using an Asset-Liability Model to Perform Value at Risk Testing

2. FORESIGHT Startup

2.a. Logging Into FORESIGHT 2.b. Manage Database Screen

2.b.i. Selecting a Database 2.b.ii. Creating a Backup 2.b.iii. Restore Function 2.b.iv. Copy Database 2.b.v. Delete Database 2.b.vi. Other Manage Database Options

2.c. Institution Tab 2.d. The All Tasks Menu

3. FORESIGHT Current Balance Sheet and Income Statement

3.a. FORESIGHT100 – The Importance of Understanding Cash Flows 3.b. Choosing Your Level of Detail - Managing your current layout

3.b.i. Inserting Items 3.b.ii. Deleting Items 3.b.iii. Renaming Items 3.b.iv. Moving Accounts and Categories

3.c. Setting Cash Flow Characteristics for New and Existing Accounts 3.d. Historical Data 3.e. Cash Flow Screen

Appendix – Amortization Types Appendix – Characteristics Setting examples

4. Populating Your Accounts - The Download

4.a. Downloading Data Automatically 4.a.i. The Download Process

Understanding Your Data Creating Descriptions or Download Mappings

4.b. Transformation Rules

4.c. Consolidation Rules 4.d. Reading the Raw Data 4.e. Mapping your Accounts 4.f. Importing Cash Flow Balances 4.g. Importing Historical Values 4.h. Importing History Data 4.i. Importing CMO Prepayments and Market Values 4.j. Manually Entering your Balances (including WARM balances)

Appendix – Download Description examples Appendix - Expression Transformation examples Appendix – Row Match Rule examples

5. FORESIGHT External Assumptions

5.a. Prepayment Speeds 5.b. Decay Rates 5.c. Market Rates

5.c.i. Historical Rates 5.c.ii. Forecast Rates 5.c.iii. Adding Instruments 5.c.iv. Environments/Global Insight rates 5.c.v. Edit Rate Behaviors 5.c.vi. Graphs

Appendix – Non Parallel Rate Environment 6. FORESIGHT Internal Assumptions - Maintaining Forecasts/Budgets

6.a. Where are You Going? 6.b. What is a Forecast/Budget? 6.c. Pricing

6.c.i. Fixed Rates 6.c.ii. Offset from Rate Driver 6.c.iii. Formula 6.c.iv. Lagging Drivers

6.d. Targets 6.d.i. Choosing a Target Method 6.d.ii. Suggestions for Target Settings 6.d.iii. Calculate

6.e. Other Tabs 6.e.i. Sales Tab 6.e.ii. Market Value Tab 6.e.iii. Gap Tab 6.e.iv. Messages Tab 6.e.v. Comments Report

6.f. Forecast and Budgets 6.f.i. Create a New Forecast/Budget 6.f.ii. Edit Forecast/Budget 6.f.iii. Edit Time Period Sets

6.g. Set up for Dividends 7. FORESIGHT Market Valuation

7.a. Net Portfolio Value and Valuation 7.a.i. The Valuation Approach

7.a.ii. Cash Flows 7.a.iii. Discounting 7.a.iv. Setting Up Valuation in FORESIGHT

7.b. An Alternative Discount Rate – Synthetic Curves 7.c. Line Item Calculations 7.d. Adding Off Balance Sheet accounts to Market Value Reports 7.e. Market Value Overrides Appendix – Non Maturing Deposits

8. FORESIGHT Reports 8.a. Overview 8.b. View an Existing Report or Chart

8.b.i. Report Features 8.b.ii. Graph Features

8.c. Change Features of an Existing Report or Chart 8.d. Creating a New Report 8.e. Export a Report to a File 8.f. Print an Existing Report or Chart 8.g. Remove a Report, Chart or Report Set 8.h. Edit Account Dictionary 8.i. Edit Policy Limits 8.j. Calculated Rows 8.k. Accounts Report 8.l. Net Income Limits Report 8.m. ROA Limits Report 8.n. Market Value Limits 8.o. Liquidity Report Adjustments 8.p. Forecast Accuracy Reports

- - - - - - - - - - - - - - Appendix – Cash Flows – The Importance of (H&R) Appendix – Download File Issues (H&R) Appendix – Farin Rate Services/List of Market Rates and Curves Appendix – Foresight Installation Requirements (H&R) Appendix – Model Rolling Checklist (H&R) Appendix – The Chart of Accounts (H&R)

Farin Foresight User Manual Page 1 of 36 © Copyright 2002-2005 by Farin & Associates, Inc.

Chapter 1

Establishing a Framework for Making Effective Asset-Liability Management Decisions

1.a. Overview Many institutions purchase asset-liability models as a result of regulatory pressure. Understandably, the use of the model focuses on producing the reporting necessary to comply with regulatory initiatives. As a result, the institution’s investment in the model is regarded as a necessary cost of regulatory compliance. Other institutions use an A/L model as a budgeting or planning tool. In this role, the model is used to formulate a plan or budget and then to track the institution’s performance relative to budget. The investment in the model is looked upon as an investment in a more efficient budgeting or planning process. Less frequently, financial institutions use A/L models as tools in making effective management decisions. The model is used by the institution in evaluating the risk/return tradeoffs associated with alternative strategies that might be employed in running the institution. All three are legitimate uses of an A/L model. But models used by institutions in making effective management decisions can often have their costs recaptured many times by the impact the decisions have on performance. This chapter is written to lay out the framework for using an A/L model in making effective management decisions. It begins with a set of definitions and concepts. From there it proceeds into the process of setting financial goals and limitations on various kinds of risk. As one of the primary uses of A/L models is to measure interest rate risk, methods for measuring interest rate risk are reviewed and discussed. Inputs into developing a plan are reviewed and discussed. Finally, an approach is suggested for dealing with change in the planning process. Throughout this chapter you will find boxes similar to this box. This chapter primarily lays out the theory of best practices in A/L management. The material in these boxes will discuss how this theory is implemented in Farin FORESIGHT and where else to look in this manual for relevant material.

1.b. Definitions and Concepts Before developing the framework for decision making we must first get a number of definitions and concepts on the table that define the objectives of the decision making framework we need to put in place. Let’s start out with a definition of asset-liability management.

1.b.i. Asset-Liability Management – the management of the relationship between a financial institution’s risk and its return.

Unless it operates in an environment where it has a competitive monopoly, a financial institution must take on risk if it hopes to produce an acceptable return. With the globalization of the economy, removal of branching restrictions, development of the Internet as a delivery channel, and the growth of non-bank competitors, competitive monopolies are rare in today’s banking world.


Risks that must be managed include (but are not limited to) credit risk, interest rate risk, liquidity risk and capital risk. These risks must be managed within acceptable ranges while producing a

desired return to the stakeholders of the organization.

1.b.ii. Static vs. Dynamic Risk Management – Traditionally, financial institutions have measured risk by looking at risk levels in old balance sheets. You might argue that traditional ‘static’ measures of risk really look at risk in current balance sheets. But by the time the analysis on a ‘current’ balance sheet is complete, the balance sheet is no longer current – it may be weeks or months old. Our industry’s history of static risk management was driven by the limitations of the regulatory agencies call report gathering process. Information contained on the call report describes where the institution has been, not where it is going. As a result static measures of liquidity (like the loan/deposit ratio) or interest rate risk (like the gap/asset ratio) evolved and became industry standards as they could be calculated from call report data.

Dynamic risk measurement looks at the risk inherent in a plan or strategy. Dynamic measurement systems recognize that risk is not static. Risk changes dynamically over time based on the business strategy being executed by the institution. If we’re going to optimize the relationship between risk and return in the A/L process, we must use forward-looking measurement systems rather than those that measure the risk in our wake. Besides, only with measurement systems that look at risk/return tradeoffs in individual plans and strategies can we choose the strategy that best optimizes the relationship between risk and return.

1.b.iii. Using Metrics that Relate Risk to Return – Traditional measurement systems used static measures like gap/asset ratios or loan/deposit ratios to measure the level of interest rate risk and liquidity risk. In this static measurement environment, less risk was ‘good’ while more risk was ‘bad’. Yet financial institutions make money by taking on risk. So in at least some circumstances, more risk has to be good. This apparent contradiction in risk management is overcome when risk is stated in the same terms as those used to set and measure an institution’s ability to achieve its long-term financial goals. Doing so makes it easier to assess risk/return tradeoffs.

Two key financial goals an institution establishes are in the area of return (ROA, ROE, or Earnings Per Share) and in the area of capital (capital/assets, risk based capital/risk based assets). In the interest rate risk management arena, evolution is away from traditional measures like gap/assets and in the direction of measures of the effect of changes in rates on income (Income at Risk) and capital (Value at Risk). The closer our asset-liability systems relate risk to our ability to achieve financial goals, the easier it is to measure our success in meeting our A/L objective of optimizing the relationship between risk and return.

1.b.iv. Book Value vs Economic Value – Financial accounting systems have traditionally carried the value of an asset or liability at the value it had when placed on the books (adjusted for depreciation, amortization, and other factors that would cause it to naturally drop in value over time). Yet the vast majority of a financial institution’s balance sheet is made up of financial instruments (loans, deposits, investments, borrowings). With few exceptions, the values of these instruments decrease in rising rate environments and increase in falling rate environments.

The accounting standards have been adjusted over the years to require that some financial instruments on a financial institution balance sheet be carried at market value (investments classified as ‘Available for Sale’ for example). Yet the majority of a financial institution’s balance sheet is still carried at book.


Interest rate risk is one of the key risks a financial institution must manage. Changes in rates cause values of most financial instruments to fluctuate, some a little, others a lot. Because the effect of changes in rate on value (value at risk) is a crucial interest rate risk measure, it makes much more sense to work with the market value of the instrument rather than book. Economic value is the difference between the market value of assets and liabilities. As a result institutions that have traditionally looked at the long-term effect of changes in rates on book value of capital, are switching over to looking at the effect of changes of rates on economic value.

Let me summarize the results of the discussion in the definitions and concepts section.

1. We must provide a planning framework that promotes actions on an institution’s part that optimize the relationship between risk and return.

2. Risk must be measured in dynamic setting allowing the institution to see risk developing in a plan before (rather than after) it affects financial performance. Only with this framework can we look at the risk/return tradeoffs between alternative strategies.

3. Metrics used in measuring risk must relate risk measures as closely as possible to measures of return and other financial goals. The closer risk measures are to financial goals, the more obvious is the relationship between cause and effect.

4. Measures must consider both book value (for accounting and regulatory compliance reasons) and economic value (a measure much closer to an institution’s value in the market).

1.c. Establishing Financial Goals Establishing a Time Frame for Goal Setting Picking a time frame for an institution to meet long-ranger goals is a relatively simple issue. Think strategic planning horizon. For most financial institutions, the strategic planning horizon is three to five years. Not that many years ago most institutions performing strategic planning focused on building a five-year annual financial plan. Three-year quarterly plans are much more common today for a number of reasons.

1. There is so much turmoil in the financial services sector that it is difficult to build a plan that reaches a full five years into the future. As the planning horizon is extended, assumptions become increasingly heroic.

2. As we’ll point out later in the chapter, institutions modeling income at risk in a 1-year plan must do so over an extended time horizon to feel the full effect of changes in rates on income. Horizons for extended income at risk modeling of three years are often used. If you are going to model income at risk over a three year time frame, your income at risk modeling horizon may match your strategic planning horizon.

3. Institutions are updating their pans much more frequently than annually to react to changing competitive and economic decisions. In many institutions plans are updated quarterly, or even monthly. Rolling a plan forward on a quarterly basis fits better with a three year quarterly plan than with a five-year annual plan.

In summary:

1. Choose your strategic planning horizon. 2. Establish your long-range goals. For most institutions, the long-range goals you establish should

be attainable within the planning horizon. If your performance is sufficiently far from the goals that they aren’t attainable within a year, you may also want to establish some intermediary goals, especially for the next budget year.

Farin FORESIGHT allows you to establish Policies (plans) that project into the future as far as you would want to project. See the end of Chapter V and all of Chapter VI for materials on developing and managing Policies.


Goals Focused on Net Income

For almost all institutions, the best measure of financial performance is the amount that shows up on the bottom line, net income. But a dollar net income target is rarely found on an institution’s list of long-range

financial goals. Rather, net income is generally divided by some denominator to convert it into a ratio which is used as the goal. Ratio goals offer the following advantages over dollar net income targets:

o With ratios, it is much easier to assess trends in an institution’s performance over time while the

institution grows or shrinks. o Ratios make it much easier to compare an institution’s performance to that of another institution

or to a peer group average. o Long-range goals can be established without worrying about the effect of the institutions growth

rate on income. A number of ratios are commonly used in setting net income goals. What they have in common is that each has net income in the numerator.

ROA

ROA is calculated by annualizing the net income produced in an accounting period and dividing by the average amount of assets producing the income during the accounting period. It answers the question, “How effectively did the institution use its assets to produce net income in the period being measured?”

ROA remains an extremely popular ratio for setting goals and measuring performance. Managers of stock and sometimes mutual organizations often receive pay incentives based on ROA. Examiners often assign the value for the E (earnings) in the CAMEL(S) ratios based on ROA. But use of ROA as an institutions most important measure of income performance is prone to two problems.

1. It encourages an institution to maintain more capital than it needs to cover its risk. Why? Because net income calculations assign no cost to capital. The cost of capital is not considered until after ROA is calculated. Capital represents a ‘free’ funding source. Institutions attempting to maximize ROA would prefer to fund with capital at a cost of funds of 0% than with an interest paying liability at an interest cost of 3%.

2. As financial institutions do more and more of their business off their balance sheet (mortgage banking, trust, investment advisory, insurance), ROA becomes less relevant as these activities cause relatively nominal growth in assets on the balance sheet.

Earnings Per Share (EPS)

EPS is calculated by taking net income and dividing by the number of shares of outstanding stock. It is communicated by specifying the ratio and the accounting period being used. For example, “Our earnings were $0.30 per share in the fourth quarter. That resulted in total earnings per share for the year of $1.00. The ratio answers the question, “How much income did the institution produce for each share of stock in

the accounting period.

EPS is an extremely useful measure of income performance for stock institutions in that the market tends to price stock as a multiple of earnings. For example, if the above institution’s stock is trading at a

price/earnings (P/E) ratio of 15, their stock would currently be priced at $15 per share.

Price = P/E x EPS = 15 x $1.00 = $15


While EPS goals are extremely useful for stock institutions, they are not terribly useful for peer group comparisons as number of shares of stock outstanding and price per share of stock vary widely from stock

institution to stock institution. Of course EPS goals are useless to mutual savings institutions and credit unions in that there is no number of shares of stock to use in the denominator of an EPS calculation. Nor is there a stock price derived from

a combination of EPS and the P/E ratio.

Return on Equity (ROE)

Return on Equity is calculated by annualizing the net income produced in an accounting period and dividing by the average amount of capital (equity) on the books during the accounting period. It answers the question, “How effectively did the institution use its book Capital to produce income during the accounting period?”

ROE can be calculated for stock institutions, mutual institutions and credit unions. All you need is income and capital. For a stock institution, it is a ratio with many of the same characteristics as EPS. It has net income in the numerator. It has a form of capital in the denominator. In comparing the two, ROE offers a less direct correlation to movement in stock price over time. But it is a much better ratio to use in comparing an institution’s performance to peers, whether the peer is a stock institution, a mutual or a credit union.

Why is this ratio relevant to mutuals and credit unions? Because if you want to keep your capital/asset ratio constant you need to grow the numerator of the C/A ratio at the same speed as the denominator. When you recognize how ROE is calculated (net income / capital) you realize that ROE represents the growth rate of the capital account. Therefore, mutuals and credit unions wishing to maintain a constant capital to asset ratio could set the following goal.

Asset Growth Rate = ROE

Because of the interaction between the ROE, the asset growth rate, and the capital/asset ratio, mutuals and credit unions that focus on setting income goals using ROE find it much easier to set goals that insure consistency between the institution’s earnings goals, growth goals and capital goals. Setting Goals for Capital Those setting capital goals are under two opposing pressures:

o Return Pressure – A lower capital/asset ratio translates into a higher equity multiplier and a higher ROE. Leveraging capital to enhance ROE is important to stock institutions wishing to provide a market return to their stockholders. It is also important to mutuals and credit unions who want to grow their institution. The higher their ROE, the faster an institution can grow assets while maintaining capital/assets at its current level.

o Capital Risk Pressure – The lower the capital/asset ratio the more likely an institution is to allow capital to become inadequate to cover the risk in its balance sheet. Institutions with low capital/asset ratios are also more likely to fail regulatory capital standards.

The ROE pressure comes from a fundamental relationship between ROE, ROA, and capital/assets. ROE = ROA x Equity Multiplier The Equity Multiplier is the inverse of the Capital/Asset ratio.

Equity Multiplier = 1 / Capital/Assets

So the ROE equation could be restated as follows.


ROE = ROA / (Capital/Assets)

The upshot of this relationship is that an institution wishing to increase ROE must either increase ROA or lower its capital/asset ratio. If it chooses the latter, it increases capital risk in the institution in that there is less capital in relation to assets to absorb the negative effects from risk.

So an institution setting goals for their capital ratio must:

o Understand the relationship between capital/assets, the equity multiplier, ROA and ROE. o Understand their regulatory capital standards. o Determine how close to regulatory capital minimums they are willing to operate. o Understand the fundamental risks in their balance sheet. o Have a solid sense for future growth plans. Institutions anticipating a future increase in the rate of

asset growth or the potential for acquisition of another organization may build up excess capital over the short term because they realize that during times of rapid growth they may be unable to generate sufficient ROE to grow capital as quickly as assets.

Once an institution has their arms around these issues they should set a goal for capital/assets that represents a compromise between their demands for ROE and their requirements to run a safe and sound institution. Because different institutions carry different levels of risk and because stakeholders of one institution may be significantly more risk adverse than others, there is no way of recommending a universal capital goal. Rather, an institution’s management and board must work through these issues and establish an institution’s goal for capital.

Once a capital goal established, the institution should manage to, but not through the capital goal. For example say an institution that sets a capital/asset goal of 8% and allows its capital/asset ratio to reach 10%. They will be more safe and sound than if the capital ratio was at 8%. But their ROE will fall below goal because of the relationship between ROE, ROA, and capital/assets. Establishing Goals for Growth

Asset Growth Goals The most important goal for growth is the institution’s goal for growth in total assets. This goal is very important because it determines what will happen to the denominator of the capital/asset ratio

over time. Once we’ve set the rate of growth for capital/asset’s denominator the actions we will need to take in setting ROE goals (as well as dividends and other stock actions for stock

institutions) will become more obvious. The key is to develop a plan that manages the relationship between growth, earnings, and capitalization.

Loan Growth Goals

Most institutions attempt to grow loans as quickly as assets. Doing so maintains a constant loan/asset ratio. Institutions may make exceptions to this rule under the following circumstances.

o The loan/asset ratio is lower than considered to be optimal. Under these circumstances, loan growth goals will exceed asset growth goals.

o The loan/asset ratio is higher than considered to be optimal. Under these circumstances, asset growth goals will exceed loan growth goals.

o Loan demand in the market is weak or loans are considered to be poorly priced in relation to investment alternatives. Under these circumstances, asset growth goals will exceed loan growth goals.

o Loan demand in the market is strong or loans are considered to be very well priced in relation to investment alternatives. Under these circumstances, loan growth goals will exceed asset growth goals.


o An institution wishes to grow extremely quickly in an attempt to reduce its capital/asset ratio. Loan demand is insufficiently strong to keep up with the proposed asset growth rate. So growth is accomplished by using deposits or borrowed money to fund investments.

Deposit Growth Goals

Ideally, an institution will set deposit growth goals equal to asset growth goals. Many institutions have found this to be impossible to accomplish due to a number of factors.

o Younger consumers are increasingly looking to market instruments (both fixed income and

equities) as a place to invest long-term funds rather than making their parent’s or grandparent’s choice of investing in financial institution CDs. As a result, the average age of CD holders is rising. As this wealth transfers from seniors to younger generations, the investment choices of those inheriting the funds will slow financial institution deposit growth rates.

o An institution may find itself in a market with relatively few net savers and a substantial population of net borrowers. As a result it finds it difficult to fund all its loan growth with deposits.

o Consumers tend to supply contractually short-term funding (short-term CDs and transaction accounts). At the same time they demand fixed-rate long-term loans. Because financial institutions need to manage the interest rate risk in their balance sheets they often find it necessary to reach into wholesale funding markets for funding to fund long-term loans.

o Competition in the institution’s market may be driving deposit rates sufficiently high that an institution finds it cheaper to raise funds in the wholesale markets than it is to raise funds from depositors.

Financial institutions of all kinds have seen significant drops in the percentage of funding provided by depositors and corresponding increases in the percentage of funds provided in the wholesale markets. This trend is likely to continue because of a continuation in a number of the factors discussed in the previous paragraph.

Farin FORESIGHT has inheritance capabilities that allow goals to be established on an overall basis. For example, the asset growth goal can be set to 8% per year. All accounts in the asset portion of the chart of accounts will inherit this goal, unless they are specifically overridden at the category or account level. As a result, a plan can be built quickly then tuned to deal with differential growth in different parts of the balance sheet. See Chapter VI for details.

1.d. Establishing Risk Tolerance Levels Developing standards for acceptable levels of tolerance for various kinds of risk is really a two part process.

1. Determine what kinds of measurement system will be used to measure risks. 2. Determine the Forecast/Policy limits given the measurement system selected in step 1.

An effective asset-liability system must at least address the following kinds of risk.

o Interest rate risk – the potential effect of changes in rates on income (income at risk) and economic value (value at risk). As indicated in this chapter, one of the objectives of an effective A/L process is to measure interest rate risk dynamically (in plans rather than old balance sheets). Tools for setting interest rate risk tolerance levels will be discussed in detail later in this chapter.

o Capital risk – the risk that the institution will find itself with inadequate capital to cover risks inherent in its balance sheet while meeting regulatory capital requirements. Capital limits are often incorporated in the capital goals established in the process of setting financial goals. Regulatory capital requirements specify both core capital requirements (generally measured by the capital/asset ratio) and risk-based capital requirements (risk-based capital divided by risk-


based assets. Some institutions specify a minimum percentage for one or the other ratio and may also specify a goal or target. Others may specify an acceptable range for one or both ratios.

o Liquidity risk – the risk that an institution will be unable to generate the cash needed to fund deposit outflows, existing credit commitments, or future loan growth. Historically, ratios that could be calculated from old balance sheets (like the loan/deposit ratio) became standard measurement tools for measuring and managing level of liquidity. With the securitization of the asset markets, and the ability of financial institutions to raise long-term borrowed money from sources like the Federal Home Loan Banks, these balance sheet based ratios are much less reliable as indicators of the level of liquidity in an institution. This is in spite of the fact the traditional liquidity ratios can be calculated from both current and future balance sheets. The trend in liquidity analysis is to measure to what extent an institution’s need for funds is covered by sources of funds -- in the context of the institution’s plan. Such a measure of liquidity is dynamic (evaluated in the context of a plan) and considers all sources and uses of funds. As this chapter was written, regulatory standards were just beginning to evolve in the area of dynamic sources and uses liquidity analysis.

o Credit risk - the risk that principal and interest associated with outstanding loans and deposits will not be repaid by borrowers and issuers of securities. Credit risk standards typically focus on leading indicators, asset quality indicators, and coverage indicators. A good example of a leading indicator might be 30 and 60-day delinquency ratios. Asset quality indicators might include non-performing loans as a percent of loans and loan losses as a percent of loans. Coverage indicators that provide a sense for the ability of the institution to cover loan losses include allowance for loan losses as a percent of non-current loans, total capital as a percentage of non-current loans and before tax net income as a percentage of net loan losses.

While all of these ratios have the potential to be calculated for current as well as forecast balance sheets, most asset-liability models fail to provide the ability to test for different levels of asset quality problems in the context of an institution’s strategy. As a result, these ratios are most commonly monitored in historical balance sheets in an attempt to evaluate historical trends, performance relative to goals and performance relative to peers. Institutions commonly set upper or lower limits for these ratios so they can identify when the institution is running into an asset quality problem. Interest Rate Risk Limits and Measurement Systems

Interest rate risk can be defined as the effect of changes in rate on income and on the market value of an

institution’s equity position. The effect of changes in rates on income is referred to as income at risk. While traditionally much of the focus on income at risk has been on changes to net interest margin caused

by changes in rates, in recent years the focus has shifted to the bottom line for at least two reasons.

o A change in rates can affect more than the net interest margin portion of the income statement. A good example is fee income from mortgages originated and sold in the secondary market. When rates fall, refinance activity is triggered, causing fee income from mortgage originations and sales to rise. When rates turn around and head back up, the refinance activity dies. Fee income from mortgage originations and sale falls.

o As pointed out in the previous section on establishing financial goals, it is at the bottom line that goals for profitability, growth and capitalization are balanced. It is entirely appropriate for a management team to ask, “How will a change in rates affect my ability to provide a reasonable return, maintain adequate capital, and grow?”

The effect of changes in rates on the market value of an institution’s equity position is referred to as value at risk. Depending on your regulatory agency, you may encounter a variety of terms being used to describe the difference between the market value of an institution’s assets and liabilities. For a number of years, OTS described this difference as the Market Value of Portfolio Equity (MVPE), then changed the term to Net


Portfolio Value (NPV). The tri-bank regulators refer to this value as Economic Value of Equity (EVE), while the NCUA uses the term Net Economic value (NEV). While there are accounting nuances that cause these calculations to be slightly different in some cases, all for terms represent essentially the same thing. – the difference between the market value of an institutions assets and its liabilities.

Because of their focus on an institution’s bottom line, management teams tend to place a greater emphasis on measurement systems that measure income at risk.

Why It Makes Sense for Management to Evaluate Both Income and Value at Risk As pointed out in earlier in this chapter, when analyzing the effect of changes in rates on income, institutions tend to model over horizons of two to three years. Take an institution with a significant percentage of its balance sheet in long-term instruments such as long-term fixed-rate mortgages or mortgage-backed securities funded by long-term FHLB advances. The income at risk modeling horizon fails to consider the significant cash flows that will come off long-term portfolios after the end of the three to five years. In other words, income at risk modeling fails to consider the interest rate risk in the tail end of these portfolios. A newly originated 30 year fixed-rate mortgage will still be 25 years away from maturity at the end of the modeling horizon.

As a result, an institution that is fairly well matched over the next few years would fly right through an income at risk analysis. But a significant mismatch beyond five years would fail to be picked up by the income at risk analysis.

In value at risk analysis, the market value of every financial instrument on the balance sheet is determined before and after rate shocks. In calculating the value, all the cash flows given off by an instrument in its remaining life are considered. While what is being calculated is value, value and income are two sides of the same coin. The reason the value of a financial instrument fluctuates in response to a rate shock is because its income stream doesn’t adjust immediately and fully to the change that happens to market rates. We’ll talk about this concept further later in the chapter. To summarize, value at risk analysis is not just an important regulatory tool. Used as part of the decision making process, it can also be an important management tool for assessing the long-term effects of changes in rate on an institution’s income stream.

Farin FORESIGHT can be used to evaluate both income at risk and value at risk. Value at risk calculations can be run on the current balance sheet or any forecast balance sheet. Discussions on setting up policies for income at risk modeling can be found in Chapter VI. Discussions on setting up assumptions for value at risk modeling can be found in Chapter VII. Both income at risk and value at risk reporting are discussed in Chapter VIII.

Static vs. Dynamic Evaluation of Interest Rate Risk

In the beginning of the chapter, we defined asset-liability management as the process of managing the relationship between risk and return. This statement implies that it is allowable for an institution to take on interest rate risk as long as:

o The institution stays within defined policy limits. o Increases in interest rate risk are rewarded with an appropriate increase in return.

So the process of optimizing the relationship between risk and return involves evaluating alternative strategies and choosing the strategy with the best risk/return tradeoffs. But here’s the rub. A large percentage of financial institutions evaluate their interest rate risk using measurement systems that focus on their most recent balance sheet. We refer to interest rate risk measurement systems that focus on the existing balance sheet as static measurement systems. But you can’t evaluate the risk/return tradeoffs associated with alternative


strategies by looking at the existing balance sheet because the risk is measured for only the strategy you implemented, not for the other strategies you might have considered or might be considering.

Static interest rate risk measurement systems are the most frequently used measurement systems by financial institutions. They are important to both the institution and to its regulators as they measure the current level of interest rate risk exposure in an institution. I’m not criticizing the use of static measurement systems. It is important for all institutions to be aware of their current interest rate risk exposure. But it is important that the limitations of static measurement systems be recognized. Because they reflect the results of only one strategy, the strategy that was implemented, they don’t provide management an effective decision making tool for evaluating risk/return tradeoffs. Dynamic measurement systems measure the interest rate risk in a plan or strategy. Modeling the plan or strategy provides the institution with the plan’s expected return by directly measuring performance using goals discussed in the goal setting section of the chapter. It also has the potential of measuring the level of interest rate risk resulting from the strategy. By modeling multiple strategies, a management team can gain a sense for the effect of each strategy on return as well as on interest rate risk. By comparing the resulting effect of each strategy on both risk and return, the institution is in a much better position to select the strategy that optimizes the relationship between risk and return. The bottom line here is if an institution’s primary objective is to measure the level of interest rate risk in their balance sheet, static measurement systems do the job. But if an institution wishes to use their interest rate risk and management expertise to select future strategies that optimize the relationship between risk and return, they need to select, implement and effectively utilize interest rate risk measurement systems that produce both static and dynamic analysis. So with the static verses dynamic issues in mind, let’s survey the different interest rate risk measurement tools available in the market.

Farin FORESIGHT supports all commonly used forms of static income at risk modeling including gap analysis, duration analysis, static income at risk testing, and testing of value at risk on the institution’s current balance sheet. Assumptions needed to perform these calculations are covered in Chapter V, VI, and VII. Static reports are covered in the reporting chapter, Chapter VIII. But Farin FORESIGHT was designed to make it easy to model interest rate risk dynamically based on the institution’s plan. Assumptions needed to perform these calculations are also covered in Chapter V, VI, and VII. Dynamic reports are also covered in the reporting chapter, Chapter VIII.

1.e. Interest Rate Risk Measurement Tools and Techniques Factors that Must Be Considered in Evaluating Interest Rate Risk .Income at risk is directly related to how quickly an institution’s yield on its earning assets responds to changes in market rates as compared the cost of its interest paying liabilities. The quickness of response is directly related to how soon the institution is allowed by contracts to adjust rates, and how much of a rate adjustment occurs at the adjustment point.

Repricing opportunities occur as fixed and adjustable-rate instruments throw off cash flows and as adjustable rate instruments reach contractual repricing points. Depending on the instrument, cash flows may be given off for one or more of the following reasons.

o The instrument matures – generally both a principal and interest cash flow occurs.


o A payment is due on an amortizing instrument. This payment typically includes principal and interest.

o An interest payment is due. o One of the parties to the instrument exercises an option that allows principal to be repaid early.

This option could be a prepayment option, a call option, or a put option.

Repricing opportunities on variable-rate instruments are usually specified in the instrument’s contract. In other instruments the amount of the rate adjustment may be driven by one or more of the following features which are specified in the contract.

o The index to which the adjustment of rates is tied. Examples are Prime, Fed funds, the 1 Year CMT Treasury, LIBOR, and the 11th District cost of Funds.

o The spread to the index. For example, a commercial loan might be priced at Prime plus 2%. o The adjustment frequency – Examples include daily, monthly, quarterly and annually. Hybrid

ARMs are often issued with an initial lock period (5 years) and annual adjustment periods thereafter.

o Annual and lifetime caps. For example, rate increases on an ARM might be limited to no more than 2% per year and 6% over the lifetime of the instrument.

In addition to affecting income at risk, these contractual features of a financial instrument also affect value at risk. That’s because when contractual features prevent an instrument’s rate to adjust quickly in response to changes in market rates, the market compensates by adjusting the instrument’s price or value in the market.

Of these issues, the options that allow a holder to pay off principal early and the restrictions on how fast rates on variable-rate instruments can be raised in response to changes in market rates (caps) wreak the greatest amount of havoc in evaluating a financial instrument’s interest rate risk. This option risk can materially affect the speed at which financial instruments respond to changes in market rates in one rate environment as opposed to another.

Let’s review the various measurement system options available to financial institutions and assess the extent to which each is successful in considering the factors just discussed in this section.

Farin FORESIGHT accurately models the income streams of a variety of complex financial instruments including fixed and adjustable rate mortgages, callable bonds and FHLB advances, CMOs and non-maturity deposits. Non-maturity deposits are complex in that their actual behavior is far different than their contractual term and repricing characteristics imply. See Chapter III for more information on instrument cash flow characteristics.

1.e.i. Gap Analysis

The Mechanics

Gap analysis is the method for measuring interest rate risk that has been in use the longest time by financial institutions. A gap report is designed to answer the question, “How soon will the institution have the opportunity to adjust the rate on its instruments the next time?” Figure 1-1 shows a simplified financial institution gap report.


Figure 1-1 – Gap Report

In displaying the answer to the question ‘how soon?’, the gap report is divided vertically into a number of future periods. In Figure 1-1, they are 0-6 months, 6-12 months, 1-3 years, and over 3 years from today. The major classes of financial instruments held by the institution are listed down the left column.

Balances are placed into the columns representing the periods based on how soon they throw off cash flows or in the case of variable-rate instruments, how soon they will be scheduled to reprice the next time. Gap reports only display principal. They ignore interest cash flows.

For example, the institution’s gap report in Figure 1-1 indicates the institution has $60 million in total investments (balances are in thousands). Of the $60 million, $20 million will reprice in the next six months, and another $20 million will reprice between six and twelve months from today. The remaining $20 million will reprice between one and three years from today. The entire $60 million portfolio will have repriced within three years.

The $20 million slotted in the 0-6 month column could be made up of maturing investments, amortization of principal for amortizing investments, early principal cash flows due to prepayments, calls, or puts, and repricing opportunities affecting variable rate investments.

On the report, loans, deposits, and borrowings have also had their repricing opportunities slotted into the appropriate gap buckets in a manner similar to that employed in slotting investments.

Totals are generated for assets repricing in each column (Total RSA) and for liabilities repricing in each column (Total RSL). The difference between the quantity of assets repricing and the quantity of liabilities repricing is the Gap. In the case of the institution shown in figure 1-1, a total of $25 million of assets are scheduled to reprice in the next six months as compared to $55 million of liabilities, resulting in a gap of -$30 million.

The gap is often shown individually for each period as well as cumulatively. For example, Figure 1-1 shows a 0-6 month gap of -$30 million, a 6-12 month gap of $0 and a 1-3 year gap of $25 million. The institution’s cumulative gap through 3 years is -$5 million.

In order to convert the gap into a ratio used in measuring the relative level of interest rate risk, the gap is often divided by total assets to compute the gap/asset ratio. The gap/asset ratio can be displayed for an individual period or on a cumulative basis. In Figure 1-1 the cumulative gap/asset ratio is -20% through six months and one year then drops to -3.3% at three years.


The Theory

Most institutions that continue to use gap analysis as an interest rate risk measurement tool focus on the gap/asset ratio. The gap/asset ratio is designed provide the institution with two key pieces of information.

1. The direction of the interest rate risk bet. Positive gaps indicate asset sensitivity. The assets are repricing more quickly than the liabilities. In a rising rate environment, yield should rise faster than cost of funds. Profits should increase. In a falling rate environment yield will fall faster than costs of funds. Profits should decline. Negative gaps indicate liability sensitivity. The liabilities are repricing more quickly than the assets. In a rising rate environment, yield should rise more slowly than cost of funds. Profits should decrease. In a falling rate environment yield will fall more slowly than cost of funds. Profits should increase.

2. The size of the interest rate risk bet. The larger the gap/asset ratio, the larger the mismatch in relation to total assets. With small gap/asset ratios institutions will see moderate fluctuations in income as rates change. With large gap/asset ratios institutions will see more significant fluctuations in income as rates change.

Limitations & Shortcomings

Gap analysis was primarily designed as a tool for gaining a sense for income at risk. That should be obvious from the fact that the time periods into which the cash flows are slotted in Figure 4-1 are much narrower for short-term cash flows than for the longer-term cash flows. Granted, Figure 4-1 is a simplified report. Most gap reports sort the cash flows into a much larger number of repricing periods and break major categories (investments, loans, deposits, and borrowings) into much greater levels of detail. Yet most gap reports devote their detail to short-term cash flows.

Within a cash flow column, gap analysis assumes assets and liabilities average to roughly the same maturity. That could be a highly inaccurate assumption for a potentially wide column like the Over 3 Year column in Figure 1-1. An institution could fund 30 year fixed-rate mortgages with four year CDs. Because both would fall into the Over 3 year column, the model would assume no interest rate risk as long as the quantity of assets was roughly equal to the quantity of liabilities. But in fact, there could be a huge mismatch leading to significant value at risk.

Beyond its income at risk focus, gap has a number of other shortcomings.

o It fails to take into consideration how much the instrument might reprice at the repricing

point. For example, a 1 year ARM with an annual cap of 1% will reprice in the next year. But the rate will rise no more than 1% in spite of the fact market rates may have increased 3%.

o The gap report fails to consider the effect of changes in rates on principal cash flows of instruments with imbedded prepayment, call, and put options. For example, a gap report will show the expected prepayment cash flows coming off a mortgage portfolio in a flat or expected rate environment. Gaps will be used to predict the effect of changes in rates on income. Yet a drop in rates will accelerate prepayments on the mortgages causing them to reprice much more quickly than the gap report predicts.

o A gap report fails to consider basis risk. Let’s say a commercial loan that resets annually off prime is funded with a 1 year CD that is priced based on 1 year Treasury


rates. The gap report says, “No problem”. But the problem could be significant if the 1 year Treasury moves by a significantly different amount than prime during the year.

o It is not a good tool for evaluating risk/return tradeoffs. While a gap report attempts to measure the level of interest rate risk it provides no guidance as to return.

It is no surprise that the Office of Thrift Supervision yanked its entire supervisory case load off gap analysis as the primary interest rate risk measurement tool at the end of the 1980s. Thrifts have big portfolios of long-term instruments and are loaded with option risk. Forward thinking banks and

credit unions have also moved on to more sophisticated measurement tools in large numbers with and without pressure from their primary regulator.

Farin FORESIGHT produces current and Forecast Gap Reports. See Chapter VIII Section 1.d. for details.

1.e.ii. Duration Analysis After dealing with all the repricing columns in a gap report you might be thinking to yourself, “Wouldn’t it be nice if someone could come up with a single number that represented the interest rate sensitivity of an instrument, a portfolio of instruments, or even the entire balance sheet?” Duration analysis attempts to provide such a measure. Of course, in order to effectively calculate the a single index representing the rate sensitivity of the cash flows of a single instrument or portfolio, we need to consider all the issues identified in the factors identified earlier in this chapter – maturity, amortization, imbedded options and contractual repricing characteristics.

The Mechanics Calculating an instrument’s duration is a multi-step process. o Step 1 – List the amount and timing of the instrument’s expected principal and interest

repricing cash flows. o Step 2 – Calculate the net present value of the cash flows listed in Step 1. o Step 3 – Weight the present value of the cash flows in step 2 by the time of receipt of the

cash flows. o Step 4 – Calculate the duration by dividing The Weighted NPV (total from step 3) by the

NPV (total from step 2). We could go through an example of a duration calculation, but most A/L models perform duration calculations and produce duration reports. More important than the duration is how we use the information once duration is calculated. The following formula uses duration to convert changes in rate into changes in value. The example calculates the change in value for a $6,000 loan with a duration of 3 years assuming a 2% increase in market rates. ChgMV = - D x ChgMR = - 3 x 2% = - 6%

The Theory If we can calculate the duration of an individual financial instrument, we can do so for a portfolio of financial instruments like our 30 year fixed-rate mortgages or a broader portfolio like an investment portfolio. It we can calculate duration of assets, we can also calculate duration of liabilities. Of course, the bigger the portfolio, the more cash flows are being evaluated. These calculations beg for automated solutions.


Once the duration of an institution’s assets and the duration of its liabilities have been calculated, the duration of its capital can also be calculated. Knowing the duration of its capital allows an institution to estimate changes in portfolio equity caused by a change in rates. As a result, duration was the first of the value at risk measurement systems adopted by the market. Financial institutions often set interest rate risk policy limits using duration for individual portfolios or for the entire institution.

Limitations & Shortcomings o Because it takes all the cash flows given off by a financial instrument and compresses

them into a single number duration is not a very good tool for evaluating the effect of short-term changes in rates on income.

o Duration, like gap, fails to take into consideration how much a variable-rate instrument might reprice at the repricing point.

o Duration fails to consider the effect of changes in rates on principal cash flows of instruments with imbedded prepayment, call, and put options.

o A duration analysis fails to consider basis risk. o Duration is not a good tool for evaluating risk/return tradeoffs. While duration attempts

to measure value at risk, it provides no guidance as to return. In spite of its shortcomings, duration is a useful tool in gaining a sense for instrument and portfolio price sensitivity. Like gap analysis, duration provides us with useful diagnostic information. But it should not be your institution’s primary measurement system for measuring value at risk.

Farin FORESIGHT produces duration reports in both current and forecast balance sheets. See Chapter VIII Section 2 for details.

1.e.iii. Dealing with Option Risk While gap and duration are subject to a number of shortcomings, those causing the greatest amount of inaccuracy occur when we start evaluating the performance of instruments with imbedded options – prepayments, calls, puts, annual interest rate caps, lifetime rate caps, etc. For both gap and duration, we must project the amount and timing of the instrument’s cash flows. With duration, we need both principal and interest cash flows. Once the cash flows are projected, gap ratios and duration can be calculated. So far, so good. Where we run into problems is when we use these calculated gaps and durations to predict the effect of changes in rate on income or value. Take duration as an example. Assume we are attempting to assess the price volatility of a 15 year fixed-rate mortgage with a duration (considering amortization and prepayment) of 4 years. The following simple formula predicts the effect of a change in rates on the value of the mortgage for a change in rates of 3% up and for 3% down. ChgMV+3% = -D x ChgMR = = -4 x 3% = -12% Chg MV-3% = -D x Chg MR = -4 x -3% = +12% These calculations assume duration and change in rates are two independent variables. Changing one doesn’t effect the other. But as the printout from Bloomberg in Figure 1-2 shows, this isn’t the case.

Figure 1-2 - Effect of Option RiskRate Chg 3 Yr Treas 15 MBS 7/1 Bond

-3% 12% 4% 3%-1% 4% 2% 1%0% 0% 0% 0%1% -4% -5% -6%3% -12% -16% -18%


Figure 1-2 compares changes in value for a 4 year strip Treasury, a 15 Year fixed-rate MBS and a 7 year Bond with a one year call under rate shocks of 100 and 300 bp in both directions. Note that the response of the Treasury with a 4 year duration is linear. The above duration formula properly predicted what rate changes would do to value. Duration and change in rates are independent variables for this instrument. So what happened with the MBS? In falling rate environments prepayments accelerated, shortening duration from 4 years to 2 years with a 100 bp drop in rates and to 1.33 years with a 300 bp drop in rates. When rates rose, prepayments slowed causing duration to extend to 5 years with a 100 bp increase in rates and 5.33 years for a 300 bp runup in rates. Duration is not an independent variable for an instrument with imbedded prepayment options. Duration is dependent on changes in rates. The duration formula would have predicted that the MBS would change in value in the same way as the Treasury in the first column. Also consider what happens to the value of the 7 year bond with a 1 year call. In a rising rate environment, the bond has a duration of six years based on its maturity and its interest cash flows. In a falling rate environment the bond will be called by the issuer who will take advantage of the opportunity to reset their debt to a lower rate. The value increases based on the 1 year duration to call. A view of a graph in figure 1-3 comparing the changes in values for the three instruments, underscores some of option risk’s less than favorable effects.

Figure 1-3 - Option Risk

-20%-15%-10%-5%0%5%

10%15%

-3% -2% -1% 0% 1% 2% 3%

Change in Rates

Cha

nge

in V

alue

3 Yr Treas15 MBS7/1 Bond

As might be suspected the value of the Treasury tracks along a straight line. Both the MBS and the callable bond show larger declines in value than the Treasury for unfavorable moves in rates. And both the MBS and the callable bond show much smaller increases in value than the Treasury for favorable movements in rates. This graph gives rise to two of the three ‘Murphys Laws’ as it relates to taking on option risk.

1. The pain associated with an unfavorable movement in rates is greater than the gain associated with a favorable movement in rates.

2. Often, as is true of the MBS, the pain from a 200 bp unfavorable movement in rates is more than twice as great as the pain from a 100 bp unfavorable movement in rates. This is because prepayments slow even more with a 200 basis point rise in rates than with a 100 bp rise in rates.


A third negative effect of option risk appears when many institutions attempt to hedge the interest rate risk in their balance sheet without hedging the option risk. In Figure 1-4, the four year initial duration of the MBS is hedged with a FHLB Advance with an identical duration. As the graph shows, value of the advance (with no imbedded options) tracks along a straight line in response to changes in rates. The MBS gains value slowly in rising rate environments while falling more quickly in rising rate environments. Netting the changes in the value of the liability against the changes in the value of the asset nets to the Change in NEV (net economic value) caused by the transaction (red line). Economic value drops with rate shocks in either direction, giving rise to the third Murphy’s law relating to option risk.

3. About the time you hedge the interest rate risk in an institution with significant option risk, no matter which way rates move, you lose.

Figure 1-4 Change in Net Economic Value

-20%-15%-10%-5%0%5%

10%15%

-3% -2% -1% 0% 1% 2% 3%

Change in Rates

Cha

nge

in V

alue

15 MBS4 AdvanceChg NEV

Financial institutions of all kinds are carrying more option risk today than a decade ago. This occurs as a result of one or more of the following reasons.

o In reaching for yield in investment portfolios, institutions are purchasing more callable bonds and mortgage-backed securities.

o Non-thrifts are increasing their portfolio concentrations of fixed and adjustable rate mortgages.

o Financial institutions are taking advantage of lower cost wholesale funding that gives the FHLB the option of putting an institution’s debt back to the institution when they can re-lend the funds at higher rates.

Given what you’ve seen in this section, you may ask why would a financial institution take on any option risk at all?

o Yields on investments containing options are higher than those not incorporating the options. For example, the yield on the MBS in Figure 5-3 was 6.5% while the yield on the comparable duration Treasury was only 5%.

o Costs of FHLB Advances that give the FHLB the option of putting the debt back to the institution have been considerably lower than advances containing those options.

o Adjustable rate mortgages lacking annual and lifetime caps are difficult to market to consumers.

o Fixed-rate mortgages with no prepayment option at all or that require payment of a penalty to exercise the option are not salable in the conventional secondary markets and are unacceptable to consumers and, in some states, illegal.


So an institution has little choice but to take on some option risk. In many cases, the decision to take on option risk is a matter of the institution assessing risk/return tradeoffs. Given the favorable affect on yield or cost of funds, is the institution being properly compensated for taking on option risk. To get answers to questions like these requires that an institution use a measurement system that:

o Properly calculates the effect of imbedded options on principal and interest cash flows as rates change.

o Allows the institution to model risk/return tradeoffs in a dynamic setting to determine the effect of the proposed strategy involving taking on option risk delivers a return that effectively compensates the institution for its risk.

o Allow the institution to assess whether incurring the option risk will take it outside its interest rate risk policy limits.

The interest rate risk measurement tools described in the following section are designed to provide answers to the above issues and questions. Farin FORESIGHT is designed so output reflects the effects of options typically imbedded in complex financial instruments including prepayment speeds that vary with rate environments, calls that occur only under specific rate scenarios, annual and lifetime caps on rate movements, decay rates and lagging pricing on non-maturity deposits, etc. Chapter III covers defining complex cash flow characteristics for financial instruments. Chapter V covers external assumptions like prepayment tables and decay rates. It also covers current and lagging rate drivers, rate shocks that are both immediate and permanent and gradual, and parallel and non-parallel shifts in the yield curve. Chapter VI covers policies including pricing and target assumptions.

1.f. Using an Asset-Liability Model to Perform Income at Risk Testing If you’ve made the right investment in an asset-liability model, feed it with good data and assumptions, and know how to use it effectively, you are well set up to perform income at risk testing at the same time as you are developing your institution’s plan. Better asset-liability management models get around the major limitations of gap analysis as an income at risk analysis tool. But before making this claim about the model you are using, you need to obtain the answers to the following questions.

• Is the model downloading sufficient information about the financial instruments in your portfolio to model them accurately.? For example, does the download include annual and lifetime caps, reset dates, indexes and spreads on ARMs? Does the download include the call dates and strike prices on callable bonds?

• As part of setting up line items in the chart of accounts are their cash flow characteristics defined? This is important to fill any holes in the information on instruments being downloaded and to allow the model to properly calculate cash flows on products originated as part of the plan?

• Is there a pricing function that allows you to define accurately how new accounts will be priced and existing accounts repriced as rates change? Does the model provide you with the flexibility of tying new account pricing and repricing to market rates most relevant in pricing these accounts.

• Is there a function that allows you to specify origination volume or total balance targets that you expect to hit as part of the plan? Are new originations modeled as financial instruments in the same way as existing instruments?

• Is there a prepayment function that provided the model with the capability of varying prepayment speeds by rate environment?

• Is there a function for modeling the cash flows and repricing behaviors of non-maturity deposits?


• In setting up rate environments is it possible to model immediate and permanent as well as gradual rate shocks? Can basis risk be modeled by establishing rate environments involving non-parallel shifts in the yield curve and by moving different market rates by different amounts?

Properly used, models with the above capabilities get around the major limitations of gap analysis as an income at risk tool.

• Repricing behavior of variable rate instruments is properly modeled by resetting rates at contractual repricing points spread off appropriated indexes. Annual and lifetime caps are enforced.

• Principal cash flows of instruments with imbedded options are properly modeled. Prepayment speeds on loans automatically vary with rate environments. Calls on callable bonds and puts on FHLB advances are automatically executed in rate environments where the option is ‘in the money’. Cash flow windows of CMOs are automatically adjusted to reflect their performance in different rate environments.

• Basis risk is taken into consideration when a strategy is tested through non-parallel shifts in the yield curve.

• Good A/L models can be used for both static and dynamic income at risk analysis. o Static income at risk is modeled over a planning horizon by holding the balance sheet

constant. o Dynamic income at risk is modeled by developing a plan or strategy that changes the

composition of the balance sheet over time and modeling the strategy through multiple rate environments.

• A good A/L model makes it much easier to evaluate risk/return tradeoffs. o First of all because the model produces a balance sheet, income statement, and key ratios

for each rate environment, income at risk as measured by net income, ROA, ROE, or earnings per share fluctuations can be directly measured.

o By comparing the results of multiple strategies through multiple rate environments, both compliance with policy limits and risk/return tradeoffs can be assessed.

Farin FORESIGHT meets all of these requirements. They will be addressed in greater detail later in this chapter. Establishing Income at Risk Limits When Using an A/L Planning System In Chapter 3 we discussed the process and importance of establishing long-range financial goals. For stock shops we recommended ROE and/or EPS. For mutuals and credit unions, we recommended ROE. Once your goals have been established, you merely need to ask yourself and your stakeholders the question, “How much of our ROE (or EPS for stocks) are we willing to place at risk in adverse rate environments. How the question is answered by stakeholders will vary somewhat depending on whether the institution is a stock or a mutual.

Stock Institutions Stockholders have a much easier time understanding income at risk limits based on fluctuations in ROE or EPS than they do understanding limits expressed as acceptable variances in the gap/asset ratio. They also have an easier time arriving at the amount of income at risk that will be acceptable. That’s because stockholders expect management to deliver an acceptable return regardless of the rate environment. In fact, once a stock institution bases its income at risk limits on ROE or EPS, they find the stockholders are much tougher in establishing income at risk limits than regulators. It takes a whole lot less income fluctuation to make a stockholder unhappy than it takes to raise safety and soundness issues with a regulator.


For example, say your goal for ROE is 15%. Would your stockholder accept a 50% drop in income in an adverse rate environment to a ROE of 7.5%? Probably not. How about a 33% drop to 10%? Possibly. A 20% drop to 12%? They may well find the 20% fluctuation to be reasonable. In this case, the institution’s policy limit may be as follows.

• Our income at risk limit is a 20% fluctuation in ROE from our plan’s expected rate environment to any adverse rate environment modeled.

Mutuals and Credit Unions

There are two reasons why mutuals and credit unions need to produce net income.

1. To grow the numerator of the capital ratio (capital) at a speed that allows it to keep up with the growth in the denominator (assets).

2. To receive an acceptable rating for the ‘E’ (earnings) in CAMEL(S) from their regulator. Given that ROE is the growth rate of the capital account for mutuals and credit unions, in defining acceptable fluctuations in ROE, the institution is answering the following questions.

1. How much of a slowdown in our ability to grow or a drop in our capital/asset ratio are we willing to accept in an adverse rate environment?

2. How much of a drop in the ‘E’ rating in CAMEL(S) are we willing to accept in an adverse rate environment?

For example, say a credit union has set a goal for ROE of 10% to match their expected asset growth rate of 10%. In establishing the limit on fluctuations in ROE they might consider a 50% drop to 5%. Let’s assume their Equity multiplier is 10. So a 50% reduction in ROE would reduce ROA from 1.0% to 0.5%. Say the regulator would deliver a 1 rating for E with a 1% ROA but only a 3 rating with a 0.50% ROA. Would management accept that big a drop in the rating as well as be willing to cut annual growth from 10% to 5% in an adverse rate environment? Probably not. How about a 25% fluctuation which would require a cut back in growth from 10% to 7.5% and cause a drop in the “E’ rating from 1 to 2? Maybe so. So in this example, the credit union or mutual’s policy limit would outwardly appear to be very similar to the stock institution’s. Yet the thought process that resulted in the limit would be significantly different.

• Our income at risk limit is a 25% fluctuation in ROE from our plan’s expected rate environment to any adverse rate environment modeled.

Optimizing the Relationship Between Risk and Return So now a goal for ROE has been established as well as a policy limit built around fluctuations in ROE. In developing strategies, we’re looking for a strategy that offers the best risk/return tradeoffs while staying within the prescribed limit. Figure 1-5 contains an example of a matrix comparing the result of three alternative strategies in three rate environments.

Figure 1-5 – ROE Decision Matrix Rate Env

Goal/ Limit

Strategy #1

Strategy #2

Strategy #3

Strategy #4

Rising -20% 10.0% (0%) 20.0% (+33%) 16.0% (+14%) 17.2% (+19%)Expected 15% 10.0% 15.0% 14.0% 14.5% Falling -20% 10.0% (0%) 10.0% (-33%) 12.0% (-14%) 11.8% (-19%) Management has established a goal for ROE of 15% in the Expected rate environment.. Income at risk limits allow for declines in ROE of up to 20% in rising and falling rate environments. Setting limits to allow


declines of up to 20% in both rate environments gives management the prerogative on betting on rising or falling rates as they wish. But the size of the bet is limited to 20% of expected rate environment ROE. The results of the runs of three different strategies for operating the institution are shown.

1. In strategy 1 all interest rate exposure has been removed from the institution’s balance sheet. ROEs are the same regardless of rate environment. But at 10%, ROE is only running at 2/3 of the institution’s 15% goal. Interest rate risk is totally under control, but at a significant cost to return.

2. Under strategy 2, the institution’s ROE goal of 15% is achieved in the expected rate environment. But with a 33% reduction in ROE in the Falling rate environment, income at risk is outside the -20% policy guideline at 33%. While it could be argued that this strategy is unacceptable because of income at risk, note that the institution’s performance in the Falling rate environment is no lower than any of the ROEs in Strategy #1. Having to choose between Strategy 1 and strategy 2, I’d be inclined to go with Strategy 2 and start working on the argument I’d use with the regulator in defending my decision.

3. At a ROE of 14% in the expected rate environment, Strategy 3 comes relatively close to achieving the ROE goal, while limiting income at risk 14%, well within the 20% guideline. Strategy 3 is an easy choice over Strategy 1 in that performance is superior in all rate environments.

4. ROE for Strategy 4 in the expected rate environment is even higher at 14.5%. But income at risk is also higher at 18%, still within the 20% guideline. Strategy 4 is an acceptable alternative to Strategy 3 and outperforms Strategy 1 in all rate environments.

So far we haven’t said anything about the strategies themselves. But based on their results, the choice is between Strategy 3 and Strategy 4. Which one would you choose? The choice is not clear. Strategy 4 offers a slightly higher return at slightly more risk. But we need to look at more than the numbers in deciding between available strategies, although the numbers eliminated two of our four choices. There are a number of qualitative issues that must be addressed before choosing between the two finalists.

• How well do the two strategies fit into our business plans and operating strategy. • Do we have the expertise on staff to execute the strategy? • What is our level of comfort with the assumptions built into the strategy? • What are the implications on credit risk, liquidity risk, and capital risk.

Once the answers to these questions are known, the institution is likely to choose either Strategy 3 or Strategy. Farin FORESIGHT has features that make it easy to model risk/return tradeoffs in alternative strategies. Comparative reports compare results on key ratios when multiple policies are run in multiple rate environments. This feature is discussed in Chapter VIII Part 4.

1.g. Using an Asset-Liability Model to Perform Value at Risk Testing Most financial institution executives look upon value at risk testing as a necessary evil. “We do it because our examiner requires that we do it.” Yet, as pointed out earlier in this chapter, value at risk testing gives a good sense for the long-term effect of changes in rate on income. That’s an important input to the decision making process as many institutions are currently booking assets (and funding) with much longer maturities than the modeling horizon they use for income at risk testing. This is particularly true for institutions contemplating the retention of long-term fixed-rate mortgages in their loan portfolio. Assume your modeling horizon for income at risk testing is three years. Should you decide to portfolio a 30 year fixed-rate mortgage as part of the three year strategy being modeled, the mortgage will have at least 27


years remaining to maturity at the end of your income at risk horizon. “What about the interest rate risk in the 27 year tail?”, you might ask. We can’t extend the modeling horizon to 30 years and make assumptions as to how we will run our institution for the next 30 years. But we can run a value at risk test on either a current or forecast balance sheet and gain a sense for the effect of changes in rates on income over the remaining life of the instrument. “How so?”, you might ask. Let’s look at a value at risk test on individual loan. Value at Risk Test – Single Instrument Our $6,000 loan calls for six annual principal payments of $1,000 along with annual interest payments on the loan’s remaining balance at a coupon rate of 10%. In the first step we list the amount of timing of the loan’s principal and interest cash flows over its remaining life. Figure 1-6 shows these cash flows in the second column. In the second step a discount rate is used to mark these cash flows to market. The results are shown in the 0 bp column of Figure 1-6.

Figure 1-6 – Single Instrument Value at Risk Period Payment -200 bp -100 bp 0 bp +100 bp +200 bp1 1,600 1,481 1,468 1,455 1,441 1,4292 1,500 1,286 1,263 1,240 1,217 1,1963 1,400 1,111 1,081 1,052 1,024 9964 1,300 956 921 888 856 8265 1,200 817 780 745 712 6816 1,100 693 656 621 588 557Total 8,100 6,344 6,168 6,000 5,839 5,685Disc Rate 8% 9% 10% 11% 12%

In a value at risk calculation we re-project all the cash flows for all the other rate environments we want to model. In Figure 1-6 we’re assessing the effect of changes in rates on value for five different rate environments. Because we’re modeling a single financial instrument and anticipate no prepayments regardless of rate environment, Figure 1-6 shows only a single set of cash flow projections. We then apply a discount rate reflecting the effect of rate shocks on market values of cash flows for each of the five rate environments using the following formula. NPV = FV / (1 + i)n Assume today’s rate for a comparable loan is 10%. An immediate and permanent 200 bp increase in rates would raise the market rate to 12%, the discount rate used in the 200 bp column. Net present value for the $1,400 cash flow in year 3 with a 12% discount rate is $996. NPV = FV / (1 + i)n = 1400 / (1 + .12)3 = $996 The present value of the individual cash flows are summed to obtain the market value of the instrument for each of the five rate shocks. By comparing the values for the five rate environments we can see how much of the instrument’s value is at risk for the different rate shocks. You might ask yourself, “Why the focus on value when my primary interest is the effect of changes in rates on income?” The reason the value of a financial instrument fluctuates is because its income doesn’t in response to changes in rate. Take the loan in Figure 1-6 with a duration of roughly 3 years for example. When market rates move from 10% to 11%, the average amount of time we have to rate to adjust the rate on this 6 year loan is approximately 3 years. The adjustments happen as the cash flows come in.


So let’s look at the loan through the eyes of a potential acquirer. Once the rates have moved to 11%, if he acquires this loan, he’ll want to earn 11%. Under the contract, the borrower would pay him 10%. Who comes up with the other 1%? As the seller, you do. You’ll sell the loan for approximately 97% of book. The buyer will book the loan and collect his 10% from the borrower. He’ll also amortize his 3% discount over the 3 year duration of the loan, picking up another 1% per year, raising his yield to 11%. Market value risk and income risk opposite sides of the same coin. In effect, the change in market value is nothing more than the net present value of the income premium offered by the loan in a falling rate environment or the income discount in a rising rate environment, calculated over the remaining life of the instrument. Because it considers all the cash flows coming off an instrument for the rest of its life, value at risk analysis is the only tool in our arsenal for calculating the interest rate risk in what’s left of any instrument in our balance sheet after the end of whatever income at risk modeling horizon we are modeling. As you’ll see as we work through the rest of this chapter, it has an important role to play in choosing between alternative strategies based on risk/return tradeoffs. “What’s the big deal with the discounted cash flow calculations shown in Figure 1-6?”, you might ask. “I could have gotten the same values by multiplying duration by change in rates.” The value comes in when the technique I have just described is used to calculate values of instruments with imbedded options. Take a $10,000 5 year bond that provides the issuer with an option to call the bond after 1 year. To simplify the example, we’ll assume the bond is newly issued and pays interest annually at a rate of 6%. Assuming current rates for similar instruments are 6%, this bond will be called in falling rate environments but will run to maturity in flat and rising rate environments. Figure 1-7 shows cash flows to maturity and cash flows to call.

Figure 1-7 – Value at Risk – 5/1 Callable Bond Year Maturity To Call -200 BP -100 BP 0 BP +100 BP +200 BP

1 600 10,600 10,192 10,095 566 561 5562 600 0 0 534 524 5143 600 0 0 504 490 4764 600 0 0 475 458 4415 10,600 0 0 7,921 7,558 7,214

Total 13,000 10,600 10,192 10,095 10,000 9,590 9,201Disc Rate 4% 5% 6% 7% 8%

Because the bond is called in falling rate environments, there is only a single cash flow to value, the $10,600 principal and interest payment at the end of year one. Because the bond isn’t called in flat and rising rate environments, there are five cash flows to value, four annual $600 interest payments and the final $10,600 principal and interest payment scheduled at maturity in year five. As can be seen from the changes in values, in rising rate environments there is nearly a 4% drop in value for each 100 basis point increase in rates. Yet in falling rate environments there is less than a 1% increase in value for each 1% increase in rates. That’s because the duration of this instrument is approximately 1 year when it is called, but approximately 4 years when it goes to maturity. The effect of option risk becomes even more apparent when the percentage changes in value for 6 year loan with its 3.015 year duration is graphed against the callable bond in Figure 1-8.

Figure 1-8 – Comparative Changes in Price


85

90

95

100

105

110

-200 -100 0 100 200

6 Yr Loan6/1 Bond

As can be seen from the graph of the price on the loan (100 being par), gains in value in falling rate environments mirror losses in value in rising rate environments. But the bond loses value more quickly than the loan in a rising rate environment because of its longer duration (4 rather than 3 years). In falling rate environments the bond gains value more slowly than the loan, because it is called in 1 year dropping its duration to one year. Discounted cash flow analysis is the technique that was used in calculating the effects of changes in rate on value in Figures 1-6, 1-7, and 1-8. It overcomes duration’s shortcoming in assessing option risk by recalculating cash flows for each rate environment being calculated. As a result, it is an accepted methodology for small to medium sized financial institutions in assessing the effect of options on value at risk. Farin FORESIGHT’s valuation reports can be produced at chart of account line item level of detail. Outputs include value, price and duration. See Chapter VIII Section 2 for details. Value At Risk - Institution Once a methodology has been established for calculating value at risk at the instrument level, rolling this information up to assess value at risk for the institution is a matter of addition, subtraction, and the calculation of a few ratios. Figure 1-9 is a value at risk calculation performed on an institution’s current balance sheet.

Figure 1-9 – Institution Current Value at Risk Value -200 BP 0 BP +200 BP

Total Assets 350,000 367,803 356,422 341,656Total Liabilities 317,100 321,272 314,163 308,507Portfolio Equity 32,900 46,531 42,268 33,149Portfolio Equity Ratio 9.45% 12.70% 11.78% 9.51%Rate Sensitivity (BP) 0 93 0 -227Equity Exposure 0.00% 10.09% 0.00% -21.57%

The book value column shows total assets, liabilities and capital as they appear on the institution’s accounting statements. Of course, most assets and liabilities are carried at historical costs, not at their market value. The institution’s book capital/assets ratio is 9.45%. The O BP column shows the current market value of the institution’s assets and liabilities. These numbers are calculated by summing the market values of its individual assets (like those in Figures 1-6 and 1-7). For the institution illustrated in Figure 1-9, the current market value of liabilities ($356.422) million is above book ($350.000 million). It would come as no great surprise that the market value of assets of this institution are


above book when you realize the values were calculated after the institution had been through a falling rate environment. When rates drop, asset values rise. Market values of liabilities are also summed resulting in total market value of all liabilities. It is interesting to note that while the market value of the institution’s assets are above book, the market value of its liabilities ($314.163 million) are below their book value ($317.100 million). This would come as a surprise to some as the market value of liabilities also rise in a falling rate environment. But there is another factor other than movement in rates that could cause values of assets and liabilities to be other than book. In this case, to the extent the institution is successful raising funds at less than wholesale market rates, they book liabilities that have market values below book. Raising funding at market values below book is a good thing in that the market value of liabilities are subtracted from the market value of assets in computing the market value of equity (Portfolio Equity). On the loan side of the balance sheet, institutions that book well priced loans also create value. The market value of the loans are above book at time of origination. As just mentioned, Portfolio Equity is the difference between the market value of assets less the market value of liabilities. The institution in Figure 1-9 has a market value of its equity ($42.268 million) well above book ($32.9 million). That occurs as a result of its assets carrying a market value above book while its liabilities carry a market value below book. The Portfolio Equity ratio (11.78%) is calculated by dividing Portfolio Equity by the market value of assets. Portfolio Equity Ratio = Portfolio Equity / MV Assets = 42,268 / 356,422 = 11.78% In Figures 1-6 and 1-7, we applied immediate and permanent rate shocks to the value of two financial institution assets to assess the effect of rate shocks on their values. Adding the market values of the of all the assets after a 200 basis point increase and decrease in rates would result in the asset values before and after the rate shock for the institution depicted in Figure 1-9. The same summation was performed for its liabilities. It should be no surprise that the value of the institution’s assets and liabilities fall in rising rate environments and rise and rise in falling rate environments. In and of itself, this is not a cause for concern. But the relative speed at which assets and liabilities respond to changes in rates is a cause for concern as inconsistency in speed of asset and liability reaction can cause changes in portfolio equity. The change in portfolio equity caused by the rate shock is the ‘value at risk’ we wish to measure. In the case of the institution depicted in Figure 1-9, portfolio equity increases in falling rate environments and decreases in rising rate environments. The institution is liability sensitive over the life of the instruments in its balance sheet. Because cost of its liabilities responds more quickly to changes in rates than the yield on its assets, the institution’s liabilities show less value fluctuations than its assets. In the 200 basis point rising rate environment, value of assets drops more quickly than liabilities causing portfolio equity to drop from $42.268 million to $33.149 million. On the other hand, as rates rise, value of assets increase more quickly than liabilities, causing portfolio equity to increase from $42.268 million to $46.531 million.


Note that for this institution, the drop in Portfolio Equity from an unfavorable movement in rates ($9.119 million) is far greater than the gain from a favorable movement in rates ($4.263 million). This is an indication of unhedged option risk in the balance sheet. Should the institution be concerned about this level of value at risk and the level of option risk illustrated in these numbers? The Office of Thrift Supervision has required all institutions they regulate to calculate their value at risk on a quarterly basis since the late 1980s. At the time this book was written, OTS was the only agency that had gone so far as to publish a table showing how they would derive the quantitative rating for the S (Sensitivity) in their CAMELS rating. Figure 1-10 illustrates this table. Note that the NPV ranges listed in the left column are revised by OTS from time to time. Check with their Web site for the most recent guidelines. OTS looks at two factors in coming up with the quantitative Sensitivity rating.

1. What is the lower of the two Portfolio Equity Ratios after immediate and permanent 200 bp rate shocks in both directions. In the case of the institution in Figure 1.10, Portfolio Equity drops to 9.51% with a 200 basis point increase in rates. The post-shock ratio falls into the 8-12% range in the left column in Figure 1.10.

2. How much of a drop happened to this ratio from the 0 BP rate environment to the worse case rate environment (measured in basis points). For the institution in Figure 1-9, the drop in the portfolio equity ratio caused by a 200 BP increase in rates was 2.27% or 227 basis points. 227 basis points of sensitivity would put the shop in the 200-400 bp column at the top of the table.

Under the OTS thought process an institution with a high portfolio equity can afford to take on more interest rate risk (show more sensitivity) than an institution with a low portfolio equity ratio. A quick look at the table in Figure 1-10 verifies this fact. We’ve color coded the various S ratings shown in the table to make this more obvious. For example, an institution with a post shock portfolio equity ratio between 4% and 8% needs to keep sensitivity at or below 100 bp to receive a 1 rating. On the other hand, an institution with a post shock portfolio equity ration over 12% qualifies for a 1 rating as long as sensitivity is under 400 bp. In the case of the institution in Figure 1-9, with a post-shock portfolio equity ratio of 9.57% and sensitivity of 227 bp, they would be a 2 rated shop for the S in CAMELS (Figure 1-10). Whether they should be concerned about this level of interest rate risk depends on:

• Their general feelings relative to the importance of a 1 rating. Many well run shops receive a 2 rating in one or more categories.

• The effectiveness of their interest rate risk management program. The 2 rating received by this shop for Sensitivity is the quantitative portion of the rating. OTS and other regulatory agencies also perform a qualitative assessment on the quality of the institution’s interest rate risk practices. If they’ve installed an effective interest rate risk management process, given that their sensitivity is only slightly above 200 basis points, they may still receive a 1 overall rating for Sensitivity from their examiners.

• The level of credit risk, capital risk, and liquidity risk in the institution. An institution with low levels of liquidity, capital and credit risk can afford to take on a relatively higher level of interest

Interest Sensitivity Measure Post-Shock NPV Ratio 0-100 bp 100-200 bp 200-400 bp Over 400 bp Over 12% Minimal (1) Minimal (1) Minimal (1) Moderate (2) 8% to 12% Minimal (1) Minimal (1) Moderate (2) Significant (3)

4% to 8% Minimal (1) Moderate (2) Significant (3)

High (4)

Below 4% Moderate (2) Significant (3)

High (4) High (4)

Figure 1-10 – OTS Table for Quantitative Sensitivity Rating


rate risk. On the other hand, if levels of credit risk and liquidity risk are high, the institution may feel uncomfortable with taking on this level of interest rate risk.

Setting Value at Risk Policy Limits Many institutions establish value at risk limits by placing constraints on level of sensitivity and post shock portfolio equity ratios. The objective of the limits is not to specify the direction of the institution’s interest rate bets. Management should have the prerogative of operating either an asset-sensitive or a liability- sensitive institution. Rather, the policy limits on value at risk should limit the size of the interest rate risk bet inherent in the institution’s balance sheet. Because OTS has the greatest amount of experience in administering value at risk measurement systems and has gone as far as to publish its standards for the Sensitivity rating in CAMELS, we often recommend that institutions use the table in Figure 1-10 for guidance in setting their value at risk limits – whether they are OTS regulated or not. Deciding on the limits from the above tables is a matter of resolving the following issues.

• How much interest rate risk do you feel comfortable taking on? If your stomach for interest rate risk is minimal, select policy limits that would keep you within the 1 range under OTS standards. If you are willing to be somewhat more aggressive, select policy limits that would keep you within the 2 range. We don’t recommend you set policy limits outside the 2 range.

• How much portfolio equity do you have and plan to maintain? The higher your minimum here, the more sensitivity you can allow with your sensitivity limit.

• How close to the limit of the category are you willing to run? Management of the institution in Figure 1-9, has decided to take on a moderate amount of interest rate risk. They need to select policy limits that will keep them within the 2 range. Given their post shock portfolio equity ratio of 9.51%, they feel comfortable with a post shock portfolio equity ratio limit of 9%. While they could go up to 400 basis points of sensitivity and still be a 2 rated shop, they elect to set a sensitivity limit of 300 basis points. Because they left a margin of error between their limits and the table minimums for a 2 rating, they might temporarily violate a limit and still stay within the 2 range. Their policy limit might be stated as follows:

• It is the policy of our institution to maintain a post 200 basis point shock portfolio equity ratio of no less than 9%. At the same time we will keep the change in the portfolio equity ratio from the 0 basis point to the worse case 200 basis point rate environment to less than 300 basis points.

Rather than focus on post shock ratios and sensitivity, other institutions set value at risk limits by specifying the maximum percentage change in the portfolio equity ratio from the 0 bp to worse case 200 basis point rate environment. This percentage change is referred to as Equity Exposure in Figure 1-9. For the institution illustrated in Figure 1-9, the Equity Exposure in the 200 BP rising rate environment is -21.57%. If you are required to set policy limits on Equity Exposure by your regulator, you may still wish to compute your post-shock Portfolio Equity Ratio and Sensitivity. If you do so, you can use the OTS table (Figure 1-10) for guidance. In reviewing their position in Figure 1-10, the institution in Figure 1-9 may feel they want to maintain a 2 rating under OTS guidelines. A little what-if work with a spreadsheet based on the mathematical relationships in Figure 1-10 might cause the institution to conclude that a 29% change in Portfolio Equity translates into 300 BP of sensitivity and a post shock Portfolio Equity Ratio of 8.78%. As a result they may settle on the following Equity Exposure Limit.

• It is the policy of our institution to maintain the Equity Exposure cause by a 200 basis point shock in both directions at 30% or less.

Evaluating Dynamic Value At Risk The value at risk analysis illustrated in Figure 1-9 is static in nature. It looks at value at risk in a recent balance sheet. But in evaluating alternative strategies for managing your shop, you might want to know where the


strategy will take you over time. “If I implement this strategy over the next year, does it have the potential for taking me outside my value at risk limits? With a competent asset-liability model, you should be able to get an answer to the question before the strategy is implemented in the real world. As Figure 1-11 shows, we model the strategy we are considering for implementation in our expected rate environment over our normal income at risk horizon (dotted green line to dotted green line). Once the balance sheet from the end of our modeling horizon has been generated (normally 2-3 years from today) we subject the forecast balance sheet to immediate and permanent rate shocks to evaluate the value at risk in the forecast balance sheet. Figure 1-11 – Dynamic Value at Risk

If we are considering two different strategies, we model both and apply the value at risk stress tests to the forecast balance sheet for both strategies. Once we have the results we can compare the impact of both strategies on potential value at risk. Figure 1-12 – Comparative Dynamic Value at Risk

Book -200 bp Flat +200 bpPortfolio Equity - Strat 1 $39,510,000 $53,738,000 $51,831,000 $47,103,000Portfolio Equity – Strat 2 $39,834,000 $57,305,000 $52,591,000 $43,223,000Port Equity Ratio - Strat 1 10.90% 14.21% 13.84% 12.75%Port Equity Ratio – Strat 2 10.59% 14.51% 13.54% 11.41%Rate Sensitivity – Strat 1 0 36.6 0 -109.2Rate Sensitivity – Strat 2 97.6 0 -212.8 Figure 1-12 shows the dynamic value at risk analysis results from modeling two alternative strategies over a three year horizon.

1. Strategy 1 – All 30 year fixed-rate mortgages were originated and sold in the secondary market during the three year forecast.

2. Strategy 2 – All 30 year mortgages were originated and re4tained in the institution’s loan portfolio during the three year forecast.

The balance sheet will look considerably different three years from now depending on whether 30 year mortgage originations are sold (Strategy 1) or kept (Strategy 2). A look at comparative book capital in Figure 1-12 shows that Strategy 2 resulted in $324,000 more capital being accumulated over 3 years. As this shop pays out 25% of net income in dividends, net income for Strategy 2 was approximately $432,000 higher with Strategy 2 than Strategy 1 over the three years. Because the institution was portfolioing well priced loans that had a market value greater than book at origination, Portfolio Equity was $720,000 higher with Strategy 2 than Strategy 1 in the flat rate environment. Both book


capital/assets and the portfolio equity ratio were lower with Strategy 2 than Strategy 1, because retaining the mortgages accelerated growth in assets. Strategy 2 did result in more interest rate risk than Strategy 1. It ends with nearly twice the sensitivity (212.8 basis points) and a post-shock Portfolio Equity Ratio over 1% lower at 11.41%. If the institution’s primary objective was to minimize interest rate risk, Strategy 1 would be chosen over Strategy 2. But under best practices of A/L Management our objective is to optimize the relationship between risk and return while living within the institution’s policy limits. A quick look at the institution’s sensitivity (212.8 BP) and post-shock ratio (11.41%) has the institution very comfortably within its value at risk limits of 300 basis points of sensitivity and a post-shock ratio of no less than 9%. Not only that, a quick glance at the OTS tables in Figure 1-10 indicates the institution just barely misses a 1 rating as Sensitivity is just over 200 bp. Given the comparative results in Figure 1-12 would you be willing to trade the higher return in Strategy 2 for a higher level of risk that is well within limits and lower than the risk in today’s balance sheet (Figure 1-9)? I’d want to look at an income at risk analysis for these alternative strategies in the format shown in Figure 1-5. But I suspect when all was said and done, I’d be recommending Strategy 2 to the board. Farin FORESIGHT produces value at risk reports on both current and forecast balance sheets. See chapter VII for inputs and assumptions necessary to perform market value calculations. Chapter VIII part 2 discusses market value reports. Developing the Base Forecast Once long-range goals and risk tolerance levels are in place, the next step involves preparing the base projections for the planning period being contemplated. The planning period could be 12 months (annual plan) or three to five years (strategic plan). Figure 1-13 lists the input data and assumptions that will go into

developing the base forecast.

Figure 1-13 - Inputs to Base Forecast Farin FORESIGHT gives you control over the inputs, internal assumptions, external assumptions and outputs coming out of a policy. See the following chapters for information on Base Forecast inputs:

• Chapter III – Defining the chart of account detail and account cash flow characteristics. • Chapter IV – Downloading your cash flow data into the model.

Balance Sheet Financial instrument cash flow characteristics Downloads of current financial instruments – balances, rates, maturities, repricing points,

caps, calls, etc. Financial Goals

Growth Capitalization Earnings

The Assumptions Loan and deposit pricing Originations, investments, borrowings Fees, operating expenses, non-earning assets Designed to meet institution’s financial goals.

Other Assumptions Prepayments, decay rates Discount rates Rate environments


• Chapter V – External assumptions including rate environments, prepayment speeds, and decay rates.

• Chapter VI – Policies including pricing assumptions, targets and sales. • Chapter VII – Assumptions needed to value balance sheet items. • Chapter VIII – Reporting.

While the institution has control over many of the assumptions in the base forecast, they have little control over rate environments and the resulting discount rates that will be used in valuing their balance sheet or the prepayment speeds they will experience in key categories of loans or investments. Because they have no control over rates, discount rates, and prepayment speeds, institutions typically model the performance of their plan through a variety of rate environments. Figure 1-14 is a graphic illustrating a plan being modeled through three rate environments. Figure 1-14

While the planning horizon is one year, the institution models the base plan over a four year extended modeling horizon (total 5 years) to get a sense for the long-term effect of changes in rates on income. The middle rate environment is the flat rate environment. The rising and falling rate environments between Today and 1 Year are designed to test the sensitivity of the base forecast’s income to changes in rates. Note that the rising and falling environments are rate ramps rather than immediate and permanent shocks. Moving rates gradually gives a strategy the opportunity to affect the

composition of the institution’s balance sheet before the rate environment has fully unwound. The extended horizon is also necessary as the full effect of the year one rate movement is not felt until the second year of the plan.

Figure 1-15: Static value at risk testing is accomplished on the Today’s balance sheet by applying immediate and permanent rate shocks to the balance sheet to see the effect of the rate shocks on economic value (value at risk). This VAR testing does not show the effect of the base strategy on economic value as the rate shocks are applied to the starting point of the plan.

Figure 1-16:

Dynamic value at risk testing involves applying immediate and permanent rate shocks to the balance sheet output by the model at the end of the base forecast. This does show the effect of the strategy on economic value as the balance sheet being tested incorporates the effect of the base strategy. Note that of the three rate environments run in the income at risk analysis, the balance sheet at the end of the base (middle rate environment) is the one to whom the value at risk testing is performed.


Farin FORESIGHT is capable of defining any rate environment you wish to run. For income at risk testing, rate ramps can be parallel or non-parallel. For market value testing, immediate and permanent rate shocks can be applied to both current and forecast balance sheets. Market value tests can also be parallel or non-parallel. Rate environments can be organized into rate sets for standardized testing. See Chapter V for details. Figure 1-17 lists the output from the planning model for the base strategy run in the base rate environment. Note that the output incorporates the effect of both the income at risk and forecast value at risk testing.

The balance sheet, income statement and key ratios are calculated from the base forecast in the base rate environment. Income at risk analysis is derived from running the base forecast in other rate environments (Figure 1-14). Ideally, value at risk analysis is performed on Today’s balance sheet (Figure 1-15) as well as on the balance sheet at the end of the base forecast (Figure 1-16). Comparison of the two value at risk analysis would give a sense for changes to value at risk that are likely to be caused by executing the base strategy over the planning horizon used in the

base forecast. Farin FORESIGHT offers a wide variety of reports and graphs. All can be customized to meet user’s needs. See Chapter VIII for details.

Liquidity risk analysis reporting would focus on sources and uses of funds during the planning horizon being analyzed in the base forecast. As liquidity can be rate environment dependent, a sources and uses of

funds analysis should be performed for each rate environment run (Figure 1-14).

Farin FORESIGHT supports both static and dynamic liquidity reports. Traditional liquidity ratios are calculated and displayed for both current and forecast periods. Sources and uses

reports track the dynamic tracking of liquidity through alternative strategies and rate environments. See Chapter VIII for details.

While the above examples imply the analysis of a single strategy, in selecting a final strategy, many alternatives would be modeled. The combination of strategies with the best risk-return tradeoffs would be implemented in preparing the base plan. Along with the output from the base strategy (Figure 1-17), the actual strategies would be documented in written form. Figure 1-18 is an example of such a document.

Investments Reduce investments as a percentage of earning assets while investing

new funds primarily in 2-3 year duration MBS to build collateral for borrowings. Given maturities and other cash flows this should keep investment portfolio duration less than 2 years.

Loans Continue to portfolio 15-year fixed-rate mortgages while selling conforming 30 year fixed-rate mortgages to continue to build loans to earning assets. Grow other loans where profitable based on internal

Figure 1-17 Base Projections

�Balance Sheet �Income Statement �Key Ratios

Risk Analysis Tools – Static & Dynamic �IRR - Income at Risk �IRR – Value at Risk �Liquidity Risk Analysis

Figure 1-18 – Asset-Liability Strategy Balance Sheet

Considerations Strategy


pricing analysis. Deposits Use pricing to control growth in rate sensitive deposits. Don’t raise

deposits when marginal cost is above wholesale borrowings. Emphasis should be placed on building non-maturity deposits because of longer duration.

Borrowings Use long-term FHLB advances to supplement deposits and extend funding duration to hedge interest rate risk in fixed-rate mortgages being added to the loan portfolio.

Derivative Securities Investigate the use of out of the money Interest Rate Caps as a tool to hedge the option risk in our balance sheet. Begin developing policy guidelines, accounting practices, and management and board education to allow us to use this tool in the future.

Other Considerations

Capital Management Manage the relationship between our slow current asset growth due to a weak economy and our capital growth by repurchasing our stock when it is available at a fair market price. If not available consider a special dividend to stockholders to slow capital growth. The objective is to maintain our current level of financial leverage, a key factor in meeting ROE goals.

Liquidity Offset the reduction in asset liquidity due to decline in investments / earning assets by focusing on adding loans (mortgages) and investments (MBS) that are potential collateral for future liability based liquidity (borrowings).

Credit Risk During current weak economy, focus on A and B credit consumer loans and mortgages that conform to secondary market credit standards. Continue to employ risk-based pricing to insure that where we take on credit risk we are compensated with higher yields.

Interest Rate Risk Income at risk and value at risk analysis both show our institution to be liability sensitive. Interest rate risk levels are minimal to moderate as defined by regulatory standards and are well within our interest rate risk guidelines. Testing of this strategy shows that interest rate risk will rise slowly as this strategy unfolds. Even so, change in income and value is moderate for moderate changes in rates. But the option risk in our balance sheet causes more severe changes in income and value for larger rate shocks. For that reason we are investigating the use of out of the money interest rate caps to hedge some of this option risk for more severe rate shocks.

The combination of the base projections and the above ALCO strategy represents the initial plan for the

institution’s fiscal year. The plan is put in place with management knowing full well that external factors such as movement in rates is likely to affect the accuracy of our base projections. As a result, changes to

both base projections and strategies will need to occur as the planning year unfolds. Dealing With Change During the Planning Period Practitioners implementing best practices in asset-liability management will reset their base forecast when necessary based on changes in both economic and competitive environments. They do so each time real world changes and internal results have evolved significantly enough to cause assumptions and strategies in the base forecast to no longer be accurate or relevant. For example, take the institution that has adopted the set of asset-liability strategies outlined in Figure 1-18. Assume the strategy was developed assuming a base rate forecast of flat rates as shown in Figure 1-14.


Assume in the next three months, rather than rates remaining flat, they have dropped following the path of the declining rate environment as shown in Figure 1-19.

Figure 1-19 - Three months into the year actual rates are tracking the falling rate environment (blue arrow). As a result

prepayments are accelerating on mortgages, causing both ARM and fixed-rate mortgages to refinance. In addition those

refinancing are rolling consumer debt (home equity, auto loans, and credit cards) into the refinance. There are not sufficient

originations in 15 year fixed-rate mortgages to keep up with the declines in existing loans. As a result, loans to earning assets are declining with cash building up in an extremely low yielding fed

funds account. Net interest margin, ROA and ROE are all falling. Should the institution continue to execute its current asset-liability strategy, it will fall far short of meeting

goals for net interest margin, ROA, ROE, and loan growth. Under a traditional planning environment, much of the work performed on management reporting would focus on variances and there would be plenty of variances to report. But the mechanism wouldn’t be in place to reset the plan in response to the changes

happening to rates.

Under best practices in A/L management the ALCO committee and management team would use the tools at its disposal to re-evaluate its position and reset its strategy. Tools would include:

Variance Reporting

Variance reporting’s primary purpose in a best practices A/L system is to separate variances that are outside an institution’s control (rate environments, prepayments, etc.) from those within its control. Those within its controls should be further broken down into assumption problems (pricing / volume assumptions for example) and those representing implementation problems (expense controls, fee collection, investment

decisions, etc.) Traditionally variance reporting has focused on balance sheet variance, income statement variance and variance in key ratios derived from balance sheet and income statement environments. Further useful information in isolating causes for changes in performance would include the following additional kinds of variance reporting:

o Variance reports comparing actual rate movements in comparison to the base rate forecast in the plan.

o Variance reports comparing actual origination volumes by product to those assumed in the plan. o Variance reports comparing actual prepayment speeds by product those assumed in the plan. o Variance reports showing actual pricing to that assumed in the plan. o Variance reports showing variances in income at risk from that projected in the plan o Variance reports showing variances in value at risk from that projected in the plan. o Variance reports showing variances in liquidity metrics from those assumed in the plan. o Variance reports showing variances in credit risk metrics from those in the plan.

Examination of variance reports would give management a sense for problem areas in key assumptions and how the institution is performing relative to the plan. Variances would be reported relative to financial goals and key risk metrics. Analysis of this information would identify opportunities for corrective action.

Formulating Strategies and Evaluating Risk/Return Tradeoffs Based on evaluation of variance reports the ALCO committee would develop a number of strategies that would need to be considered in determining what kinds of modifications might be made to the institution’s strategy.


In the case of the previously discussed institution experiencing accelerated prepayments on mortgages in the falling rate environments, the following strategy alternatives might be considered as possibilities.

o Begin investing the growing investment portfolio into 1) similar duration assets to those already indicated in the current ALCO strategy, 2) or 15 year fixed-rate mortgage-backed securities.

o Begin portfolioing all 30 year fixed-rate mortgages in addition to the 15 year fixed-rates already portfolioed. Drop rates on short-term CDs and replace runoff with long-term FHLB Advances that are duration matched to the mortgages. Shorten the investment portfolio purchases to offset the increasing duration of loans.

o Sell all 15-year fixed-rate mortgage production as well as all 30 year FRMs, reasoning that nobody should hold fixed-rate mortgages with rates this low. Accept the declines in loan growth, margin, ROA and ROE as inevitable given the rate and economic environment.

Each of the above three strategies would be modeled under a variety of rate environments as shown in Figure 1-20.

Figure 1-20 – Given the changes that have already happened to rates (blue arrow) a new set of rate environments is developed to test income at risk for the proposed strategy modifications. The middle rate environment would become the new base rate environment. Once all three strategy revisions are tested, the strategy with the best risk/return tradeoffs would be selected. Then that strategy would be modified for the other rate environments to consider the major issues that would be addressed in those rate environments.

The ALCO committee would then approve the revised strategy. Assuming that the 30 year mortgage retention strategy is assumed to have the best risk/return tradeoffs, the revised strategy might appear as in Figure 1-21. Major changes from the original strategy in Figure 1-18 are highlighted in bold blue.

Investments Reduce investments as a percentage of earning assets while investing new funds primarily in 1-2 year duration MBS to build collateral for borrowings. Given maturities and other cash flows this should keep investment portfolio duration less than 1.5 years.

Loans Portfolio all 15-year and 30-year fixed-rate mortgages to continue to build loans to earning assets. Grow other loans where profitable based on internal pricing analysis.

Deposits Use pricing to reduce rate sensitive short-term CDs. Don’t raise deposits when marginal cost is above wholesale borrowings. Emphasis should be placed on building non-maturity deposits because of longer duration.

Borrowings Increase use of long-term FHLB advances to supplement deposits and extend funding duration to hedge interest rate risk in fixed-rate mortgages being added to the loan portfolio.

Derivative Securities Investigate the use of out of the money Interest Rate Caps as a tool to hedge the option risk in our balance sheet. Begin developing policy guidelines, accounting practices, and management and board education to allow us to use this tool in the future.

Figure 1-21 – Revised Asset-Liability Strategy Balance Sheet

Considerations Strategy


Other Considerations

Capital Management Manage the relationship between our slow current asset growth due to a weak economy and our capital growth by repurchasing our stock when it is available at a fair market price. If not available consider a special dividend to stockholders to slow capital growth. The objective is to maintain our current level of financial leverage, a key factor in meeting ROE goals.

Liquidity Offset the reduction in asset liquidity due to decline in investments / earning assets by focusing on adding loans (mortgages) and investments (MBS) that are potential collateral for future liability based liquidity (borrowings).

Credit Risk During current weak economy, focus on A and B credit consumer loans and mortgages that conform to secondary market credit standards. Continue to employ risk-based pricing to insure that where we take on credit risk we are compensated with higher yields.

Interest Rate Risk Income at risk and value at risk analysis both show our institution to be liability sensitive. Interest rate risk levels are minimal to moderate as defined by regulatory standards and are well within our interest rate risk guidelines. Testing of this strategy shows that interest rate risk will rise slowly as this strategy unfolds. Even so, change in income and value is moderate for moderate changes in rates. But the option risk in our balance sheet causes more severe changes in income and value for larger rate shocks. For that reason we are investigating the use of out of the money interest rate caps to hedge some of this option risk for more severe rate shocks. Given the decision to portfolio 30-year fixed-rate mortgages, perform IAR and VAR testing monthly. 30 Year fixed-rate mortgages have even more option risk than 15’s.

Depending on how quickly rates are moving, the strategy revision cycle could happen as often as monthly. During times of stagnant economic conditions, strategy revisions would happen less quickly, especially if it turns out that variance analysis doesn’t uncover major problems in the institution’s execution and/or internal assumptions. Many managers familiar with the traditional budgeting/planning process may express concern over this concept of changing the plan on the fly as the year evolves. Keep in mind, that in the above examples, the institution’s goals were not changed. What did change is the path the institution needs to follow to reach the goals. Farin FORESIGHT offers the ability to clone an existing policy or plan so what-if testing can be performed on potential strategy revisions. See Chapter VI for details.

Conclusion This chapter has provide a basic structure for running a best practices ALCO process that relies on the sophistication provided by a quality asset-liability model. The process begins with establishing long-range financial goals over a strategic time frame. Goals focus on earnings, growth, dividends, and capital. Measurement systems are selected to measure interest rate risk. The institution is best served selecting measurement systems that are dynamic and focus on measuring risk using the same metrics as those used to


measure return. With a common set of metrics and a dynamic environment, potential strategies can be evaluated based on risk/return tradeoffs. Institutions of all kinds will want to choose a measurement system for measuring interest rate risk that focuses on income at risk. Institutions with long-term balance sheets will need to supplement their income at risk measures with value at risk measures. An effective process will lead to meaningful income and value at risk policy limits. With a quality A/L model, goals, and limits, an institution can model risk/return tradeoffs in plans, choosing strategies that optimize return while keeping risk within guidelines. An effective ALCO process will allow an institution to react to economic and competitive change within a planning period by evaluating risk return tradeoffs in alternative strategies, choosing those that will bring the institution closer to achieving its long-range financial goals.

Farin Foresight User Manual 5.2.3.5 Page 1 of 9 © Copyright 2002-2008 by Farin & Associates, Inc.

Chapter 2

FORESIGHT Startup In this chapter we will outline a basic overview of database management and give instructions for opening the software, restoring a current database, logging into the institution and verifying the preset institution information. 2.a Logging Into FORESIGHT The first screen the user sees when they open FORESIGHT is the Login screen (shown to the right). From the Login screen the user will be able to navigate to Institutions, All Tasks and Manage Database Connections. The Login screen contains the settings which allow the user to control the total or partial access of others to the model.

1. Username/ Password Not currently being used. This option will be employed in future versions to control model access. This will allow some users complete access while giving others “read only” access.

2. Start Allows the user to easily enter a Default Database (bypasses the Manage Database Connections). 2. Default Database to Load Shows the user the name and location of the server and database that is selected as the Default Database. A default database can be set on the Manage Database screen (see 2.b.vi) or by using the selection option here. Manage Database Connections Takes you to Manage Databases. Enter Authorization Code Entered during installation. Needed to receive updates, market rates, prepay and decay rates. Configuration Used to display messages about the machine configuration that may prevent you from getting expected behavior with certain Foresight features. Version Shows the version of the model the user is currently running. The version number will change each time the model is automatically updated. Version Info Takes you to a web site with version changes listed.


2.b. Manage Database Screen From the Login screen, select Manage Database (right side of the screen). The Manage Databases screen opens.

2.b.i. Selecting a Database The Manage Databases screen allows the user to select the database with which they would like to work. Most users will only manage one database. If the database to be opened does not appear in the Servers/Databases list on the left, the user can restore that database from a previously made Backup. Directions to backup and restore a database are provided in the next section. After restoring, the institution’s name will appear in the Server/Database list in the box on the left side of the screen. If the user knows the name of a database and the name of the server on which it resides, they can connect to that institution using the ADD Database Connection option. This option is outlined in Chapter 2.c.vi. To select the institution that one would like to work with, highlight the name on the left hand side and click the Login button in the bottom center. FORESIGHT will open the database and the user will see the Institution screen (outlined below). The user has now ‘opened’ their Institution.

2.b.ii. Creating a Backup The user will need to protect their data by creating Backups often. The user can Restore the Backup at any time - this returns the database back to an exact copy of when the Backup was created (accounts, balances, etc.). Therefore, if the user plans on making substantial changes to their model, they should make a Backup first. The user may also need to email the database or save it to a CD for travel or troubleshooting purposes. The Backup function saves the database in a compressed file. Remember, there is no UNDO and restoring the wrong database will delete all of a user’s work; therefore, it is very important to know the contents of a Backup BEFORE it is restored.

The backup function is used to make a file copy of the selected database. Databases do not exist as separate files - the data is stored in proprietary formats and managed by SQL Server or MSDE. When you make a backup, you actually get a file copy which can be archived, moved, emailed to others and managed as a separate file.

To make a Backup, from the Manage Databases screen, choose the database to be backed up by clicking on the name in the Servers/ Databases window and Click on the Backup/Archive button. The Select Backup File Name window will open.


Browse to the location1 that you would like to store the Backup - the name of the Backup is preset by the model and is based on the institution name and the current date. After the process is complete, the user is returned to the Manage Databases screen. The user now has a ‘Known Good’ model they can return to if needed.

2.b.iii. Restore Function If a user’s computer crashes or they make changes which they do not like, they can get back to a ‘Known Good’ model by restoring a previously-made Backup. Also, users who travel with their model may have to restore a Backup of their work once they return from traveling. To restore a database, go to the Manage Databases screen and click on the database to be restored OVER (therefore, anything currently inside of the DB will be written over). This means that the user must be careful not to accidentally choose an ‘older’ Backup and restore it over the current database rather than a newer Backup. To alleviate this possibility, the user should verify the Backup by checking the Date stamp in the name.

Once the database destination has been selected, click on the Restore Database from Backup link on the right side of the Manage Databases screen. The Select Backup File Name window will open. Browse to the Backup that you would like to restore. Select Done. The model will take the information contained in the Backup file and restore that information OVER the selected database. Be CAREFUL. Users may overwrite a useful database with garbage from an incorrect/older Backup. Make sure you know what is contained in a Backup before you restore it OVER any existing database. This will most often occur when the user accidentally restores an older version of a Backup rather than the desired newer version. The contents of a Backup can be discerned from the name, which is derived from the name of the institution and the date the Backup was made. If one is unsure as to whether they are restoring an older Backup over a database, they should make a Backup of the current database before performing the restore function.

2.b.iv. Copy Database The Copy Database function allows the user to copy one existing database over another. This process will overwrite the contents of the Destination database (the one being copied over) and replace it with the data from the source database. This process is most often performed when a user copies a database from a network server to a local server (or vice versa). Be careful not to have more than one MASTER database2. As outlined earlier, recombining databases is impossible. See Chapter 2.a.ii for more information on databases and database management. The user may also create a blank database and copy an existing database into that blank database.

1 Though the user can save Backups anywhere (i.e., there is no physical restrictions) we suggest storing the Backup in a network folder. If the users machine crashes and backup’s are only on the C drive, everything is lost! The best place to save the backup is a network folder. 2 That means, don’t copy the database from the network to a laptop and still allow others to make changes to the database on the network. The moment the model is copied from the network to a local drive, the database on the LOCAL drive becomes the MASTER and the database on the network should not be changed.


Be careful of creating extraneous databases and be careful when writing over existing databases. To copy a database, highlight the source database shown in the list on the left hand side of the Copy Database screen (shown above). Then highlight the Destination on the right. Click Copy. Remember, the SOURCE information will overwrite the DESTINATION information. The user must be certain they are not overwriting important information in the DESTINATION database. Again, there is no UNDO. Once the data has been written over, it is gone forever. The user should make a Backup of BOTH databases prior to Copy.

It cannot be stated enough - the user should manage only one MASTER database, as multiple databases can NEVER be recombined.

2.b.v. Delete Database The delete process allows the user to delete an entire database. Deleting an entire database may, in most cases, be an extreme tragedy. Rebuilding a deleted database from scratch is a time-consuming process. Please contact a Farin & Associates employee prior to deleting a database. Most times when using DELETE, the user really wants to simply REMOVE a connection (bottom left of the Login Screen) not DELETE.

2.b.vi. Other Manage Database Options Numerous other database management options appear on the right side of the Manage Databases screen. These options allow the user to add and remove databases and database connections, as well as servers and server connections. Deleting and removing connections, databases, or servers is very dangerous. Please contact a Farin & Associates employee prior to deleting or removing an item.

Edit Database Connection Use this to create and edit the connection string for the selected database. This option will be employed to re-map databases (or servers) after an institution’s IT department moves the data or changes other critical settings. Set Default Database Make the selected database the default database when you login. By clicking on this option the user can set a specific database as the default database. This will allow the user to open that database from the Login Screen.

Create New Database Create a new database on an existing server. The new database will contain a blank institution. This option allows the user to create a new database on their server. The database created will be a blank, ‘template’ database. The user can populate this new database with the Copy Database function outlined above. Having numerous Dummy, or Test Databases, is a dangerous thing to do.

Add Database Connection This option allows the user to add a server connection to an existing database. This is often done when installing FORESIGHT on a new computer. The database already exists on the server. The user just needs to make a connection to that database.


Remove Database Connection This option removes the server connection to an existing database. This will not delete the database. It only removes the connection. Remove Server Connection This option removes the server connection from an existing server. It also severs all connections with databases on that server.

2.c. Institutions Tab Once the user has selected an institution and logged-in (done on the Login tab) they can select the Institutions tab (shown below). From this tab, the user can verify the database, the current model’s “as of” date, and select the branch and institution within a holding company that they would like to work with. There are numerous institution settings residing on the institutions tab. When the user selects the ‘Update Settings’ button, the Update Institution screen appears (shown below). The data on this screen shot seldom changes. Most institutions will be able to maintain the initial settings created by Farin & Associates. Characteristics Name: User entered alphanumeric string that will be displayed on reports. It is not required but it is best to avoid illegal DOS characters (?/\”.*:’) in the name. Type: Bank / US Credit Union / Savings and Loan / Holding Company / Non Bank Subsidiary. The type chosen has an effect on regulatory reports that FORESIGHT generates. Fiscal Year End: The calendar month the institutions fiscal year ends. This has an effect on various reporting options and supplies the necessary information for totaling of Income on a Fiscal Year to Date as opposed to totaling of Income on a Calendar Year to Date. Currency: US Dollar / Canadian Dollar. Only has an effect if there are multiple institutions in the same database with different Currency Types. This would likely occur in a multi-national holding company. A currency conversion table needs to be filled out if multiple currency types are selected. Subchapter S Corporation: True or False.


Model Settings Default Growth Rate: This is the growth rate for the default target settings (institution wide). Prepay Method: Prepayment Interpolation Method. By Step /Nearest/ Linear/Geometric. The Prepay/Decay method allows the user to define how they would like the model to transition from table to table and yield curve to yield curve during a rate shock. See chapter 5 for explanation of calculations. As Of Date: This is the date that will be used for a default for the ‘Current Date of the Model’. It doesn’t have to reflect the last period for which historical data is stored in the model, but usually does. History Starts: The date of the first period of history that is stored in the model. Period Length: The period size of the historical data stored in the model. This is not changeable once the database has been created. Tax Brackets The user can define their tax rate at the Federal, State and Local level. The user can insert additional tax lines with the INS button. The user can remove extraneous options using the DEL button. Income taxes are calculated by the Tax Calculation routine. This is called from the Balancing routine. The Tax Calculation routine first calculates taxable income and then expense. It then applies the Tax Brackets (found on the Update Institution Settings screen) against the calculated Net Income before Taxes. If a forecast period has a tax loss it creates a tax loss carryforward. This will reduce the amount of taxes calculated in subsequent periods. When balancing, Credit Unions don't call the Tax Calculation routine. Taxable Income is calculated separately at the Local, State and Federal levels. Each level of Taxable Income is calculated by multiplying the accounts income by the Taxable Income % (found on the Manage Chart of Accounts screen). The individual accounts taxable income is summed and then expense is calculated. NIBT (Net Income before Taxes) is then calculated. The correct Tax Brackets are used to calculate taxes (matched to the appropriate NIBT). While the Local, State and Federal taxes can be stored in different accounts, they should be arranged under a single category. This makes it easier to reference the total amount of taxes in the reports dictionary. Special Accounts – for all institution types The user will learn in Chapter 3 the extent to which the Chart of Accounts is user-defined. In fact, the Chart is nearly 100% user-defined. This level of flexibility adds complexity to the model when trying to maintain accounting conventions; therefore, the user must ‘force’ the model to treat certain accounts as ‘special’ accounts. The user must determine which accounts will be the Asset and Liability Balancing accounts, as these accounts help the model maintain the relationship (Assets = Liabilities +Equity) throughout the forecast and budget. Asset Balancing Account – In order to balance the Balance Sheet, any surplus liabilities are lent out in this account. Typically, this is done in some fashion of a Federal Funds Sold or Overnight account. There does not have to be a related account on the Income Statement. In the case where there is no related account on the Income Statement then the cost of these funds will be zero. Liability Balancing Account – In order to balance the Balance Sheet, any amount equal to the unfunded assets are borrowed in this account. Typically, this is done in some fashion of a Fed Funds Purchased account.


There does not have to be a related account on the Income Statement. In the case where there is no related account on the Income Statement the cost of these funds will be zero. Retained Earnings or Undivided Earnings Account – This is the account that is used to handle the flow from the footing of the income statement to the equity portion of the balance sheet in order to balance the balance sheet. Gain/Loss on Sale – The institutions appropriate investment gain/ loss income statement account. This is for future use, when sales capabilities are fully supported in Foresight. FASB 115 Gain/Loss – FORESIGHT will calculate the difference between Amortized Cost and Market Value in all accounts with a FASB Type of Available for Sale (AFS). This amount less the FASB 115 Tax will be inserted into this equity account during forecast calculation runs. FASB 115 Tax – Using the tax accounts, an amount equal to the taxable portion of FASB 115 Gain will be inserted into this Liability account. Trading Account Gain/Loss – FORESIGHT will calculate the difference between Amortized Cost and Market Value in all accounts with a FASB Type of Trading. This amount will be inserted into this account. This is similar to the FASB 115 Gain special account. Dividends Account – If you choose to use the model to calculate stock dividends and transfer the net income after dividends to the equity than this account will need to be assigned. The dividends account assigned is the account on the Income Statement after the Net Income. Special Accounts – for taxable institution types Federal Tax Account – This is the income statement account that will receive calculated federal taxes from the tax tables. If there were no applicable federal taxes this special account would not have to be setup. An institution with a tax status of Subchapter S would be an example of an institution that did not pay any federal taxes. This should be an expense account. State Tax Account – This is the income statement account that will receive calculated state taxes from the tax tables. This special account doesn’t have to be filled out if there are no applicable state taxes. This should be an expense account Local Tax Account – This is the income statement account that will receive calculated local taxes from the tax tables. This special account doesn’t have to be filled out if there are no applicable local taxes. This should be an expense account. Special Accounts – specific to credit union institution types Reserve Transfer Account – This is the account that will receive the calculated, periodic reserve transfer amount. See below for information on reserve transfer. Total Loans Account – The reserve transfer calculations need to know where to find a total loans category. This should specify the category or footing category that best represents total loans in the chart of accounts. Loans Sold with Recourse Account – If any such accounts exist, this should point to such an account or category (usually a memo account). It is also used for the reserve transfer and PCA calculations. Unused Member Business Loan Commitments Account - If any such accounts exist, this should point to such an account or category (usually a memo account). It is also used for the reserve transfer and PCA calculations.


As the Chart of Accounts is modified, the user may choose to adjust the Special Accounts. To change the Special Account settings, click on the X next to the Special Account to be changed. A copy of the Chart of Accounts will appear in a window. Scroll through the chart and select the correct item in the chart. Select Done. Setting up Reserve Transfer for Credit Unions Each asset account on the balance sheet must be designated for the type of risk portfolio it falls into. There are 8 risk portfolio categories; 1) Long-Term Real Estate Loans, 2) Member Business Loans, 3) Investments, 4) Low Risk Assets, 5) Average Risk Assets, 6) Loans Sold with Recourse, 7) Unused Member Business Loan Commitments, and 8) Allowance. Numbers six and seven are risk portfolios that are not on the balance sheet. In addition the Equity Accounts can be designated. The best way to designate risk categories is using the Manage Chart - Characteristics screen. Go to the category which represents total assets and set the Risk Category to "Average Risk Assets". Then with the cursor on the drop down symbol, right click the mouse and select "Push Down". This will push the setting from the category level down to all accounts below it. Then move to, for instance the category that represents Total Investments and set that category’s Risk Category to "Investments" and push it down in the same way. In this fashion you can quickly set the Risk Categories of all the accounts on the balance sheet. (You can directly set any account’s Risk Category without having to push it down.) Then set the Equity Accounts that are needed to determine the regulatory Net Worth Ratio by selecting the Risk Category Equity Accounts. You can use the push down method as stated above. After setting the accounts on the balance sheet you next have to go over to the Update Institution Settings screen. On this screen you need to set the following special accounts: 1) Loans Sold With Recourse, and 2) Unused Member Loan Commitments. This completes filling out the risk portfolios. Remember that if your credit union has broken out either Loans Sold With Recourse or Unused Member Loan Commitments you can create a footing category that sums up all the disparate pieces and designate that footing category as the special account. While you are on the Update Institution Setting screen you also need to designate the "Total Loans" account so the financial engine knows where it is when it needs it to make ratio calculations for the regular reserve transfer. Finally you need to designate the account that will receive the regular reserve transfer if it takes place.


2.d. The All Tasks Menu With the software installed, the database restored, and the institution information verified, the user is now ready to start using the model. The All Tasks tab is the center of what a user will be doing in FORESIGHT. Any operation that needs to be completed can be accessed from this tab. After the user has logged in and selected their institution or branch, they can move around the model starting from the All Tasks tab. Descriptions under the links are mostly self-explanatory, and we will refer to them as we outline the specific tasks to be carried out in FORESIGHT.

Farin Foresight User Manual 5.2.3.5 Page Page 1 of 23 © Copyright 2002-2008 by Farin & Associates, Inc.

Chapter 3

FORESIGHT Current Balance Sheet and Income Statement FORESIGHT allows institutions nearly 100% user flexibility in designing a Chart of Accounts. Users are free to insert as many accounts as they need to reproduce their institution’s unique cash flows. Users are also free to insert unlimited categories to help them group accounts making reconciliation quicker and the display more efficient. Finally, the user can insert footer categories, which provide additional accounting and memo information on reports. Because of this flexibility, the user also has to be aware of the possible dangers of inserting chart items incorrectly. Users will need to make strategic decisions about which items to add to the chart and which items to leave out1. In this chapter, we will outline a process to choose the level of detail to be displayed in the Chart of Accounts. We will then outline the process of changing the size, shape and scope of the Chart of Accounts using the Insert, Delete, Move and Rename functions. Next, we will focus on setting the characteristics of accounts correctly: matures, amortizes, repricing, caps/floors, prepay/decay tables etc. We end the chapter by looking at the historical data and the cash flow data displayed on the Manage Chart of Account screen. As a final reminder, before we get started, the user will initially spend the majority of their time working within the Manage Chart of Accounts screen. Once the model is set up and the user is confident that it will properly reproduce the institution’s actual cash flows, much of the user’s time will be spent in the Forecast Assumption section, Download section and Reports. At that point, the user spends very little time from period to period adjusting the Chart of Accounts. 3.a. SAM100 - The Importance of Understanding Cash Flows As discussed in Chapter 1, every decision that an Asset Liability (AL) Manager makes emanates from their expectation of future cash flows. By anticipating the cash flows that will roll off or reprice in different scenarios, the AL Manager can construct and study groups of rate-sensitive assets and rate-sensitive liabilities. By interpreting changes in these groups, the AL Manager can measure liquidity, interest rate and solvency risks – to name a few. Therefore, properly projected cash flows can provide a solid picture of an institution’s future risk exposure. Poorly projected cash flows can lead to a general misunderstanding of risk, at best, and a total miss-management of the bank, at worst. To help the user identify the level of detail required in their model, this section will explain the basic characteristics of cash flows. It is important to start by reviewing the various types of financial instruments found in banking. While looking at Investments, Loans, NMD (non-maturity deposits), CDs, and Other Borrowings, we will focus on cash flow characteristics (CFC’s) such as amortization, balloons, repricing, callables, prepayments speeds, and decay rates. This instrument review will provide a common starting point for understanding the cash flow characteristics associated with the different types of financial instruments. Without this understanding, one cannot model instruments correctly in the FORESIGHT model. It is paramount for every modeler to intuitively understand the theoretical cash flow characteristics of the financial instruments used by banks. If the modeler cannot mentally visualize the cash flow differences between a MBS (mortgage backed security) Pass-Through and a CMO (collateralized mortgage obligation), they will be unable to model such accounts in FORESIGHT (the CMO requires much different setup information than the MBS). Without a clear idea of how balloon loans differ from bullet loans, the user will fail to locate cash flow errors in their raw data downloads by looking at a GAP report. Furthermore, without

1 Certainly a user could insert hundreds of Accounts in FORESIGHT, but the small level of improved detail gained by inserting so many accounts would be outweighed by the massive reduction in processing speed and the additional time spent managing/ administering the system.


understanding the effects of the embedded options in Agency Callables (assets) and FHLB Callables (liabilities), the user will fail to anticipate, or explain, changes in market value in different rate environments. The examples provided above describe just a few of the many types of cash flows the user must understand; therefore, the first step towards modeling financial instruments is to truly understand the instruments. These generic characteristics are the basic building blocks for understanding all financial instruments. Moreover, the characteristics are universal across the balance sheet. An amortizing MBS investment, an amortizing loan, and an amortizing FHLB advance are basically the same thing. A bullet bond, a bullet loan, a CD, and a bullet FHLB advance exhibit similar cash flows. Individually, the generic characteristics are seldom problematic; however, as we combine the pieces into real-world instruments, the concepts become much more difficult to grasp.

Immediate Instruments An immediate instrument is an instrument that can be retired, or which ‘matures,’ immediately. This instrument has no ‘Term to Maturity.’ Since the instrument has no term, it also has no prepayment or repricing characteristics. Examples of immediate accounts are Fed Funds (overnight investing/ borrowing), Credit Cards and NMDs (Non-Maturity Deposits). When modeling, the user needs to be concerned about balance, coupon rate and decay rate. These are the characteristics that create cash flows. Amortizing Instruments An amortizing instrument is one in which the issuer receives principle and interest payments through out the life of the instrument. The amount paid to reduce principle each month is the ‘amortizing amount.’ The typical loan is an amortizing instrument. The standard amortizing instrument has the same original maturity and amortizing period, but there are amortizing ‘hybrid’ instruments that have different amortizing and maturity periods, for example the balloon instrument. When modeling amortizing instruments, the user needs to be concerned with coupon rate, maturity term, amortizing period, repricing, and prepayments. These are the characteristics that create cash flows. Balloon Instruments A balloon instrument is a subset of the amortizing instruments. A balloon is an instrument with a very short absolute maturity (for example five years), but it will amortize over a much longer period. The balloon allows a lender to lend for a shorter period of time (reducing various risk exposures to the lender) while making it affordable for borrowers to borrow (reducing the payments to low levels). For example, many borrowers could not afford the amortization on a short-term instrument, so the lender allows the borrower to pay back the instrument (amortize) on a schedule as if it were a longer-term instrument, say 25 or 30 years. Since the amortization is much lower on a 30-year instrument than it is on a five-year instrument, the borrower will still owe a substantial portion of the principle at maturity. This final payment is the ‘balloon payment.’ The user needs to be concerned about coupon rate, maturity term, amortizing period, repricing and prepayments. These are the characteristics that create cash flows. Bullet Instrument A bullet instrument is one in which the entire principle amount is paid in one lump sum at maturity, for example, a bond or a certificate of deposit. The borrower will pay interest through the life of the instrument and then re-pay the entire principal amount upon maturity. A variation of the bullet is a discount instrument. A discount instrument has only one lump sum payment at maturity which contains both interest and principle. An example of a discount instrument is a US Treasury Strip. The user needs to be concerned about coupon rate, maturity term, interest payment frequency, repricing and prepayments. These are the characteristics that create cash flows.

Users will find additional embedded options in the generic cash flows listed above. Often, the user will be confronted with repricing, callables, putables, prepayments and CMO options.


Repricing Variable rate instruments can be found in all areas of the balance sheet. Nearly all repricing instruments reprice through some contractual2 agreement between the bank and the borrower (loans), the bank and the issuer (investments), or the bank and the lender (other borrowings). To model variable rate cash flows, the user must possess complete detail about the repricing contract for each instrument. Specifically, the user must understand the repricing schedule of every instrument. Contractually, how many months from the origination date is the first reprice date? How often does the instrument reprice after that? With this information, the AL manager can combine instruments with similar repricing characteristics and separate dissimilar instruments. Callables The seller of a callable has the right to repurchase the instrument at some point in the future. The buyer has the obligation to relinquish the instrument. For example, a municipal that issues a callable bond has the right to recall that bond in the future, while the buyer (the bank) has the obligation to sell the bond back. This will happen in a falling rate environment. In another example a FHLB that issues callable advances has the right to recall the advance in the future while the borrower (the bank) has the obligation to repay the money on demand. This happens in a rising rate environment. Given these characteristics, one can see that the purchaser of the callable bears the risk. Callable instruments can either call once during their life, or they can call at some set interval throughout their life. To understand a callable, the user must understand the entire call schedule: first call and subsequent call frequency. Putables The seller of a putable instrument has the obligation to repurchase the instrument in some point at the future. The buyer has the right to relinquish the instrument to the issuer at a set price in the future. Given these characteristics, one can see that the seller of the putable bears the risk. Banks typically have very few wholesale putable investments. Most wholesale putable instruments are found in the liability section, specifically FHLB Callable Advances. Prepayments Prepayments occur on almost all types of loans. Prepayments affect projected cash flows by speeding up the principle payments and reducing the average life of the instrument. Prepayments occur because of re-financing and early retirement of the loan related to job changes, deaths, etc. Prepayments can have a massive effect on cash flows in changing rate environments. CMOs A true CMO is a unique subset of amortizing MBS, or ABS (asset backed securities). A basic MBS pass-through amortizes (pays down principle) throughout the life of the instrument exactly like an amortizing mortgage loan. The investor simply buys a pro rata share of all of the cash flows in the portfolio for the entire life of the portfolio. For example, the investor will get a percentage of the interest and principle cash flows each month for the entire life of the 30-year mortgages that comprise the underlying collateral of the MBS pass-through. A CMO is unique in that the buyer purchases a percentage of a specific set of cash flows rather than a percentage of all the cash flows throughout the life. For example, the purchaser of a certain CMO may buy an investment (tranche) that entitles them to principle payments for months 120 through 240 only. They will receive interest payments while other investors are receiving the principle from months 1 through 119. During the principle payments from months 120 through 240, the investor will receive a pro rata share of principle and their interest payment. After the principle payment from month 240 has been paid, the investment is viewed as matured and the investor receives no further cash flows even though the investors holding principle

2 There are some instruments that reprice by a calendar schedule regardless of origination or maturity. Pass-through MBS’s often reprice by calendar month for convenience.


cash flows from months 241 through 360 are still receiving money. CMOs are highly volatile instruments and their exact cash flows are never known in advance. Cash flows need to be estimated for numerous future rate environments to truly understand the risks related to owning various CMOs.

With the basic financial building blocks in hand, one can now begin to modify the Chart of Accounts to match their unique risks. The user should create an account for each unique cash flow type identified above. To aid in this process, the user can find specific Cash Flow examples in Appendix C. 3.b. Choosing Your Level of Detail - Managing Your Current Layout As outlined above, banking is risk management and managers need to identify and track their unique risks. In this section, the user will learn how to modify Foresight to better match and model their risks. The steps included are Inserting items, Deleting items, Moving and Renaming items. Once the Chart contains the correct combination of items, the user must ensure that each item exhibits the correct Cash Flow Characteristics (CFC’s). It is these unique items and their specific CFC’s that will help managers track, measure and project risk. Since the user can insert accounts, categories, and footing items into the balance sheet, it is useful to have a clear understanding of the difference between these items.

1. Account An account is the basic building block in FORESIGHT. The cash flow risk is measured by accounts, as each account has a unique CFC setting. The user should create an account in the Chart of Accounts for EACH unique cash flow characteristic they identified in the previous section. FORESIGHT’s balance sheet, since it is mapping ALL of our risks in unique accounts, will be full of many detailed accounts. Accounts are identified by a single black box in the Chart of Accounts. 2. Category As bankers, we are used to seeing our balance sheet consolidated by our processor into major groups. This can make reconciliation to the highly detailed FORESIGHT balance sheet somewhat difficult. This is where categories are useful. The user can insert categories and group accounts into these categories. The categories subtotal all of the accounts that they hold. The user can then reconcile back to their traditional ‘consolidated’ GL by comparing it to the Categories in FORESIGHT. Categories can also reduce the time needed to set Account Assumptions. Users can set Growth Targets at the Category level and ‘force’ the information down to the Accounts within the Category. Finally, Categories can be used to group Accounts and Contra Accounts together. Categories are identified by a stack of black boxes in the Chart of Accounts. 3. Footing Category Footing categories are subtotals built into a Chart of Accounts. They are somewhat like categories, which automatically give subtotals of the accounts that belong to them, but they can add up accounts and categories from anywhere in the chart (unlike a category which can only sum the items ‘inside itself’). A typical footing account would be "Net Interest Income". That number should be equal to the total of "Interest Income" minus the total of "Interest Expense". Because interest income and interest expense are not likely to be next to each other in the chart, we need footing categories to calculate Net Interest Income. In addition, we may not have a subtotal for interest income, and may prefer to just add up the income from various earning asset categories. Consequently footing accounts allow the user to add and subtract any number of accounts (or other categories) from any other items in the chart. Footer Accounts are identified by a single blue box in the Chart of Accounts.


Armed now with an understanding of one’s unique cash flows and the Chart Items available in FORESIGHT, let’s review the layout of the Manage Chart of Account screen prior to actually modifying the Chart Items. A screen shot of the Manage Chart of Account Items can be seen below. The user can see their Chart of Accounts displayed on the left side of the screen. Again, the chart is nearly 100% user-defined. The user can insert, delete, move, and rename items on the chart. On the right side of the screen the user can find the Cash Flow Characteristics (CFC’s), Historical balances and current Cash Flows. As the user cycles through the Chart of Accounts, the highlighted item’s data is displayed in the tabs on the right. On the Characteristics tab the user will assign characteristics to accounts. From the Historical Data tab, the user can easily access historical data. On the Cash Flows tab, the user can hand enter or verify and adjust imported balances. In the upper left corner, the user can find controls to move forward to the Account Assumptions (Targets, Pricing, Market Value etc.) or move back to the All Tasks screen. Also note the link labeled ‘Change view’ that is located directly below the box that contains the institution name and above the chart of accounts. When selected a drop down box appears with available options for color coding the chart. The choices are as follows: Normal: this is the default setting which actually means there is no color coding. All accounts appear in black type with a white background. Accounting Type: this setting is probably the most useful on a day to day basis and is also the setting that will help avoid the confusion whether the account selected is in the balance sheet or the income statement. It highlights Assets in light green, Liabilities in light yellow, Equity accounts in gold, which completes the Balance Sheet Accounts. On the Income Statement side Income items are highlighted in dark green, Expense items in dark orange. AFS: this setting highlights only those accounts that have a ‘FASB Designation:’ on the Characteristics screen set to ‘Available for Sale’. Has Related Account: this setting identifies Interest Bearing balance sheet accounts that necessarily have a related account on the Income Statement and vice versa. Interest Bearing: similar to ‘Has Related Account’, but only shows the Balance Sheet item. Interest Sensitive: only those accounts that have the Interest Sensitive box checked on the Characteristics screen for purposes of gap and valuation are highlighted. Instrument Timing: color coding based on the ‘New Instrument Timing:’ found on the Characteristics screen. ‘End of Period’ is yellow, ‘Middle of Period’ pink, ‘Start of Period’ green, and ‘Weighted by Day’ blue. Special Account: shows those accounts designated on the ‘Update Institution Settings’ screen as Special Accounts. Trading Account: highlights the accounts with the ‘FASB Designation;’ set to Trading. Use Default Characteristics: designates those accounts that have the ‘These Characteristics override any existing cash flows’ box checked on the Characteristics screen. User Supplied: reserved for a future enhancement.


Finally, the user can modify the model’s sensitivity to automatic saving of changes. With this check box selected, the model will prompt the user to ‘save’ after every change. This may become very tedious. With the check box unchecked, the model will automatically save all changes prior to leaving any screen.

3.b.i. Inserting New Items There is no limit to the number of chart items the user can add to their model. The only real constraint, when inserting items, is how the user will populate a new account3. For example, the user may want to insert numerous accounts, but there may be no way to import their corresponding balances automatically because of insufficient information in the raw data. That would mean that the user would be forced to ‘manually’ enter balances each month. In this case, adding the extra accounts may actually be inefficient as the time to manually enter balances outweighs any marginal forecasting benefit. New users will find that inserting items and adjusting the Chart of Accounts is an iterative process. As the user gains confidence with the basic steps and general functionality of the model, they will choose to branch out and increase detail by inserting new items. To help understand what items a user should have, and how they will relate to the import function, please refer back to Chapter 3.a. Understanding Cash Flows and Chapter 4.a. Downloading Balances.

Inserting an Account Start by highlighting an account in the Chart of Accounts that is near or above the location where the new account will appear, then click on the Insert New Item link found in the lower left. Here we will insert a New Mortgage, so we will start by highlighting the current Mortgages 30-Year Fixed account in the Chart of Accounts. After clicking Insert New Item, the Insert Chart Screen appears (shown below). If the user selects the incorrect insertion position, it is ok. The user can simply move the account in the future.

The first step in inserting an item is to name it. Users should employ detailed names that clearly telegraph the characteristic’s of the item. Next the user needs to select a type4(Account, Category, Footer). As we are inserting an Account, the user should select “Account” in the Type box. At this point, the user needs to decide between Copying the characteristics of an existing account or Creating a completely new account. When possible, the user should default to using copy characteristics of an existing account. They will save an

3 There are no constraints when inserting Categories and Footers. This is because the model loads balances directly to the Account. Categories and Footers just recombine and display the information that is already present in the Accounts. 4 Refer to the outline above for a clear understanding of the various Chart Item types and their operations.


enormous amount of time and energy that the user would otherwise have to expend setting CFC’s and Assumptions. Minor adjustments may need to be made but it will be easier than building an account from scratch. Use of Certified Accounts are outlined in Chapter 3.c. Copy Characteristics and/or Copy All Assumptions - This copies, or clones, the account the user has selected to create the new account. It saves a lot of time for the user by copying the existing account’s CFC’s and/or Internal Assumptions to the new account.

Update with Selected Account - If the Copy Characteristics check box is selected, this control shows which item’s characteristics will be copied when creating new items. Interest Bearing - With this box checked, the model will insert a corresponding income statement account for the new interest bearing balance sheet account. With the box unchecked, no income account is created. If an error is made, the user can always insert an omitted income account at some future point.

Location – Displays where the model will insert the new item. The user selects the category to place the new item in, or account to place the new item under.

Clicking on the Add button, inserts the new Balance Sheet Account. The Select Related Inc/Exp Account Location screen will appear. This allows the user to define the related Income Statement Account’s location. The user can change the location of insertion for the new Income Statement item by selecting an existing Income Statement item which is directly above where the user would like the new item to be placed. Click on ‘Done’ and the user returns to the Chart of Accounts, displaying the new account inserted. To Move, or Rename the account, see directions in 3.b.ii. below. To adjust the cash flow characteristic, see 3.c. below. Insert A Category As mentioned above, Categories are mainly used to group and sub total accounts. Categories allow the users to easily reconcile the very detailed FORESIGHT balance sheet with the often consolidated Processor balance sheet. Though the user may not want to copy the characteristics of an existing item into the new Category, they will still need to inform the model where the new Category should be placed. This is done on the LOCATION section in the bottom of the Insert Chart Item screen. The user will learn in Chapter 6 that Categories can maintain Internal Assumptions. The user may choose to Clone an existing Category and copy it’s assumptions into the new category. In this case the user would utilize the ‘Copy Characteristics of’ box.


Insert a Footing Category A Footing Category is for display purposes only. It will allow the user to create accounting notes on the balance sheet and income statement. In the example to the right, we will insert a footer named Footing Account. There is no reason to check the Copy Characteristics box when inserting a new Footer item. To assign balances see Edit Footing Category in section 3.c. under Special Characteristics.

Once the user has inserted a new item, they will need to set the Cash Flow Characteristics related to the new item. Chapter 3.c provided directions on adjusting Cash Flow Characteristics for the various Chart Item types. 3.b.ii. Deleting Items The user can Delete Items as easily as they Insert them. To delete an item, highlight that item in the chart of accounts and click the ‘Delete Selected Item’ link in the bottom left of the screen. The user must remember that there is NO undo function. Once an item has been deleted, it is gone forever; therefore, items that are still part of an active budget, forecast, or with needed historical balances should not be deleted. In most cases the users do not want to delete accounts. Instead users want to make accounts inactive. This removes the accounts from the current reports, while leaving them in the budget and maintaining its history detail. The “Inactive” selection is in the Account Status section of the Characteristics screen. An inactive account will not show on reports unless you select that function on the reports Change Features screen. The user may select to show or not show the inactive accounts in the Chart of Accounts screen by checking or unchecking the “Show inactive items” box in the lower left.

3.b.iii. Renaming Items To Rename an Item, click on the item in the Chart of Accounts. That item’s characteristics will appear in the Characteristics tab. Click in the Name box on the upper right and retype the new, correct name. When you select another account the name will change on the chart of accounts. Or click on ‘Done’. This will also change the name of the item and return the user to the ‘All Tasks’ screen. For all interest bearing accounts, the user will get a message box asking if they would also like to rename the related Income Statement account. In almost all cases, the user should maintain their income statement and balance sheet as identical; therefore, they should allow the model to rename the Income Statement Account to match the Balance Sheet account.

3.b.iv. Moving Accounts and Categories The user can move any item in the Chart of Accounts to make their reports easier to read and understand. There are three ‘typical’ types of moves in FORESIGHT. First, the user can move


categories around within the Balance Sheet or Income Statement. Second, the user can reorder items within categories. Finally, the user can move items from one category to another. Moving Categories around the Balance Sheet and Income Statement In the example below, we will move the category named NEW LOAN GROUP out of the REAL ESTATE category and into the LOAN Category. Start by clicking on the category to be moved. Drag and Drop the highlighted category to its location. In this example, the user will drag and drop New Loan Group into the Loans category.

Note: Moving items in one section only affects that area, not the related area. For example, reordering the loans on the balance sheet will not change the order of items in the income statement. The user would have to manually reorder the income statement to match the new layout of the balance sheet.

If the user is just re-ordering the Category within its current Category, the model will simply post a “Yes/No” question as to whether the category should be moved. If the user is moving the category into a new category, the user will be prompted with the message shown below in the lower left. By selecting YES, the user will ‘drop’ NEW LOAN GROUP category into the REAL ESTATE category. Since it started there, essentially the user is simply moving a category around inside of its current category. If the user selects NO, the NEW LOAN GROUP is moved out of REAL ESTATE and ‘dropped’ inside of the LOANS category. In this example, we will say NO and move the category out of its old location and into the new.

Moving Items within Categories (Re-ordering Items) To rearrange items within a category, click on the item to be moved and drag and drop it to its new location. Since the user is only moving the item within a category, the model will simply post a “Yes/No” question as to whether the item should be moved.


Moving Items Between or Into Categories In this example, we will move the Account “RENAMED 30 Year Fixed” account out of the Real Estate Category and into the New Loan Category. To move this account, the user must first highlight the account. The user then drags the account and ‘drops’ it on top of the account or category into which they would like to move it. In this example we will say Yes – move into the category.

A message box appears to prompt the user to decide if they would like the account dropped INTO the category or UNDER. In this example, we choose INTO and the RENAMED MORT account is shown displayed inside the New Loan Group (below on the right). Had the user chosen ‘Under’, the account would be dropped into the LOAN account, but it would be displayed ‘Under’ the NEW LOAN GROUP category.

By this point, the user has identified their unique cash flows and decided what Items are necessary in the Chart of Accounts model to reproduce them (Chapter 3.a). The user should have also inserted the needed items, modifying the Chart where appropriate (Chapter 3.b). Now we need to focus on verifying the settings attached to each Chart item. It is these settings that will direct the model to reproduce the correct cash flows for originated dollars or possibly use some characteristics if they are missing in the download. 3.c. Setting Cash Flow Characteristics for New and Existing Accounts As outlined earlier, FORESIGHT offers the analytical tools an institution needs to make prudent risk management decisions, quantify any existing and projected balance-sheet performance and risk. It integrates interest-rate risk management, strategic planning, and budgeting. To perform these and other tasks, FORESIGHT must be able to reproduce an institution’s unique Cash Flows. With one’s Chart set, we will


now check and verify that information relating to Item characteristics is correct. Periodically the user should walk through the Chart of Accounts and re-verify the settings of the CFC’s. To truly grasp the concept of CFC, the user needs to see their model in two separate time frames: Current and Forecast. When the user ‘closes the books’ on a particular period and loads the balances, these balances represent current information in the FORESIGHT model. Once these balances are loaded, the user can look at a current balance sheet, a current gap report and current market value estimations. These reports all focus on how the current balances will run-off in the future. There is no assumption (at this point) about new balances. In real life, what controls how the current balances run off? According to information provided in Chapters 1 and Chapter 3, one can see that Cash Flow Characteristics and market changes drive how current balances will run off. In the model, what will control how the current balances run off? The model allows the user to either import specific CFC’s from the RAW data (for each loan, CD, checking account, etc.) and attach those characteristics to accounts, or specific instruments within accounts. These download-driven cash flow characteristics will produce the most robust cash flow estimations. When the RAW data is incomplete, the user can still estimate the balance run off by using the CFC’s set in the model. During a forecast, the balances that are added in the future will always generate their future cash-flows based on the cash flow settings found on Characteristics tab. Since account CFC’s can be set and maintained three different ways (CERTIFIED, USER DEFINED or download data driven) it is important to understand these options clearly. Certified Accounts (removed in version 5.2) Certified Accounts are created and managed by Farin and Associates. They provide a quick and efficient way for users to insert accounts and guarantee that the instruments react in an industry accepted way. If the user chooses to define their own instruments, it will be their responsibility to verify the settings. Users can insert Certified accounts during the insert process or a user can overwrite the current characteristics settings by changing an existing account into a Certified account. Either way FORESIGHT will access the Certified Accounts database at Farin only once (at insertion or when overwriting). After that there is no communication linkage between Farin and the client database regarding this issue. While an account is ‘Certified’ there are only limited parts of the provided definition that can be changed. The ‘Certified’ status can be removed from an account. This can be helpful as a starting place when creating an account’s definition. However, Farin & Associates make no assumptions about the logic of the definition or expected results from non-certified accounts. User-Defined Accounts User-defined accounts are ones in which the user sets the cash flow characteristics on the Characteristics tab to create an account. User-defined accounts are no longer Certified as correct. Users trying to employ user-defined accounts bear the responsibility that an account has the correct CFC’s settings. The user-defined settings will be used as the CFC’s for all originated forecast balances or missing CFC’s from the download. If the Apply These Characteristics button is checked, the model will write over any CFC’s that are being imported with the User-Defined settings on the Characteristics tab.


Download Data Driven When the user imports their RAW data, cash flow characteristics are imported into the model. The user can ‘force’ the current balances to take on those characteristics by unchecking the Apply These Characteristics button. Characteristic that can be imported are outlined in Chapter 3.e. As noted above, regardless of the condition of the Apply These Characteristics button, all money originated during a forecast will carry the cash flow characteristics displayed on the characteristics tab. Below the user will find directions to set the Cash Flow Characteristics for all FORESIGHT Chart Items. The Characteristics tab is broken into four main parts: Account Identifiers, General Characteristics, Cash Flow Characteristics, and Special Characteristics. We will first outline the setting up of Balance sheet accounts - as they are the most complex. Then we will outline the settings for Income Statement Accounts and Footers. Name & Type The name is a combination of alphanumeric characters and is not a unique identifier. The first step in setting CFC’s is to verify that the name clearly identifies the account. For example, “Mortgages - Adjustable” does not tell us anything about the account. Is it a 1-1 ARM, 3-1 ARM, 7-1 ARM? Instead, the user should create an account for each type of ARM they will carry in the future and assign it a detailed name. Again, since there is no limit to the number of accounts a user can employ in FORESIGHT, one should strive to build their model with as much detail as possible while maintaining an efficient chart of accounts. The second step is to choose the correct Type of account. These types will restrict what the user can do with each item. The options available match the various sections of the common financial reports. Since the Chart is nearly 100% user defined, there is nothing to stop the user from displaying an Asset account in the Liability section. In fact the model does not require users to break their balance sheet up into Assets, Liabilities and Equity. This same thing is true for the Income Statement. The user can order their income accounts as they see fit. The flexibility requires that the model user ‘know’ where the accounts ‘should’ go under normal accounting rules.

1. Undefined - Not used. 2. Asset - Balance Sheet Account 3. Liability - Balance Sheet Account 4. Equity - Balance Sheet Account 5. Income - Income Statement Account 6. Expense - Income Statement Account 7. Statistical Balance- Other Accounting Notes (behave like a B/S account) 8. Statistical Flow - Other Accounting Notes (behave like a I/S account)

Contra Account: A contra account will offset itself from its own Type.


General Characteristics Next the user will need to define the basic Payment and Interest Calculation methods related to the account. These settings include designating Interest Bearing and Non Interest Bearing accounts, Interest Payment Frequency and Interest Calculation Methods. Interest Bearing This check box determines if interest will accrue in the selected account. The related income account for this item can then be identified in the Related Account box (in the bottom center of the Special Characteristics section of the Characteristics tab). Interest Sensitive This check box determines if the line item will appear on a GAP report and if market value is calculated. Payment Frequency This is the default setting on how often the selected account will ‘book’ either interest or dividends for accrual purposes. This is different from the actual receipt of the interest or dividends. For example, U.S. Treasury Bonds pay interest semiannually, but a financial institution will ‘book’ the accrued interest on their holdings of U.S. Treasury Bonds every month. Interest Method This determines the annualization and deannualization method that will be used in the selected account’s interest calculations. For example…

Given a 5% coupon and a $1200.00 average balance in the month of August, the following illustrates the differences:

Average Balance Coupon Income/Expense 30/360 $ 1,200.00 5.00% $5.00 Actual/365 $ 1,200.00 5.00% $5.10 Actual/360 $ 1,200.00 5.00% $5.17 Actual/Actual $ 1,200.00 5.00% $5.10 ($5.08 – leap year)

Cash Flow Characteristics (CFC) The top priority in setting CFC’s is to make sure that the model’s mathematical expectations of the cash flows match the user’s real world expectations. Reviewing the information presented in Chapter 3.a will be helpful at this point. In this section the user will define the amortization, reprice, call, prepay, and decay characteristics of their instruments. First, it is helpful to identify the meaning of: Matures in, Amortizes in, Balloons in and Non Maturing.

• Amortizes In: Used for Amortizing Instruments. Relative to beginning date, this characteristic determines over how long an instrument in this account will amortize. If the account is interest bearing the Payment Frequency will determine how often interest is accrued. The period length chosen here will determine principal payment schedule. The Delay option can be used for delaying amortization on construction loans, i.e. construction to permanent mortgages (Amortizes 186 Months and Delay 6 Months). Amortization Types are: Level Yield, Level Amount and Sum of the Digits (see end of chapter for more information).

• Balloon In: Used for Balloon Instruments. Relative to a beginning date, this will determine when an instrument in this account, if amortizing, will balloon.


• CMO: Tells the model this item is a CMO and cash flows are prepared for the open/closed window information.

• Matures In: Used for Bullet Instruments. Relative to a beginning date, this characteristic determines when an instrument in this account will mature. If the account is interest bearing the Payment Frequency will determine how often interest is generated.

• Non Maturing: Used to designate immediate liability instruments like Checking, Savings, etc. Any instrument without a defined maturity date, which the user expects to DECAY off the books at an identifiable speed, should be treated as Non Maturing.

• Servicing: By entering your servicing percent on a Statistical Balance (OBS) account that holds the cash flows, the model will forecast the servicing income for the related account (non interest). The cash flows will also need the Servicing field completed either by hand entering or through your download using the system field called ‘Servicing’. The only income realized on this account will be the servicing income.

The user must employ the proper combination of these options to create the proper CFC’s for each account. For example, see Appendix C. Then the user must designate the repricing and call characteristics of the instrument.

Repricing Schedule Relative to a beginning date, this will determine when the individual instruments in this account will first reprice and then how often after that date they will continue to reprice. The first reprice is set in the ‘Reprices in’ option and the frequency is set up in the ‘and every’ option. We recommend consistently using the same time period (month, year etc) that is used in the download file. If the download is missing any cash flow information Foresight will default to the characteristics screen for the missing data. Call Options Relative to a beginning date, this will determine when instruments in this account will first attempt to call and then how often after that date they will continue to attempt to call. The first call date is set in the ‘Call Option in’ option and the frequency is set up in the ‘and every’ option. We recommend consistently using the same time period (month, year etc) that is used in the download file. If the download is missing any cash flow information Foresight will default to the characteristics screen for the missing data. The ‘If’ selection box is for the Call Rule and governs the new money in a forecast. Caps/Floors If the instruments in this account reprice, originated money or existing instruments that might be missing data from the download file will be governed by the cap/floor entered here. There are implicit caps of 100%. There are implicit floors of 0%. Caps and floors are expressed as decimal rates, not basis points. A periodic cap of 0.25 represents a quarter percent, or 25 basis points per month.

• Periodic Cap/Floor – Specifies the maximum amount that an instrument’s repricing rate can change in one month.

• Anniversary Cap/Floor – Specifies the maximum amount that an instrument’s repricing rate can change in one year.

• Lifetime Cap/Floor – Specifies the maximum amount that an instrument’s repricing rate can change throughout the instruments life. (Lifetime rates are relative to the pricing rate on originated dollars or relative to the current rate on existing instruments that might be missing data from the download file.) A lifetime cap/floor overrides annual or periodic caps and floors that could result in larger values. An annual cap/floor overrides a periodic cap/floor that results in a larger value.

• Absolute Cap/Floor – Specifies the maximum rate that the instrument will reprice up or down to. For originated dollars this would be the unconditional cap or floor.


Has Fees By entering a loan fee percent here the model can forecast the fee and include it in income for yield purposes. The cash flows will also need the Fees field completed either by hand entering or through your download using the system field called ‘Fees’. The amount added to income is calculated as: (Amort. + Prepays/Beg. Balance) x Fees (Fees = % x Beg. Balance). Collateral Spread By entering a spread percent here the model can add this to the current rate for Prepay table use. The cash flows will also need the Collateral Spread field completed either by hand entering or through your download using the system field called ‘Alternate/Underlying Coupon’. Prepayment/ Decay Tables These tables can be calculated by outside sources, such as McGuire or the OTS, and imported directly into the model. The user may insert user-defined tables if they have calculated those values. To see more about inserting and populating prepayment and decay tables see Chapter 5. Prepay Factor Changing the factor will increase or decrease the prepay speeds used from the tables. Credit Quality/Early Withdrawal Table Choose a previous created prepay table (user defined prepay table set up, see chapter 5, page 2) and apply that table to the balances on the account. A table or tables could be created and used to forecast non-accruals or early withdrawals from Certificates of Deposit. Non-accrual use: one or more tables can be created under the Prepayment/Decay Tables screen using the Table Type as ‘Credit Quality Table’. Specific tables are then selected on the Characteristics screen. The factor (in basis points) times the balance after prepays and amortization is entered on the Non accruals line of the target screen and reduces the ending balance of the account. This amount is added to your Non accrual accounts balance each period. The non accrual account is placed on the same level as Total Loans and uses Related Account target method. Set method option to Percentage of (Summed); Use Account select Total Loans; Balance to Use select NonAccurals; and Target is set as Ending Balance. Early Withdrawal use: one or more tables can be created under the Prepayment/Decay Tables screen using the Table Type as ‘Early Withdrawal Table’. Specific tables are then selected on the Characteristics screen. The factor (in basis points) times the balance after maturities is entered on the Prepay/Decay line of the target screen and reduces the ending balance of the account. New Instrument Timing: The user can select when in the forecast month newly originated balances are added to the account. The current selections are End of Period, Middle of Period, Start of Period and Weighted by Day. The first three timing parameters are self-explanatory. Weighted by day means that the cash flows being thrown off during the period are day weighted to determine the day that the new instruments will be placed on the books. One caution here is that any additional originations will also be added on that day. Best use would be with Investments where you are not consolidating and larger dollars are involved. These Characteristics Override . . . . As mentioned above, based on the status of this check box, the CFC’s for current balances can either be set by the Characteristic screen or by the information found in the RAW data. With this box checked, current balances will run off according to the characteristics on this screen. With this box unchecked, CFC’s will come from the raw data, except for NEW money originated in a forecast, those CFC’s will come from the Characteristics screen.


Special Characteristics The final step is to put the ‘finishing touches’ on the account. These final settings include Risk Weighting, FASB settings, Account Status, Related Accounts and Taxes Risk Weight / Risk Category The risk weighting, between 0 and 100%, is used for risk weighted capital estimates. This generally only applies to Assets and some Off Balance Sheet items. Specific risk weighting figures are provided by the regulatory agencies. This setting will differ between banks and credit unions. Credit Unions will select an appropriate category from the drop down selection. FASB Designation This setting determines the accounting rules that are applied to an Asset account for FASB 115 purposes. The three FASB control accounts must be set to AFS. They are the investment account ‘FASB 115 Adj’, the liability account ‘FASB 115 Tax Payable’, and the equity account ‘Net Unreal G/L on AFS’. If you do not wish the model to calculate the FASB 115 and target formula is added to the ‘FASB 115 Adj’. See ‘FASB 115 Accounts – Setting Up’ in the Help and Reference Index for additional information.

Related Account If an account is interest-bearing then Related Account is the default account on the income statement that receives the income/expense from the balance sheet account. This is non-exclusive. An income statement account could receive income from many balance sheet accounts. If the selected account is a Footing Category, the user will change the footing calculation via the Edit button to the right of the related account option. Account Status - Inactive The Inactive checkbox allows the user to maintain the account in Budgets and History, but it hides the account so that it will not show up on current reports or forecast reports. Often when a user believes they want to “delete” an account, they really want to make it inactive. Account Status – Certified (removed in version 5.2) The Certified checkbox indicates whether the account is a certified account. A user defined account can be turned into a certified account by clicking on “Use Certified Accounts” and selecting an appropriate certified account from the list. Taxable Income % For all income accounts the user can assign the appropriate taxable income percent for Fed, State and Local taxes. The combinations of the settings outlined above allow the user to reproduce the Cash Flows of any type of ‘real-life’ financial instrument. Setting Up FASB 115 Accounts Special Accounts Two out of the three FASB 115 accounts need to be assigned in the special accounts section of the Update Institution Settings screen: FASB 115 Gain/Loss is assigned to the appropriate equity account and FASB 115 Tax is assigned to the appropriate liability account. Follows find information on the proper way of setting all the FASB 115 accounts. How to properly set up the FASB 115 accounts:


The model will calculate Net Unreal G/L on AFS by checking the 'Apply FAS115 calculations' box at the bottom of the 'Select a Report Task' screen. This will run a market value on all of the investments that have been set to 'Available for Sale' and calculate the difference between that and the book value of those accounts. If there is an Unrealized Gain, then the model will compute the Tax payable and enter that in the NIB Liability, and the balance will be applied to the Equity account. If the difference is a loss, the entire amount will be applied to the equity account. If the user does not want to run the reports with 'Apply FAS115 calculations' selected there is a way to have Foresight simply carry the historical values forward in a forecast report. To accomplish this a reserved word is available to be used in setting a formula target for the FASB 115 Adj account in the investment section of the Chart of Accounts. The formula to be used is: PE.CARRYFORWARDHISTORICALFAS115LIABILITYTOTAL(). In order for this to work properly the model needs the three FASB 115 accounts set correctly. The three accounts are the investment account 'FASB 115 Adj', the liability account 'FASB 115 Tax Payable', and the equity account 'Net Unreal G/L on AFS'. All three of these accounts need to be set as 'Available for Sale' in the 'FASB Designation' box in the Special characteristics section of the Characteristics screen. The final step is to place the above formula into the target screen for the FASB 115 investment account. Once these steps are completed and a forecast is run without FASB 115 Calculations checked, the model will carry forward the historical values found in the three FASB 115 accounts. Note that FASB 115 Adjustments are made after balancing. This is possible as the total adjustment is the same to both sides of the balance sheet. Setting Category Characteristics The user needs to verify the NAME, TYPE and the RELATED ACCOUNT. Again, the related account shows where the model will display the related income/expense on the income statement (another category). There are no other special settings for the Characteristics of a Category. Setting Income Statement Characteristics The user needs to verify the NAME, RELATED ACCOUNT and TAXBLE INCOME %. The related accounts show the balance sheet accounts which will drive the income/expense (another category). There are no other special settings for the Characteristics of an Income account. Setting Footer Account Characteristics The user needs to verify the NAME, TYPE and the RELATED ACCOUNT. There are no other special settings for the Characteristics of a Footer Account. To verify the RELATED ACCOUNT, the user is verifying the Footer Formula, as shown to the right. The formula currently set is to add total Liabilities to total Equity. To modify the formula, the user would click and drag an item from the Accounts list on the right into the formula box on the left. The model assumes you wish to add it, but you can just type a minus sign to subtract instead. Footing category definitions are currently limited to addition and subtraction.


3.d. Historical Data The Historical Data tab provides the user with quick access to Historical data and trends. The user can manually enter balances (current or historic) into the History Balance grid. There is very little actual work that is done on the history tab. For users that import their GL with their other files, the Historical Data tab is really a reference or a display. For institution that hand enter balances into the model each month, they may enter those balances in the Historical Data. To hand enter ‘other’ balances (i.e. balances of non-cash flow accounts like Building, REO, Other Liabilities) the user will highlight in the End of Period box for their current data and enter the correct numbers. Users may also hand enter actual income into income accounts in the same manner. 3.e. Cash Flow Screen For the user downloading all of their balances, this screen provides a reference and checkpoint for their data. Some users may have to hand enter some types of balances, such as FHLB Advances or Investments. From this screen the user can verify loaded balances, modify loaded balances and hand enter balances if needed. Ending Balance Displays the total balance associated with the selected account. As of


Displays the ‘as-of-date’ for the cash flow data shown. Weighted Ending Rate Displays the weighted ending rate of all the cash flows. Hide Unused Columns Not all of the columns on this screen are utilized for all accounts. For example, FLOORS and CAPS would not be needed for a Fixed Rate account. Since these fields would be blank, the user can automatically collapse these fields. Insert Blank/Default Row When manually entering detailed cash flow balances, the insert row option will insert a new line for entering cash flow data. For example, if the user is hand entering FHLB Advances, they will need to Insert a new Row for each advance. You can either insert a blank row or a default row. The default row will look at the characteristics screen and enter the information needed to build a cash flow. Delete Row This option allows the user to delete unused rows. Copy from Allows the user to copy cash flows from a prior period into the current period. This includes all balances, rates, dates, call rules, open/closed window information etc. Enter in This option allows the user to see cash flow balances in dollars, thousand, or millions. Show XML A diagnostic tool for our developers relating to cash flows.


Appendix - Amortization Types Foresight currently supports three amortization methods – Level Yield, Level Amount and Sum of the Digits. Level Yield Assume a cash flow statement in the amount of $100,000,000 amortizing for 60 months at 5.00% with a 30/360 interest method. Payments are received on the 16th of the month for this instrument because we assume the payment date is the same as the specified maturity day. However, if the cash flow statement does not have a specific maturity date, but just a maturity term, we use middle of the period for payment date. A 30/360 interest method yields a deannualization factor of 0.08333. The payment comes out to 1,887,123.36, which can be verified in Excel using the same principal, 60 periods, and a rate of 5% * .08333. Note this amount includes principal and interest. As interest is paid in arrears, the interest portion of this for the prior month is 416,666.67. That leaves a maturity paydown for the current period of 1,470,456.70. Interest for the current month is calculated as total interest on the starting balance less interest forgone as a result of the principal payment on the payment date. In this case the interest payment for the current period will be 416,666.67- 3,063.45 = 413,603.22, but will clearly vary depending on payment date: 100,000,000.00 * 0.05 * 0.08333 = 416,666.67 full interest 1,470,456.70 * 0.05 * 0.08333 * 15/30 = 3,063.45 less principal paid Note that all financial instruments or cash flow statements in Foresight are assumed to be pools* and as such the payment amount is recalculated each period. This means the payment may vary particularly as prepayments reduce the principal in the pool. But without options being exercised, and assuming a 30/360 interest method, the payment will remain the same. The payment on instruments with no prepayment will vary slightly with different interest methods because the annualization factor will vary depending on the current month. If the payments are received on the last day of the month then interest amount computed out of the payment function will coincidence with the interest paid for the month. This is because the accounting period for the interest that is paid in arrears and the current month begin and end at the same time. *The alternative, which is to treat each cash flow as a single account requires that history be stored for every single transaction, and also would require that a prepayment probability model be applied to transactions every period. That technique would add considerably to the calculation complexity and time for most models. Level Amount Assume a $100,000,000 loan for 60 months at 5.00% with a 30/360 annualization technique. The payment will be 1/60 of $100,000,000. Sum of the Digits Beginning Balance / (n * (n + 1)) / 2, where n is the number of periods remaining.


Appendix - Characteristics Setting Examples (the Use Certified Accounts feature was removed in version 5.2, samples do not include new features from version 5.2.5.7)


Chapter 4

Populating your Account Balances (The Download) Populating the model is the most complex process in the FORESIGHT product. One needs to automatically import, or manually insert, cash flow balances in such a way that the model can identify the detailed and unique CFC’s for each and every instrument. Without detailed cash flows the model will be unable to perform complex Market Value calculations, GAP Reports, or run accurate forecasts. While reading through the next few sections the user must bear in mind that every institution’s data will be complex and different. Even institutions that use the same processor may receive files with drastically different layouts. Or, if the layouts are similar, institutions may use different definitions for Type Codes, Class Codes, Purpose Codes etc. Therefore, there is no one way to set up a download. The user is required to make numerous assumptions and may find themselves changing their downloads several times before they capture their true cash flows. 4.a. Downloading Data Automatically FORESIGHT allows the user to import their unique RAW data files directly into the model. To make this completely user-defined process work, the user must understand how their data is being created by the processor and how FORESIGHT interprets that data. Chapter 4.a. will cover understanding your RAW data, designing the FORESIGHT download components, and importing different types of data into the model. Typically the user will not have to adjust the download or build new descriptions from period to period. Only in the initial set-up, when the user is trying to adjust the model, will the user make drastic download changes. For most users, the process of loading data each month is very streamlined. The streamlined-monthly process is outlined below starting in Chapter 4.a.ii. Before skipping to that section, it is important for the user to understand how their current download was built (around their unique files) and how to make adjustments if needed.

4.a.i. The Download Process When one ‘looks’ at an institution’s RAW data, it may appear as a line of confusing numbers, or it may look like an Excel spreadsheet with randomly filled columns (see diagrams below). Either way, without an understanding of what information is contained in the data file and where the information is located within that file, the user will not be able to import their data. In the next section the user will learn how to identify the important components of their data. The user will then learn how to translate these components into a Download Description.

Fixed position file


Delimited Data (tab, comma, etc.)

4.a.i.1. Understanding Your Data There is a lot of information inside of an institution’s RAW data (see diagrams above). There are balances, rates and maturity dates. These are often easy to locate. There are also codes that identify the cash flow characteristics (see Chapter 3) of each account – these codes are called Cross Reference Codes. These codes are sometimes difficult to identify and understand. Examples of Instrument Characteristics that will affect future Cash Flows are listed below. Again, these Characteristics will only appear in RAW data files – not in the GL. Furthermore, this list is not meant to be all-inclusive.

• Original Term Code • Amortizing Term Code • Repricing Schedule (time to first reprice and time to subsequent reprice) • Call Schedule (Time to 1st & Time to Subsequent Calls) • Balloon/ Bullet Characteristics • Floors & Caps (Periodic & Lifetime) • Participations (Bought or Sold) • Reprice Dates • Call Dates • Indexes and Offsets


If setup correctly, FORESIGHT will employ various combinations of Cross Reference codes in order to separate and identify an institution’s unique CFC’s. As outlined in Chapter 3 the user can then choose to use these CFC’s from the RAW data or employ the user defined CFC’s at the account level. Either way, the codes will help the user ‘map’ the lines from the RAW data into the specific accounts in the model. ‘Mapping’ the raw data to specific accounts is how FORESIGHT populates the Chart of Accounts with current balances and rates. As mentioned above, without detailed cross-references, one cannot build detailed instrument cash flows. Cross-reference code identification (users may refer to them as Type codes, Product codes, Class code, Collateral code, Major Type, Minor Type, Original Term, Reprice Frequency, Purpose codes, Cusip, Investment Types, etc) is the area that tends to cause the most problems while building a download and loading data. The list below shows the most common types of cross-reference codes that are required to separate raw data. Each bank will need to employ different combinations of codes according to their mix of real-world instruments. At a minimum, the model will require the following information.

• The user needs cross-reference codes to help them separate commercial, mortgage, and consumer loans. Then they need codes to help them sub-divide those categories further into HELOCs, Auto, Student, Construction, Land, Term Loans, Time Notes, Letters of Credit, and LOC commercial, to name just a few.

• On the liability side, users will need the codes to separate savings, checking, NOWs, MMDA, etc. They will also need to be able to separate NIB (non-interest bearing) and IB (interest bearing) accounts. Furthermore, one needs codes to separate IRAs, Roths, and regular CD’s if the cash flows are different between the instruments.

• If your Type/Class codes do not denote original term, users will need to employ an Original Term field in their data. Users will need to segregate their loans between 15 year and 30 year maturities or 5 year and 10 year maturities, and all appropriate maturities between. They will also need codes to separate 30 day CDs from 5 year CDs, and all appropriate maturities between.

• If not clearly defined in the Type/Class codes, additional fields are required to separate fixed and variable rate instruments. Also, if not clearly defined by the Type/ Class codes, additional codes are needed to outline when the FIRST rate change occurs, next reprice date and the FREQUENCY of rate changes after that.

• For Investments or Borrowings, users need to be able to separate callable and putable instruments. One needs to know the time of the FIRST call, next call date and the FREQUENCY of subsequent calls. Instruments with different call characteristics need to be separated, just like instruments with different repricing schedules in order to originate balances properly.

• CMO’s – users will need a Bloomberg print out (or an equivalent) for each CMO. This sheet should provide detailed information about the changes in payment windows and weighted average life due to future rate changes. Specific cash flows are not needed; FORESIGHT will create the cash flows using the open/closed windows, PSA and average life.

These type of codes will help the user separate their RAW data into specific instruments with unique cash flows. In Appendix A and B the user will find a set of procedures to separate different types of instruments from the RAW data. Examples are provided for Investment files, Loan files, Deposit files, and Other Borrowing files.


4.a.i.2. Identifying the Important Components of Your RAW Data Files To identify the location of specific components in their RAW data (balances, rates, maturity dates, cross reference codes), the user must contact their processor. The process can provide file layout information to the user for each data file. An example of a Processor’s layout mapping can be seen below. The user should note that the layout report displays a lot of useful information. First, the field description provides clear and understandable descriptions of the data. Second, the exact length, field start and field end positions are readily identifiable. Finally, the last column provides “translations,” which show the meaning of the codes that will be found in each field. It is imperative that the user obtain layout information that contains ALL of this information. 4.a.i.3. Creating ‘Descriptions’ or Download Mappings To import a data file into FORESIGHT, the user must first build a download description. The description contains a set of rules which describe for the model what data is contained in the files, where it resides, and what type of format it has. Below we will outline how to Insert a new download description. Prior to creating a description, the user must read through the information above and in Appendix A and B. Without a clear understanding of the unique data, the user will be unable to build or modify a description.

From the All Tasks screen select ‘Manage File Descriptions’. The Edit File Descriptions screen will open. In the upper left the user can see their current list of File Sources. All users should start with descriptions preset by Fain support staff. If a description already exists, the user can modify that description by following the same steps outlined in the building process.

To insert a new file description click on the ‘Insert New File’ link. The Add a File Source window opens.


The user will select the purpose of the import file.

• Import Cash Flows – This description should be used for current files with cash flow characteristics like loan files, savings files, FHLB Advance files, investment files, etc.

• Import History Values - This will be used only for G/L balance sheet and income statements. There will be no cash flow characteristics associated with these balances.

• Import CMO Prepayments – This file source will be used to import your CMO window information from a raw data file.

• Import Market Values – This file source will be used to import override market values in selected rate environments from a raw data file.

• Import Forecast Values – This file source will be used to import current budget numbers into the Foresight model.

The user must then select the specific file layout. Examples of these two types were shown above (4.a.i.).

• Fixed (txt, dat) • Delimited (csv, txt)

The user must carefully select the proper choices at this stage. Once a description type and date type have been set, this cannot be changed. The user would have to insert new descriptions and delete the improperly set descriptions. When the user clicks insert, a blank description will appear in the file source box. It will be named [New Data Source]. The user must now modify this blank description to fit their new file. First, the user will want to update the options of the description found in the location of the upper right corner. The user should change the name by highlighting the Name box and typing in the correct name. The user will then browse to the input file location. This is the location of the RAW data file to be loaded. Finally, the user will want to complete the file format description in the gray box. The gray-box settings allow users to specify their data layouts. Often, the user can skip header and footers and identify page breaks and data.

Next, the user must define the fields they plan on importing into the model from the data. To add fields, click on the insert button. New import fields appear in the bottom center of the screen. While working in the download it is useful to remember that a File Source is the complete set of rules to load a file, while a Field rule is the information required to import a single piece of data; Balance, Rate, Maturity Date, Etc. The user must create a detailed rule for each field to be imported. Field positions and length show the model where to find that information in the raw data. This information can be found in the File Layout document from the data processor.


The user then identifies the type of data found in that field. Options are:

Amt/Rate - a numeric sequence of characters that makes up a number with decimal places (012345678901.) - there might be no decimal place but it is an allowable option (i.e. 123 is treated as the same number as 123.000). Integer - a numeric sequence of characters that makes up a number with no decimal place (0123456789) - a decimal place is not allowed. Even though 123.000 is the same number 123, 123.000 is not allowed. Note: Typically used for Cross Reference codes. When using an integer there is a maximum allowable length of nine characters the model will read. String - a collection of alpha numeric characters. Punctuation is allowed. A string can't be add, subtracted, multiplied or divided. Reserved as cross-reference codes only. Date - a group of characters in a supported format that is recognizable as a date (MM/dd/yyyy, yyyy/MM/dd, dd/MM/yyyy, etc) True/False - a set of data with only two possibilities (0,1, -1,0,-1,1, T,F, True,False)

The non system box allows the user to designate a field as a cross reference code or as a unique model field. Cross reference codes are non-system accounts. That means, the model will not do any ‘math’ or ‘manipulation’ with these fields. Examples of System Accounts are Balance, Rate, Maturity Date, etc. For a complete list of System Accounts, see the table below. Examples of different completed Descriptions can be found in Appendix D.

Designating a row as a certain type of system field is the way the user forces the model to treat balances like balances, rates like rates, premiums like premiums, etc. To specify an account’s system field designator the users clicks on the ‘x’ to the right of the system field. When selected the “Select Data Item” window opens. The system fields are outlined below – separated into three tables (characteristic types, balance types and cash flow types). The settings, defaults and related rules are outlined for each option. For detailed examples, see Appendix D.

Characteristic Types Field Type Description Default Rules Account Name String Name of account. Branch ID String Identifies the Branch by

matching to the Branch's nickname field.

Asof Date Date When matching values for any historical or forecast period, this date becomes the index used for all values in the same data record.


Balance Types Field Type Description Default Rules Accretion Amortization Rate Amount Amortizing Points Rate Amount Average Balance Rate Amount Average Rate Rate Amount Overruled by any cash flow

rate Amortized Cost Rate Amount Overruled by any cash flow

Amount Balloons Rate Amount Calls Rate Amount End Of Period Rate Amount Overruled by any cash flow

Amount Fees Rate Amount Income Expense Rate Amount Same as Flow Data, just

alternate name Flow Data Rate Amount Same as Income Expense MarketValue Rate Amount Maturing Amount Rate Amount Originations Rate Amount Premium Discount Rate Amount Expressed as +(premium) or -(discount)

Prepayments Rate Amount Principal Prepaid Rate Amount Repricing Rate Amount

Cash Flow Type Field Type Description Default Rules Amortize Date Date The maturity date of an amortizing instrument, which

is the end of the amortization period. In general it is better to use End Date which allows the characteristics of an account to determine if it amortizes. Use this field if an account can contain both bullet and amortizing instruments.

Amortize Period Type Integer Length of an amortize period Monthly Amortize Periods Integer Number of periods to amortize, alternative to

specifying Amortize Date

Amortize Type Integer Enumeration of Amortization Types 2 (level yield) Amortizes Yes/No Defines an amortizing instrument Amount Double Becomes the book amount for this cash flow

definition.

Annual Cap Double Annual Floor Double Balloon Date Date Balloon Period Type Integer Length of a balloon period Balloon Periods Integer Number of periods until balloon, alternative to

specifying Balloon Date.


Balloons Yes/No Defines a balloon instrument Call Date Date Call Index String Call Offset Double Call Period Type Integer Length of a call period From account

characteristics.

Call Periods Integer Number of periods between call events. From account characteristics.

Call Price Double Absolute call price. Call Start Integer Number of periods until first call, alternative to

specifying call date.

Call Start Type Integer Length of Call Start periods. Monthly Calls Yes/No Defines a call event End Date Date This date is interpreted as either the bullet maturity

date or the ending amortization date depending on the characteristics of the account it is applied to.

ID String Identifies this particular instrument in multiple periods of cash flow data. CUSIP is an ideal value to assign to ID for securities.

Internally generated id

Life Cap Double Life Floor Double Mature Date Date The maturity date of a bullet maturity instrument. In

general it is better to use End Date which allows the characteristics of an account to determine if it is a bullet maturity. Use this field if an account can contain both bullet and amortizing instruments.

Mature Period Type Integer Length of a maturity period Monthly Mature Periods Double Number of periods until maturity, an alternative to

specifying Mature Date or End Date.

Matures Yes/No Defines a bullet maturing instrument Original Term Integer Original term of instrument in months. Origination Date Date Origination date of instrument. Par Double Par value. If no Amount (book value) found for this

instrument, Book = Par + Premium/Discount

Payment Amount Double Regular payment amount of an amortizing instrument. Use this field if using transformation rule to calculate an end date from the payment amount.

Percent of Total Double Can be used instead of amount to designate a cash flow definition amount, refers to percent of current balance.

Period Cap Double Period Floor Double


Premium/Discount Double Premium (positive) or Discount (negative) amount for this instrument. If Par not supplied, then Par = Book - Premium/Discount.

Rate Double Becomes the stated rate for this instrument. Reprice Index String Reprice Next Integer Not used. Reprice Offset Double Reprice Period Type Integer Length of a reprice period From account

characteristics.

Reprice Periods Integer Number of periods in reprice cycle. From account characteristics.

Reprice Rate Double New rate. Uses account pricing driver.

Reprice Start Integer Number of periods before first reprice, as an alternative to Reprice Start Date.

Reprice Start Date Date First reprice date. Reprice Start Period Type Integer Length of a Reprice Start period Monthly Reprices Yes/No Defines a repricing event

Once the System Field designator has been set, the user then needs to insert the appropriate name in the NAME field. The name will be used during the account mapping stage. It is useful for the user to select a name that helps to clearly identify fields without being overly complex or cumbersome. Examples of field description names can be seen in Appendix D.

Period Lengths The following period lengths are available with importing data. The period data size determines the

physical size of each period of data. 1 = Day 6 = Quarter 2 = Week 7 = Half Year 3 = Bi Week 8 = Year 4 = Month 9 = At Maturity 5 = Bi Month

The final stage in creating/updating a new rule is to set Transformation or Consolidate rules if necessary.

4.b. Transformation Rules

To enable the user to automatically manipulate their raw data on a global level, Transformation Rules can be set for each field. Clicking on the Transformation Rule Edit button next to the field you wish to transform causes the Transformation window to open.


Stage 1 1. Date Format String: All date fields require a date format to be set. All date fields shown in

the tables on the previous pages require a Transformation Rule. Though called a Transformation Rule, this setting allows the user to define the layout of the maturity date, as found in the raw data. No transforming is done here.

Formatting Dates in Load Recent History These formats are case sensitive: / - Date separator dd – Day as two numbers (for example, 01 as the first day of the month) d – Day as either one or two numbers (for example, January 1st as 1, November 11th as 11) MM – Month as two numbers (for example, 12 as December) M – Month as one or two numbers MMM – Month as abbreviation (for example, Nov) yy – two digit year (for example, 03) yyyy – four digit year (for example, 2003) Depending on what is found in the client file you would mix and match from the above possibilities. Remember the formatting is case sensitive. MM is not mm. Examples: 01/09/2007 MM/dd/yyyy 1/9/2007 M/d/yyyy 010907 MMddyy 15MAY2007 ddMMMyyyy 15052007 ddMMyyyy 150207 ddMMyy 15/MAY/2007 dd/MMM/yyyy 15-05-2007 dd-MM-yyyy 15-5-2007 d-M-yyyy

Two-digit Century is Misinterpreted When you have a date field with a two-character year, the century that year gets translated into is interpreted based on settings for the particular computer that the import is running on. By default, Windows XP computers set the threshold to 30. This means that a year of 20 is correctly interpreted as the year 2020, but the number 30 is interpreted as 1930. In the current release, this can be corrected by the following control panel settings (note – these instructions are for most versions of Windows XP. In Windows 2000 and possibly other XP versions, the location of the culture settings can wander): - From Control Panel, choose Regional and Language Options. - Next to Standards and formats language selection, choose the Customize button. - Choose the Date tab from the top-level tabs. - Note the setting for “When a two-digit year is entered…” - Suggested update is to set this to 2099, which will make the interpretation 2000-2099 If you update this setting, any import files affected will have to be re-read as well as re-running Calculate Values from the pro-forma screen.


2. Minimum /Maximum/Set Date To As Of Date: These settings allow the user to adjust dates earlier than the As of Date of the model or shorten dates that might be beyond 30 to 40 years in the future.

3. Set Value To: Sets a numeric field to the specified value.

4. Use lookup table: Lookup tables allow users to transform in-house codes into recognizable

data (i.e. rate indexes) or they can be used to substitute data. With the Lookup Table selected in the Edit Transformation Rule screen raw data codes are substituted with items recognized by the model (i.e. market instruments).

5. Simple Replace: Can be used instead of a Lookup Table for simple replacements of raw data

(8.50%, replace % with ). 6. Insert Location: Inserts the specified string into a string. 7. Absolute Value: Adjusts the sign of the field value to positive if it is negative.

8. Scale values by: Used to adjust rates, balances and other numbers during the download

process. A typical example scales interest rates: the download file has the format of rates as .0625 so the example to the right converts this rate to 6.25.

9. Imply Decimals: Used to adjust rates, balances and other numbers during the download

process. If decimals are not included in the download file and the balance is $1,568.95 but shown as 156895 the user needs to imply two decimals.

10. Amount has COBAL Trailing Sign: Informs the model the ‘+’ or ‘-‘ sign is after the

number not before

11. Pad to length. This works with other transformations like Date Format. For example, if the month is a single digit, many processors simply drop the preceding zero. So from January to September the date looks like Mddyy and October thru December looks like MMddyy, since the month is now two digits. Since the longest is six characters, we can pad the shorter layout to look like the longer layout by adding a preceding 0 to the left. Check Pad to length and select 6, check on the left, With 0. Then check Trim spaces and left only.

12. ITI Last Character Substitute: Used to replace a letter in the cent digit of certain down load

files. 13. Trim Spaces: This can work with other transformations like the Date Format or as a stand

alone transformation. With a fixed position file it can trim or delete erroneous characters from a field.

14. Truncate to length: Truncates a string value at the specified location.

Stage 2

15. Use Expression (Stage 2): Expressions can be used on all accounts from the file description or individual accounts on the row match rules. See page 16 for examples that can also be used on the File Description ‘s.

16. Reverse Asset or Liability Sign (Stage 2 All Accounts): Many times the trial balance is

designed to ‘sum to zero’ . This normally means that liabilities and capital are set as


negative. Checking the correct box on the Edit Transformation Rule screen overcomes this processor setting.

17. Reverse Income or Expense Sign (Stage 2 All Accounts): Many times the trail balance is

designed to ‘sum to zero’. This normally means that expense accounts are set as negative. Checking the correct box on the Edit Transformation Rule screen overcomes this processor setting.

4.c. Consolidation Rule FORESIGHT allows users to import all of their cash flow instruments individually. This is an incredible amount of data and it allows the user to create nearly perfect cash flows. But, such a large quantity of data forces the model to compute an enormous amount of calculations – slowing forecasts and market valuation calculations. To enhance calculation speeds, the user can consolidate their balances1. Consolidation should be common sense combinations of cash flows. As a general rule the user will consolidate all fixed rate instruments and keep the individual cash flows on all variable rate instruments. Specific field settings for consolidation are: Balance: Type will be Sum. Rate: Type will be Weighted Average and Column will be Balance. Maturity Date: Type will a meaningful selection from the drop down box (Min, Mid Month, Max).

Preview: File Source vs. Down Load File Once the user has completed setting the details for each File Source to be imported, they can test their rules with the Preview Button. The data preview screen allows the user to verify that they have chosen the correct fields and that the data has the layout that they expected. For example, the user would want to verify that numbers that ‘look’ like interest rates appear in the rate column. One would also want to see numbers that ‘looked’ like a date to appear in the maturity date column. To see a Preview of their download, the user should set the number of lines they would like to view and click the Preview button. The File Preview window will open. It is not necessary to preview more than a few lines. The purpose of the preview is to quickly verify that you have separated items correctly. For example, the user will quickly see if they broke their balances up incorrectly, or if a part of the maturity date were missing. Example 1 is a typical CD/Savings example. Each field row in the description has a corresponding column in the preview. Example 2 is a typical loan example. Notice from our previous discussions that loans require more information to identify their unique cash flows. This can be seen by the larger number of field rules in the description. When looking at the Preview, we can see each Field Rule in example 2 also has a corresponding column in the Preview screen.

1 A recorded consolidation step needs to be done during the mapping process.


Example 1

Example 2


4.d. Reading the Raw Data With the file descriptions reviewed for accuracy the user is ready to read the files into the model. Click on the Read Import Files link in the upper left of the Edit File Descriptions page and the user will be taken to the Load Recent History screen as seen to the right. From this page the user will import balances into the model. Instead of going directly from the data file to the account, the model will first read the files and save the appropriate ALM data that was selected in the file description. Reading import files needs to be done only once a period. In this step FORESIGHT will load the selected information from the RAW data file into the database. As the user loads data to the accounts, they are actually loading from the database not the raw data files. The first step to reading files is to select the correct data as of date. Second, the user should select all of the files they would like to read prior to populating the accounts. Third, the user clicks on the Read Import Files link. The last step will be populating your accounts. Click on Next and the user will be taken to the Load Recent History page. On the Load Recent History page the user can Calculate Values or edit the mapping rules. It is important to differentiate between Building/Modifying the mapping rules and simply Calculating Values each period. Building/ Modifying would require the user to change the mapping rules for accounts by clicking on the Edit button next to the account or by selecting the View All Row Match Rules option. Once the model has been setup, the user will not spend a lot of time adjusting the row match rules. The monthly process will be to select the Calculate Values option to populate the accounts using the row match rules that have been written.


4.e. Mapping your Accounts

When the model is first built, the user/builder must set specific rules regarding how each account will be populated. These rules can be seen by clicking on the Edit button found directly to the right of each chart name or by selecting the View All Row Match Rules option. From period to period a user will not necessarily need to change the existing rules. Therefore, on a monthly basis, a user will populate their accounts by simply clicking on the Calculate Values link on the left side of the Pro Forma Values screen.

To begin mapping an account, click on the edit button next to its name. The Edit Row Match Rule window will open. To change or modify the rules, the user must first have a clear understanding of how the cross reference codes, in this unique file, identify the specific account.

Row Match Rules determine which rows from a data source will get applied to a particular account. The rule also indicates whether rows are consolidated or not. The actual rule is an expression that describes what data source field values are used to determine the rows that belong with an account. See Appendix E for additional samples of Row Match Rules. Account: The accounts on the Pro Forma Values page match the accounts on the Chart of Accounts. File: Allows the user to choose which data source will be providing data rows for this account. After you choose a file, the Fields list will display all the known columns in that data source. Rule: An expression written using the Fields and Operations. The data type of the column must match the data type of the value it is being compared to. Amounts and numbers are specified without any special characters. Character values, or strings, must be enclosed in single quotes (as in BranchID =’M’). Fields: The user defined name given to each system and non system fields in the file source. Operations: Common comparisons use typical algebraic comparison operators like =,<,>,<=,>=. IN – the field is compared to a list of values in brackets. As in Type IN(2,5,9). LIKE – the field is compared to a pattern of characters where a wildcard can be specified.

Type LIKE ‘B%’ will match any type that begins with the letter B. BETWEEN – the field will match if it is in the set of numbers specified by the start and end

range. As in Term BETWEEN 7 and 12. CONTAINS – the field will match if it includes the phrase specified. The syntax is slightly

different from other operators. CONTAINS (ClassCode ‘K’) will match any class code that has the letter K in it.

AND/OR – can be mixed with the AND/OR keywords to create compound conditions.


Once all of the Row Match Rules have been set the user can populate the balances in the Pro Forma Values page by clicking on Calculate Values. At this time, if the balance sheet does not ‘tie out’ the user will need to go back to the Row Matching Rules for those accounts out of balance and adjust their rules. Once the rules have been adjusted, the user will re-Calculate Values. This is an iterative process until the balances load correctly.

Finally, the user’s rules may contain duplicated or unused items. The duplicated items cause the model to load the same instrument to more than one FORESIGHT account while unused items are not accounted for in any account. That means that after loading, the TOTAL balances will be too high or too low and individual accounts will not reconcile. To see if one’s rules may contain duplicates or unused items: after the file has been read and calculated click on the ‘View File Usage Statistics’ and review/correct any duplicated or unused items from files. To see the detailed list of instruments contained in an account, the user can select the VIEW under the data or detail column. The Cash Flow Detail screen opens and it displays the detail of all the loaded instruments related to that account.

In addition to mapping the combination of cross-reference codes into a mapping rule, the user may need to set consolidation and transformation rules at the account level.

• Consolidation Rules (Individual account selection)

Consolidation Rules are used to consolidate downloaded data for fixed rate accounts or accounts with no cash flows, where the consolidating of individual cash flows does not alter the integrity of the data. This example consolidates REM – 20 Yr Fixed loans by month using the maturity date of the individual cash flows (see Detail Processing section). Non Maturity Deposits should be combined into one row.

• Transformation Rules (Individual account verses global): See end of chapter for examples.

Using Expressions in Transformation Rules When creating row match rules, transformations can be applied to any particular column, and that transformation is only applied to the particular account the row match rule is associated with. For example, the sign can be reversed on a balance just for this account. These account-level transformation rules are accessed from the Edit Row Match Rule screen:

By clicking on any of the Edit buttons next to the Fields in the Column Transformation Rules grid, you get access to the rules for that column: Any of the supported transformations can be applied at this account level, but only a few make sense to use here; the reverse sign rules and the Use Expression rule. The other rules are typically applied at the File source level -- they are setup in the Edit File Descriptions screen which means they apply to all rows in the source file regardless of what account the row gets assigned to.


Use Expression Rule (these can also be used on the file description, Stage 2, covering all accounts) This rule is designed to allow custom calculations and data lookups to be applied when determining what the final value of the field in question should be.

Example; a file has records where loan participations are included in the book value reported. A separate field indicates what percent of the book value is participated out, and therefore the true book value of the instrument is really the stated book minus the participations. Let’s assume such a file has, among other values, columns called BookValue and ParticipationPercent. We also assume that both these columns are type Amount/Rate and that ParticipationPercent is a whole decimal (e.g. 45.0 is 45%). In this case, we need a transformation rule for the BookValue column to use a calculation – and for that we setup an expression transformation.

Expressions are based on the VBA language, and also contain specific functions provided by SAM. The value we would type into the Use Expression in this case would be:

Eutil.ColumnValue(“BookValue”) * (1 – (Eutil.ColumnValue(“ParticipationPercent”)/100)

Here is how the syntax of this expression breaks down:

Eutil. – this is a reserved word which must be put in front of some built-in functions ColumnValue – this is a function that returns the value of the column name put in parenthesis and quotes. This function is used to “obtain” the value of another field in the same record. The column named must be a numeric column (amount/rate or integer). * / - + – operators that can be used anywhere 100 – any fixed numeric value can be used in an expression. Available Values Eutil.ColumnValue(column name) – returns a numeric column value. Eutil.ColumnValueDate(column name) – returns a date column value. Eutil.ColumnValueString(column name) – return a string column value. Eutil.InstitutionValueDate(“History Asof Date”) – returns one of the specified values. Eutil.AccountNextReprice(column name) – returns the next logical reprice date for this account based on the maturity date. Column name must specify the name of the column that contains the maturity date – and it must be a date type field. This looks at the reprice cycle for the account, assumes the maturity date would specify the month/day of the end of a cycle, then calculates the next logical cycle that would occur after the current date. It handles first reprice period by assuming using term from account characteristics to calculate an origination date. Eutil.AccountValueDate(“NextRepriceDate”) – uses the account’s characteristics to calculate the next reprice date given that the data asof date would be the origination date. Eutil.MaturityFromPayment(balance column name, payment column name, rate column name) – Calculates a maturity date from the balance, payment and rate columns specified, and using the payment frequency of the account. The three columns specified must be numeric (Amt/Rate) columns. Invalid values (such as a negative payment, or invalid rate) result in a date returned equal to the first day of the first forecast period. Eutil.MaturityFromPayment(balance column name, payment column name, rate column name, pay frequency column name) – Calculates a maturity date, only in this case the payment frequency is specific by a column of data in the import. The payment frequency column must be an integer column that contains a value equal to one of the Foresight period type values:


1=Days, 2=Weeks, 3=Biweeks, 4=Months, 5=BiMonths, 6=Quarters, 7=HalfYears, 8=Years, 9=AtMaturity

If an invalid payment frequency is specified, it defaults to monthly. Note that if a frequency type in a file does not match the Foresight number scheme, a lookup transformation rule can be used to transform the incoming value to the appropriate integer value (e.g. a character M can be converted to the number 4).

In all the statements listed above, case is not important, but spelling and spaces are. LookupDbValue Function

This function can be used to perform database lookups in order to calculate what a field value should be. For example, let’s assume that participation loans are in separate records and therefore we cannot use the functions in the example above to determine what the true book value is. In this example we assume the participation values are in records that can be identified by a specific type code, BUT, we still need to subtract them from the current record’s book value. For that we would use the following expression for the BookValue column:

Eutil.ColumnValue(“BookValue”) - Eutil.LookupDbValue(“Sum(BookValue)”, “ClassCode=’A9’”)

LookupDbValue takes two parameters, the first is the column from the data that we want to access and the second is the criteria for record selection. Note that the column value can use SQL statistical operations such as SUM, AVG, MAX, etc. The criteria value must be expressed exactly as any row match rule is using the same syntax.

Other build-in VBA (Visual Basic for Applications) functions can be used – those will be documented more completely in a later version of this documentation.

Date Examples

1. Determine next logical reprice date based on maturity date:

Eutil.AccountNextReprice("MaturityDate")

2. Use a maturity date and current history date to determine a reprice date based on maturity month and day, and calculating current year:

IIf(Eutil.ColumnValueDate("MaturityDate").Month <= Eutil.InstitutionValueDate("History Asof Date").Month, DateValue(CStr(Eutil.ColumnValueDate("MaturityDate").Month) + "/" + CStr(Eutil.ColumnValueDate("MaturityDate").Day) + "/" + CStr(Eutil.InstitutionValueDate("History Asof Date").Year+1)), DateValue(CStr(Eutil.ColumnValueDate("MaturityDate").Month) + "/" + CStr(Eutil.ColumnValueDate("MaturityDate").Day) + "/" + CStr(Eutil.InstitutionValueDate("History Asof Date").Year))) Current Restrictions:

- LookupDbValue can only be used on records in the current file – the file that the row match rule belongs to.


3. Quarterly Repricing on January 1, April 1, July 1 and October 1 when next reprice date is not in the download file. This is a stage 2 transformation on the individual account level.

iif((eutil.asofdateplus1day.month mod 3) = 1,DateAdd("m",3, Eutil.asofdateplus1day), iif((eutil.asofdateplus1day.month mod 3) = 2,DateAdd("m",2, Eutil.asofdateplus1day),DateAdd("m",1, Eutil.asofdateplus1day)))

4.f. Importing Cash Flow Balances

Importing cash flow balances into FORESIGHT is a two step process. Read Files and Calculate Values. Read Files, as outlined above, imports the selected raw data into the model, but nothing is done with the data. Calculate values actually assigns the cash flow balances to the proper accounts.

4.g. Importing Historical Values Importing historical Values (NIB Balance Sheet and Income Statement balances) into FORESIGHT is again a two step process. Once a file description is created and properly mapped, the Read Files and Calculate Values functions should be completed. Read Files, as outlined above, imports the selected raw data into the model. Calculate Values actually assigns the ‘Read’ balances to the proper accounts.

4.h. Importing History Data Importing SAM4.1 history data balances into FORESIGHT is again a two step process that needs to be done only once. Once a file description is created and properly mapped, the Read Files and Calculate Values functions were completed by Farin Support. Read Files, as outlined above, imported the SAM4.1 data into the model. The history balances were imported if the institution had monthly history and the balance sheet was in balance.

4.i. Importing CMO Prepayments and Market Values

Importing CMO window information and over ride market values into FORESIGHT is again a two step process. Once a file description is created and properly mapped, the Read Files and Calculate Values functions should be completed. Read Files, as outlined above, imports the raw data into the model. Calculate Values actually assigns the ‘Read’ balances to the proper accounts. These must be calculated after the investment file cash flows have been calculated. A sample file description for importing CMO prepayments is included in Appendix – Download Description Examples. Market Values for previously imported Financial Instruments can be imported. The steps follow: 1) In "Manage File Descriptions" build a file source of the type "Import Market Values". This would include the ID and Market Value System Fields you wish to import from -400bp to +400bp. 2) In "Read Import Files", read the file in for the correct month. 3) In "Load Recent History", Calculate Values. Remember the import match’s on the ID (or CUISP). A report will be generated showing the duplicate CUSIP’s in the import file (if any) and the CUSIP’s from the import file that were not found in Foresight for that month.


4.j. Manually entering your balances

Users can manually input cash flows, balances, income/expenses, history or memo data into the model. When possible, the user should refrain from hand entering data into the model. It is much more efficient to import the data from a file. The Foresight database stores cash flow instruments for each account which represent the instruments and their characteristics that make up the book balance of a balance sheet account. In some cases, such as with non-interest bearing deposits, there is usually no need for more than one statement. In other cases, such as with adjustable rate mortgages, it is very useful to have separate statements for similar accounts, or possibly a separate statement for each individual loan where each statement contains the particular rate and repricing information. The cash flows tab on the manage chart of accounts screen provides a view of the instruments for each account, and also allows editing, copying and other maintenance functions. Most of the decisions about how to create and consolidate cash flow statements are made when setting up the automatic download of historical data from sub systems. The cash flows tab then allows a way to verify and edit these statements - although it is always possible to manually create all of the cash flow statements for any account. Editing Statements To make changes to an existing statement, just start typing in any of the existing cells. The cash flows grid functions very much like a Microsoft Excel worksheet. Changes are saved when you move to a different account, select a different period to view, or leave the chart of accounts screen. If you have the "prompt me to confirm saving changes" option selected, you will be warned that you made changes that may need to be saved. If this option is not selected any changes are saved automatically. With any changes to cash flow statements, Foresight re-calculates the period's book balance, average balance, and the period's weighted ending rate for interest bearing accounts. Inserting New Statements To insert a new row into the cash flows list grid, just click the Insert Blank Row or Insert Default Row (recommended) button in the bottom left of the grid. A new, empty row will be added at the end of the list. Deleting Statements To delete a single row, click on the row number, which is the first column on the left. Click the Delete Row button and answer the confirmation message. To delete multiple rows, you must select a group of rows by clicking and dragging down in any one of the columns (the amount column is easiest). Then click the Delete Row button. You will be prompted with a confirmation message indicating which row numbers it is about to delete. Copying Statements From a Prior Period If you do not import cash flow instruments into certain accounts or if you have complex cash flow statements that you would like to preserve as you add historical data to the model, you can use the copy button to copy statements into the current period from any other historical period. To use the copy function, first make sure you have selected the correct historical period in the As Of date field in the upper left. (Foresight will always show the most recent historical data in the cash flows grid when you first select an account or move to a different account). Click the Copy from... button at the bottom of the grid. You will be presented with a list of historical data periods you can copy from - with the prior month pre-selected. When you click Copy, the cash flow instruments from that period will be added to the current period's list. If instruments already exist for the current period, the copied instruments will be added below them. After copying, you can edit, insert, or delete as you normally would to update.


Sorting Statements It is often useful to sort the existing list of cash flow instruments by maturity date, amount, or some other value that lets you quickly determine if a particular value or statement is present. To sort, simply click on one of the column titles. Clicking on the Amount title in the second column will display the instruments in ascending order by amount.

1. NIB Asset or Liabilities: If a user does not load general ledger balances from a download file, but instead has elected to hand enter the balances, this is done at the Historical Data Tab under the Manage Chart of Account section. The user will need to enter the required information for the appropriate month.

2. Statistical Accounts: If a user does not load statistical balances from a download file, but instead

has elected to hand enter the balances, this is done at the Historical Data Tab under the Manage Chart of Account section. The user will need to enter the required information for the appropriate month.

3. Income or Expense: If a user does not load income statement balances from a download file, but

instead has elected to hand enter the balances, this is done at the Historical Data Tab under the Manage Chart of Account section. The user will need to enter the required information for the appropriate month.

4. Immediate Interest Bearing

Account: If a user does not load certain balance sheet balances from a download file, but instead has elected to hand enter the balances (ie FFS, FFP), this is done at the Historical Data Tab under the Manage Chart of Account section. The user will need to enter the required information for the appropriate month. At a minimum the two balance amounts and the two rate amounts should be entered.

5. Term Interest Bearing Account: If a user does not load certain cash flow balances from a download file, but instead has elected to hand enter the balances (ie Investment account), this is done at the Cash Flows tab under the Manage Chart of Account section. The user will need to insert a row and enter the appropriate cash flow information for the month. You can enter cash flow information relating to the type of instrument (i.e. bullet, amortizing, balloon), call data, reprice data and CMO detail.


6. WARM Balances WARM, or weighted average remaining maturity balances represent cash flows that combine multiple instruments into a single amount and weighted rate, with the maturity expressed as a remaining average maturity in months. WARM balances are typically found in some regulatory reporting systems, such as OTS CMR reports, and are also produced by some processing systems in lieu of representing individual instruments. These balances can be imported or manually entered into cash flow statements, and then get processed by the forecast calculation engine as another type of amortizing balance. If WARM balances are going to be entered by hand, you enter the cash flow information as though it were a maturing or amortizing instrument. Then in addition, simply enter the WARM in periods in the WARM Periods column of the cash flow statement. If the system sees a WARM Periods value other than zero, it will use it to amortize rather than the term of the instrument (which is typically expressed as either the Mature On or Amortize Until date). Note that the system will not allow you to enter a WARM Periods value for a non-maturing account. The account must be interest bearing, and its characteristics must be set to either an amortizing or maturing (bullet) instrument. To import a WARM balance, import the amounts and rates as usual, then assign the field that represents the WARM period number to the cash flow field type called WARM Periods. Amortization Method Used When the system encounters a cash flow statement with a WARM Periods number, it uses the following algorithm to calculate the runoff amounts for each period: Low Mid Point = WARM / 2 High Mid Point = (WARM + Maturity Period) / 2 High Percent = (WARM - Low Mid Point) / (High Mid Point - Low Mid Point) Low Percent = 1.0 - High Percent High Percent = High Percent / (Maturity - WARM) Low Percent = Low Percent / WARM From the first period to the WARM Period: Monthly Maturity = Amount * Low Percent From the WARM Period to the Maturity Period Monthly Maturity = Amount * High Percent Then the system will construct traditional Amortize statements using the computed Monthly Maturity. These constructed on-the-fly Amortize statements will include any prepayments, premium & discount (prorated) or repricing that was included. The following example illustrates this: Amount = 100, WARM = 18, Maturity = 24 So Low Mid Point = 18 / 2 = 9 High Mid Point = (18 + 24) / 2 = 21 High Percent = 75% / 6 = 12.5% Low Percent = 25% / 18 = 1.39% Therefore


Amortize 100 * 1.39% over 1 months Amortize 100 * 1.39% over 2 months Amortize 100 * 1.39% over 3 months . . . Amortize 100 * 1.39% over 18 months Amortize 100 * 12.5% over 19 months Amortize 100 * 12.5% over 20 months . . . Amortize 100 * 12.5% over 24 months The constructed distribution (before prepayments) will return the initial WARM number.


Appendix - Download Descriptions


Appendix - Expression Transformations When importing historical data into Foresight, it is sometimes necessary to convert or transform data from one form into another, more recognizable by Foresight. This expression rule is designed to allow custom calculations and data lookups to be applied when determining what the final value of the field in question should be. For example, it is sometimes necessary to calculate book value of an account after taking into consideration participations with other institutions. In another case you sometimes have to calculate a value - say a maturity date - when that value is not found in the import file. Using expressions in transformation rules can be applied to any particular column on the file description level or the account level. Examples Calculate the correct balance with participations. Example description: a file has records where loan participations are included in the balance reported. A separate field indicates what percent of the balance is participated out, and therefore the true balance of the instrument is really the stated balance minus the participations. Let's assume such a file has, among other values, columns call Balance and PercentSold. We also assume that both these columns are type Amount/Rate and that PercentSold is in decimal form (i.e. .45 is 45%). In this case, we need a transformation rule for the Balance column to use a calculation - and for that we setup an expression transformation. Eutil.ColumnValue ("Balance") * (1-(Eutil.ColumnValue("PercentSold")) Create a Next Reprice Date because it is not in the download file. Eutil.AccountNextReprice("Maturity Date") Create a Next Call Date because it is not in the download file. Eutil.AccountNextCall("Maturity Date") Create a maturity date from an origination date and a number of periods. It is assumed that "Term" is in months. DateAdd("m", Eutil.ColumnValue("Term"), Eutil.ColumnValueDate("Origination Date")) Turn off Balloons if the maturity date is equal to the balloon date. It is assumed that this expression is in a column with the system field "Balloons" IIf (Eutil.ColumnValueDate("MaturityDate") = Eutil.ColumnValueDate("BalloonDate"), False, True) When Balloon Amortize Date or Periods is not available. Create a dummy field using the System Field 'AmortizeDate' and add the the expression on the individual account in the Load Recent History screen. DateAdd("m",300, Eutil.columnValueDate("BalloonDate")) Where 'm' means months, 300 (number of months between AmortizeDate and BalloonDate (360-60), 'BalloonDate' is the reference point for adding months. Calculate a maturity date from a balance, payment and rate. Eutil.MaturityFromPayment("Balance", "Payment", "IntRate")


For Delay Amortization loans: Calculate when principal payments start. Example is a 180 month amortizing loan with a 6 month delay, characteristics screen is set as Amortizes in 186 months and Delay 6 months and Origination Date is in download. 1. Create a new file description and add two dummy fields: Call the first Delay Periods, set as an integer, system field is Next Payment Periods and call the second Delay Period Type, set as an integer, system field is Next Payment Period Type. 2. Add a transform rule to Next Payment Period Type, ‘Set Value To’, set it to ‘4’ (monthly). 3. Add the expression to the field ‘Delay Periods’ on each individual account on the Load Recent History screen (where 6 is the monthly delay): IIF(6 - DateDiff("m", EUtil.ColumnValueDate("OriginationDate"), EUtil.AsOfDatePlus1Day)>=0,6 - DateDiff("m", EUtil.ColumnValueDate("OriginationDate"), EUtil.AsOfDatePlus1Day),0) Limit a date to a maximum of 15 years from the import date (on an individual account level and not on the file description level). It is assumed that this expression is in a column with the system field "Maturity Date" IIf(EUtil.ColumnValueDate("MaturityDate") > EUtil.AsOfDate.AddYears(15), EUtil.AsOfDate.AddYears(15), EUtil.ColumnValueDate("MaturityDate")) Transform an offset rate to be a negative when separate field code specifies a negative versus positive offset. IIF (Eutil.ColumnValueString("RateOffsetPointer") = "L", Eutil.ColumnValue("Offset") * -1, Eutil.ColumnValue("Offset")) ---------- A note on IIF: IIF has 3 parameters, the TEST, the TRUE branch and the FALSE branch. IIF(TEST, TRUE BRANCH, FALSE BRANCH) Quarterly Repricing on January 1, April 1, July 1 and October 1 when next reprice date is not in the download file. iif((eutil.asofdateplus1day.month mod 3) = 1,DateAdd("m",3, Eutil.asofdateplus1day), iif((eutil.asofdateplus1day.month mod 3) = 2,DateAdd("m",2, Eutil.asofdateplus1day),DateAdd("m",1, Eutil.asofdateplus1day)))


Appendix - Row Matching Rules Row match rules determine which rows from a data source will get applied to a particular account. The rule also indicates how rows are consolidated, if at all, to save space and calculation time. The actual rule is an expression that describes what data source field values are used to determine the rows that belong with this account. A simple rule could be something like:

ClassCode = 21, which says all rows with a value of 21 in the ClassCode column will match this account.

Rules can be as complex as necessary, as in this example that matches 2 conditions; a ClassCode that matches one of the 3 specified values and a Term between 6 and 12:

ClassCode IN ('B1','B2','B7') AND Term BETWEEN 6 and 12

File:

Allows you to choose which data source will be providing data rows for this account. After you choose a file, the Fields list will display all the known columns in that data source.

Rule:

The rule is a sentence that describes the conditions that create a match. The rule must refer to at least one field and at least one value. Rules follow SQL search condition syntax rules, the most common of which are summarized below:

Data types: The data type of the column must match the data type of the value it is being compared to. Amounts and numbers are specified without any special characters. Character values, or strings, must be enclosed in single quotes (as in BranchID = 'M').

Comparsion Operators: Common comparisions use typical algebraic comparison operators like =, <, >, <=, >= (Term > 6).

Special Operators: The SQL syntax includes some special operators to perform more complex comparisions: IN - the field is compared to a list of values in brackets (ClassCode IN ['B1','B7']) LIKE - the field is compared to a pattern of characters where a wildcard can be specified. ClassCode LIKE 'B%' will match any ClassCode that begins with the letter B. BETWEEN - the field will match if it is in the set of numbers specified by the start and end range (Term BETWEEN 6 and 12) CONTAINS - the field will match if it includes the phrase specified. The syntax is slightly different form other operators. CONTAINS(ClassCode,'K') will match any ClassCode that has the letter K in it.

And/Or: Conditions can be mixed with the AND / OR keywords to create compound conditions.

Fields:

This is a list of the fields associated with the selected file, or data source. Since one or more of the fields will be used in a rule, this list is provided for convenience. Double click on any field name and that name will be added to the rule box.


Operations:

This is a list of common operations that can be used on fields and values when constructing rules. Double click on any operation to insert it into the rule box. Operations can also be typed in manually where needed.

Detail Processing:

This choice determines whether and how rows are consolidated, or added up, in order to reduce the total number of cash flow statements created for this particular account.

Keep all detail rows - In this case no consolidation is done, and one cash flow statement is constructed for this account for every row in the data source that matches this rule Summarize by month - All rows that have the same month/year in the specified Summary Field will be combined into one row. This is a common way to create cash flow statements with one statement per maturity month. When you use this method, you must specify a Summary Field and it must be a field from the data source with dates. Summarize by week - All rows that have a Summary Field date in the same week will be combined into one row. Again the Summary Field must specify a date column. Summarize by day - All rows that have the same Summary Field date will be combined into one row. Combine all rows into one - all matching rows are consolidated into one cash flow statement.

Reserved Words to Avoid in Row Match Rules When importing historical data in Foresight (calculating values), you may see an error like this: Incorrect syntax near the keyword 'xxxxx'. Statement(s) could not be prepared. It is likely you are using a reserved word in something like a column name or perhaps a table name. Reserved words are parts of the SQL database language that is used to process row match rules. If you are using one of these, you will need to change the name of a column in the file description of the file being processed at that time. Reserved Words List ABSOLUTE ACTION ADA ADD ADMIN AFTER AGGREGATE ALIAS ALL ALLOCATE ALTER AND ANY ARE ARRAY AS ASC ASSERTION AT AUTHORIZATION AVG BACKUP BEFORE BEGIN BETWEEN BINARY BIT BIT_LENGTH BLOB BOOLEAN BOTH BREADTH BREAK BROWSE BULK BY CALL CASCADE CASCADED CASE CAST CATALOG CHAR CHAR_LENGTH CHARACTER CHARACTER_LENGTH CHECK CHECKPOINT CLASS CLOB CLOSE CLUSTERED COALESCE COLLATE COLLATION COLUMN COMMIT COMPLETION COMPUTE CONNECT CONNECTION CONSTRAINT CONSTRAINTS CONSTRUCTOR CONTAINS CONTAINSTABLE CONTINUE CONVERT CORRESPONDING COUNT CREATE CROSS CUBE CURRENT CURRENT_DATE CURRENT_PATH CURRENT_ROLE CURRENT_TIME CURRENT_TIMESTAMP CURRENT_USER CURSOR CYCLE DATA DATABASE DATE DAY DBCC DEALLOCATE DEC DECIMAL DECLARE DEFAULT DEFERRABLE DEFERRED DELETE DENY DEPTH DEREF DESC DESCRIBE DESCRIPTOR DESTROY DESTRUCTOR DETERMINISTIC DIAGNOSTICS DICTIONARY DISCONNECT DISK DISTINCT DISTRIBUTED DOMAIN DOUBLE DROP DUMMY DUMP DYNAMIC EACH ELSE END END-EXEC EQUALS ERRLVL ESCAPE EVERY EXCEPT EXCEPTION


EXEC EXECUTE EXISTS EXIT EXTERNAL EXTRACT FALSE FETCH FILE FILLFACTOR FIRST FLOAT FOR FOREIGN FORTRAN FOUND FREE FREETEXT FREETEXTTABLE FROM FULL FUNCTION GENERAL GET GLOBAL GO GOTO GRANT GROUP GROUPING HAVING HOLDLOCK HOST HOUR IDENTITY IDENTITY_INSERT IDENTITYCOL IF IGNORE IMMEDIATE IN INCLUDE INDEX INDICATOR INITIALIZE INITIALLY INNER INOUT INPUT INSENSITIVE INSERT INT INTEGER INTERSECT INTERVAL INTO IS ISOLATION ITERATE JOIN KEY KILL LANGUAGE LARGE LAST LATERAL LEADING LEFT LESS LEVEL LIKE LIMIT LINENO LOAD LOCAL LOCALTIME LOCALTIMESTAMP LOCATOR LOWER MAP MATCH MAX MIN MINUTE MODIFIES MODIFY MODULE MONTH NAMES NATIONAL NATURAL NCHAR NCLOB NEW NEXT NO NOCHECK NONCLUSTERED NONE NOT NULL NULLIF NUMERIC OBJECT OCTET_LENGTH OF OFF OFFSETS OLD ON ONLY OPEN OPENDATASOURCE OPENQUERY OPENROWSET OPENXML OPERATION OPTION OR ORDER ORDINALITY OUT OUTER OUTPUT OVER OVERLAPS PAD PARAMETER PARAMETERS PARTIAL PASCAL PATH POSITION POSTFIX PRECISION PREFIX PREORDER PREPARE PRESERVE PRIMARY PRIOR PRIVILEGES PROCEDURE PUBLIC READ READS REAL RECURSIVE REF REFERENCES REFERENCING RELATIVE RESTRICT RESULT RETURNS REVOKE RIGHT ROLE ROLLBACK ROLLUP ROUTINE ROW ROWS SAVEPOINT SCHEMA SCOPE SCROLL SEARCH SECOND SECTION SELECT SEQUENCE SESSION SESSION_USER SET SETS SIZE SMALLINT SOME SPACE SPECIFIC SPECIFICTYPE SQL SQLCA SQLCODE SQLERROR SQLEXCEPTION SQLSTATE SQLWARNING STATE STATEMENT STATIC STATISTICS STRUCTURE SUBSTRING SUM SYSTEM_USER TABLE TEMPORARY TERMINATE TEXTSIZE THAN THEN TIME TIMESTAMP TIMEZONE_HOUR TIMEZONE_MINUTE TO TOP TRAILING TRAN TRANSACTION TRANSLATE TRANSLATION TREAT TRIGGER TRIM TRUE TRUNCATE TSEQUAL UNDER UNION UNIQUE UNKNOWN UNNEST UPDATE UPDATETEXT UPPER USAGE USE USER USING VALUE VALUES VARCHAR VARIABLE VARYING VIEW WAITFOR WHEN WHENEVER WHERE WHILE WITH WITHOUT WORK WRITE WRITETEXT YEAR ZONE


Chapter 5

FORESIGHT External Assumptions External market forces can have massive affects on an institution. In this chapter, we will outline the types of external assumptions included in the FORESIGHT product and the methods to keeping them up to date. There are three major external assumption sections; prepayment speeds, decay rates and market rates. All three sections can be automatically imported into the model via an internet connection. Users are able to insert user defined prepayment and decay tables. User may also create user defined market rate drivers in the interest rate section and have complete flexibility to forecast market rates into the future. The Update External Assumptions link from the All Tasks screen will take the user to Market Rates and this screen includes a link to Prepay/Decay Assumptions.

5.a. Prepayment Speeds Prepayments are early, unscheduled and non-contractual return of principal. FORESIGHT supports a number of prepayment methods when using a prepayment table. Prepayment methods determine how the ‘cell’ or ‘cells’ used to determine the prepayment amount is chosen. Inputs to this process are a coupon rate, a rate change and the time into the forecast that the prepayment takes place. Prepayment Speeds are set on the account Characteristics screen. Prepayment rate tables match the OTS Asset Liability Pricing Tables. The prepayment method to be used is selected on the “Update Institution Settings” screen.


5.b. Decay Rates Since Non Maturity Deposits do not have guaranteed cash flows (contractually immediate accounts) like CD’s or loans, institutions need to estimate when dollars will roll out. The OTS provides acceptable Decay Rates to use when estimating NMD roll off. Users can also calculate their own in house numbers or utilize external providers such as McGuire Solutions. Foresight uses the decay rate to estimate cash flows for the NMD’s. Decay Rates are set on the account Characteristics screen. Both Prepayment and Decay rates are updated at the same time by clicking on the “Load Most Recent Prepayment/Decay Tables” link. Create User Defined Tables: User’s can define their own prepay and decay tables through the “Create a New Prepayment/Decay Table” link from the Prepayment/Decay Tables screen. A dialogue box will pop up showing you the additional table options that must be filled out. To create a user defined Prepayment/Decay Table you will: 1. Table Name: Give the table an appropriate name that telegraphs the contents to users of the model. 2. Table Type: Select Prepayment Table, Decay Table, Credit Quality Table or Early Withdrawal Table. 3. Value Type: For Prepayment Table select from CPR (Conditional Prepay Rate, used most often), CPP (Constant Prepay Percent), or PSA. Credit Quality and Early Withdrawal Tables on have one choice for each. For Decay Table select from Annual (used most often) or Remaining. 4. Period Type: Determines how many columns exist in the table and which time periods the values for each column correspond to. For Prepayment Table the selection is Standard Prepay Buckets. For Decay Table select from OTS or McGuire. 5. Rate Environments: Select all rate environments you will use. 6. Click Continue 7. Use one prepay/decay rate for all interest rates or insert additional rows for prepay/decay rates by interest rates. Complete by inserting prepay/decay rates by interest rate, time period and rate environment. To make changes to your user defined tables click on the ‘Update Table’ box and make the appropriate changes. This update will retain the old rates for back testing. To delete a user-defined table, select a table and click the “Delete Selected Prepayment/Decay Table” link. To delete a row, select the row and click the “delete Rows” button. Multiple rows can be selected. You can only delete rows in a new table that has not yet been saved. If you are working with a waved table, it is necessary to delete and re-create the table in order to get specific rows removed. Not that you cannot delete rows or otherwise make changes to a table that is downloaded from the Farin web service. The new table is created with one row per rate environment. To add a new row(s) for different coupon rates, click the “Insert Rows” button and then enter the coupon rate for that row (the rate that represents the high


value of the range used during a table lookup). To edit row values, including the coupon rate, simply type values into the cells of the table.

Prepayment and Decay Calculations FORESIGHT supports prepayments and decay calculations via prepay and decay tables. The following details the algorithm:

Rate Environment Type

Shock: Prepayments: The financial instruments Coupon Rate is matched to the Coupon rows in the table. The age of the financial instrument is matched to the time period in the table. The amount of the shock is used to index into the Rate Change vector of the table. Decays: The financial instruments Coupon Rate is matched to the Coupon rows in the table. The amount of time from the beginning of the forecast is matched to the time period in the table. The amount of the shock is used to index into the Rate Change vector of the table.

Yield Curve Driven: Prepayments: The financial instruments Coupon Rate is matched to the Coupon rows in the table. The age of the financial instrument is matched to the time period in the table. Decays: The financial instruments Coupon Rate is matched to the Coupon rows in the table. The age of the financial instrument is matched to the time period in the table Note: For Decays, the above is only true for calculations on Gap or Market Value / Duration Reports. In Income Simulations, all Financial Instruments that Decay will be assumed to Mature in 1 Month. The Income Simulation doesn’t need to generate the extra detail as there is be no difference in Balance or Income between the two approaches.

Prepayment Methods FORESIGHT supports a number of prepayment methods when using a prepayment table. Prepayment methods determine how the ‘cell’ or ‘cells’ used to determine the prepayment amount is chosen. Inputs to this process are a coupon rate, a rate change and the time into the forecast that the prepayment takes place. The prepayment method to be used is selected on the “Update Institution Settings” screen.


Using this transposed prepayment table, examples of how the different prepayment methods work is shown below: OTS 30 Yr FRMs - FHA & VA Months 0-6 -400.00 -300.00 -200.00 -100.00 0.00 100.00 200.00 300.00 400.00 06.00 34.8609 32.7680 26.7701 15.3604 10.2126 08.3521 07.4538 06.9319 06.5923 06.50 35.1311 33.5900 29.8325 20.1958 12.1900 09.2806 08.0013 07.3024 06.8655 07.00 35.3325 34.1335 31.5809 24.9641 15.1519 10.5669 08.7025 07.7531 07.1866 07.50 35.4883 34.5177 32.6564 28.2978 19.1853 12.3970 09.6233 08.3111 07.5686 08.00 35.6124 34.8030 33.3712 30.3713 23.4618 14.9992 10.8627 09.0153 08.0293 08.50 35.7136 35.0229 33.8765 31.6912 26.8581 18.4416 12.5621 09.9221 08.5934 09.00 35.7976 35.1975 34.2511 32.5797 29.1701 22.2386 14.8827 11.1130 09.2956 09.50 35.8684 35.3393 34.5392 33.2105 30.7093 25.5517 17.8765 12.6969 10.1846 10.00 35.9290 35.4569 34.7673 33.6785 31.7666 28.0069 21.2479 14.7907 11.3272 The following data is used in the examples: Coupon Rate = 9.20%, Rate Change = 45.00 basis points and the prepayment takes place 1 month into the forecast. The first thing done is to gather the data for the correct time period. Prepayment tables are analogous to an Microsoft Excel workbook. A prepayment workbook is made up of worksheets. Each worksheet represents a single prepayment table time period. The worksheet for the OTS 30 Yr FRMs – FHA & VA prepayment table for the first 6 months into the future is shown above. The following methods are selected on the Update Institution Settings page (see chapter 2).

By Step – the cell chosen will be equal to the last row (moving from top to bottom) that the coupon rate is larger than. In this case 9.20%, the row used will be the row with the coupon rate of 9.00%. The column used will be the last column (moving from left to right) that the rate change is larger than. In this case 45.00, the column used will be the column with a rate change of 0.00. The value returned will be the exact (9.00, 0.00) cell value of 29.1701.

Nearest – the cell chosen will be determined by using the By Step method as a starting point. Then if [45.00 = rate change – By Step column rate change] is greater than half the difference of [100.00 = rate change in column to right By Step column rate change – By Step column rate change] the base cell will move to the right one rate change column. In this case 45.00 is less then 50.00, so it will not move. Then the coupon rate is measured against its By Step row coupon rate of 9.00% in a similar manner. In shorthand, [(9.20 – 9.00) / (9.50 – 9.00)] is not greater than 50% so again the base cell from the By Step calculation will not change. Linear – the value chosen will be linearly interpolated using the cell to the right of and the cell below the cell that would have been chosen in the By Step method. These are the values that surround the inputs. So: (9.00, 0.00) = 29.1701 from the By Step method is used as a base. (9.00,100.00) = 22.2386 is used for any Rate Change Adjustment. (9.50,0.00) = 30.7093 is used for any Coupon Rate Adjustment. = 29.1701 - ((45 / 100) * (29.1701 – 22.2386)) + ((0.20 / 0.50) * (30.7093 – 29.1701)) = 29.1701 – 3.1192 + 0.6157


So Base - Rate Change Adjustment + Coupon Rate Adjustment or 26.6666 is returned. 5.c. Market Rates The External Assumptions Rates screen has several components that use Market Drivers, Yield Curves and User Defined Instruments. 5.c.i. Historical Rates

These recent market interest rates are downloaded from Farin & Associates web service. The historical rates will be updated when you select the Update Historical Rates option. A Select Rates to Import screen allows the user to choose starting and ending date ranges to import.

5.c.ii. Forecast Rates

When the user selects this option, future interest rates will be displayed for the Rate Environment chosen. This option can be very useful visualizing how future rates are affected in different rate environments. See Environments and Behaviors sections below relating to how the model interprets rates changes.

5.c.iii. Adding Instruments

Individual Instrument To add an institutions user defined interest rate click on the Add Your Own Instrument link. On the Add Your Own Instrument screen select whether you are creating a New yield curve or New single point instrument. Enter an appropriate name (the Market Instrument list is in apha order) and Term in months. If creating a new instrument, most users will create a single point instrument similar to their NOW, MMDA and Savings instruments. When the user defined instrument is created the rate will need to be manually entered monthly (this includes both ending and average rates unless you check the “Upon entry of User Defined historical rates, set Average and Ending Rates to the same value”). Yield Curve To add a curve with multiple points (instruments) follow this order: Click the Add Your Own Instrument link; choose New Yield Curve button; enter a name for the yield curve; enter the name of the first point in this yield curve; enter the term in months of this first instrument; click continue and you will see your new yield curve/market instrument in the list of instruments. To add additional points to the yield curve follow this order: click on the cell that contains the yield curve name; click the Add Your Own Instrument link; this time select the option New Point to Existing Curve; fill in the


next point’s name and term in months; click continue. Continue this last step to complete your curve. Once completed you may enter your user defined interest rates.

5.c.iv. Environments

To view, modify or insert a rate environment the user will select the Edit Rate Environments on the Update External Assumptions (i.e. Rates screen). To view how the rate environment reacts, highlight your selection in the upper left and review the setup on the right. To modify an existing instrument the user might change the information on the target rows. An example of a change would be to adjust the Offset and/or Over Months selections. When the user selects Insert New Item they must first give the new rate environment an appropriate name. Next select whether it will be a “Shock”, “Yield Curve Driven” or “User Supplied”. If the new item is a Shock the user must enter the basis points in the Shock Offset box. This will be an immediate and parallel shock in the first month of the forecast. When the user selects Yield Curve Driven Foresight requires at least one row under Yield Curve Targets. Click on the Pick button to select a Yield Curve from the Market Rates screen (i.e. US Treasury). Enter the Offset in basis points and Over Months for the number of months the change should move over. The user may select the “Relative to Forecast Start” box so that the Yield curve continues to be the date of the model or uncheck the Relative and enter a Date that will not change when the model is rolled forward (i.e. Non Parallel). A Yield Curve Driven Rate Environment uses the base yield curve to generate all other market rates. The offset that existed between any derived instrument and the base curve as of the base curve date is used as a constant to the base yield curve to generate forecast market rates. When the user selects User Supplied Foresight requires at least one row under Yield Curve Targets. Click on the Pick button to select a Yield Curve from the Market Rates screen (i.e. US Treasury). Then click on the Edit button to select the forecast month(s) you wish to enter the selected curves rates. You may define all the points in the curve selected for the number of months you entered (i.e. Non Parallel). Samples of Rate Environments can be found in your model and in your training material. Always verify the set up of the environment by checking the forecasted rates in the Update External Assumptions screen. A user Supplied Rate Environment allows you to enter future rates in the selected yield curve by selected forecast periods. The offset that existed between any derived instrument and the base curve as of the base curve date is used as a constant to the base yield curve to generate forecast market rates the same as in Yield Curve Driven Rate Environment. For information on the Global Insight Rates please contact our Farin Sales staff.


Using the Global Insight Rate Forecasts in FARIN Foresight (fee based service) Each month the Global Insight rate forecasts are updated and available for importing into model. The GI file name includes the month of the last period of Actual data. For example the GI Base Jun-2007 rate environment contains forecast information starting in July.

Steps to import the information into Foresight: 1. From the All Tasks screen select Update External Assumptions. From this screen select the Edit Rate Environments options. 2. Select the Import Global Insight Rates link at the bottom of the screen. Remember you will need an active internet connection to import the rates. 3. From the available dates select the month you wish to import and click on Import.

In Foresight the Global Insight rates are imported as a Rate Set, meaning they are organized in a group with the Month/Year of the period as the name. Each set contains only the three rate forecasts for that period, Base, High and Low. If you wish, once the rates are loaded you can add more rate forecasts to that set using the Edit Rate Sets option. If you are using the Edit Rate Behaviors option on the External Assumptions page review for updates to those movements for the new rate forecasts you have added this period. You are now ready to use the GI rates in forecasting your income at risk. Information about the rates may be obtained by entering your authorization code at the website address http://globalinsight.farin.com McGuire Smart Ramps (fee based service) Foresight now supports the downloading of McGuire Smart Ramps with an additional import selection. This feature supplies Driver rate exposure ramp scenarios and Yield curve shape change exposure scenarios.

5.c.v. Edit Rate Behaviors


The user can go one step further with Rate Environments and change the behavior of a market instrument different from the chosen Yield Curve. With the rate environment selected on the Update External Assumptions page click on the Edit Rate Behaviors link in the lower right. With the Edit Rate Behaviors screen open select Insert New Item and pick the market instrument you would like to react differently from the curve. Next you will select one of three Method’s: Fixed Rates, Offset from Rate Driver or Offset from Ourselves. With the Fixed Rates method selected enter your forecasted rates in the Driver Rate row. With the Offset from Rate Driver method entered you will first browse and select a Driver Rate from the list of market instruments (.i.e. similar term). You will then be able to enter an offset from the rate driver or enter the rate used and the model will calculate the offset (similar to the pricing screen). The user can also use a lagging driver that will be user defined for the instrument inserted and the rate environment you are working in. The lagging drivers work similar to the pricing screen where you select the lag periods and then the percent change in each period.

The Offset from Ourselves works similar to Offset from Rate Driver except you only use the market instrument you are saying reacts differently. Always verify the set up of the rate behavior by checking the forecasted rates in the Update External Assumptions screen.

5.c.vi. Graphs

The user can chart external rate graphs from two separate locations, i.e. Edit Rate Environments screen or External Assumptions – Rates screen.


External Assumptions-Rates “View Yield Curve” selection: The View Yield Curves and Rates screen appears. Select the instrument curve on the left and the appropriate date(s) on the right and click Yield Curve to display the results in a graph. Edit Rate Environments “Graph” selection: After selecting your Forecast the Choose Yield Curve to Graph screen appears. Select the instrument curve on the left and the appropriate date on the right and click Done to display current rates and the Rate Environment rates in a graph.

Example: Compare Treasury Yield Curve from June 2004 with June 2003.

By changing items on the charts tool bar the user can change the looks of the graph.


Appendix - Non Parallel Rate Environment

You hear Interest Rate Risk (IRR) discussed more and more in our troubled economy. What is Interest Rate Risk? It is the potential for loss of future income or economic value of equity caused by an adverse change in interest rates. What is an adverse change in interest rates? That could be different for each financial institution. For you an increase in rates could reduce your future income and for another a decrease in rates could have a negative effect. As we look into the future, the mix of our current products and the option risks associated with them come into play. Whether you have longer term fixed rate products with substantial prepay option risk, adjustable rate instruments that can adjust to market in the future or a large portfolio of callable investments, that mix and timing can cause you to lose sleep at night. Should you change your mix? Should my marketing plan attempt to book fixed rate loans or adjustable? How do I fund my loan demand? Should I increase deposit rates or borrow the funds needed. There are many questions. What about answers. What about market rates? We know that economic stimulus generally comes from two government sources. The Federal Reserve can lower interest rates, as it has been doing since September 2007. This is meant to promote borrowing and spending. Also Congress can lower taxes or boost spending to provide an added lift to the economy. Of course we do not know if or when the Fed or Congress will act or how the action will affect our bottom line. So you can see that there seems to be more questions than answers when we try to forecast the future. This is normal since we are talking about the future, the unknown. But how can we make good decisions with the information we do have? Whether written or not, most of us have a net income limit that specifies the maximum our net income can drop in selected rate environments. Let’s be realistic, any reduction in net income is considered a negative. Now that we have some of the unknowns and worries on the table, let’s consider how Farin Foresight can help you sleep at night. The Foresight model is designed to help you make better decisions. Since you are forecasting the future, there are a number of things in the model you must be comfortable with. The four basic ones are:

1. You must have an accurate source of data dividing rate sensitive assets and liabilities entered into Foresight. As we start looking into the future, your largest source of cash flow information is your current position. The more detail you can extract from your download files and enter into the model the better current cash flows you have to start with.

2. The user must have reasonable prepayment assumptions entered by rate environment. The option risk of prepaying a loan is a large portion of future principal and interest cash flows. Feel comfortable about the reliability of the models prepay schedules or create your own.

3. The user must make reasonable assumptions about the future for each line item in their Chart of Accounts. These are your assumptions relating to what the balances will be in the future and what the rate will be on originations in interest bearing/paying accounts.

4. The user must develop a set of rate environments to use in testing sensitivity. This is not just an assumption or best guess on what market rates will be in the future but a test of ‘what if market rates do this’. What we need to test is how sensitive our net income is in selected rate environments.

There are many pieces to consider in managing our balance sheet to maximize profits while ensuring you

maintain the ability to meet the needs of loan and deposit customers. Volumes have been written on how to achieve consistent profits, liquidity and safety. However what I am going to do is to continue specifically with number four from above. This can be a sensitive part of modeling because unlike setting your budget in one rate environment, for Income at Risk (IAR) sensitivity testing you must run forecast reports in various market rate scenarios.

In the past the test rate environments normally consisted of Flat, a rate ramp increase and a rate ramp decrease. Remember you are trying to answer the ‘What if’ question. As an example: what if rates increased or decreased 100bps, 200bps or 300bps over 12 months or 36 months? How would that impact my bottom line in each case? If the result of the forecast is a decrease in net income; is the risk of that happening acceptable? In the past one short coming we as an industry had, was not testing a broad enough range of rate fluctuations.


Testing with rates increasing/decreasing 25 to 50bps is not sufficient. We always need to test in outside ranges or extreme changes to check the sensitivity of movements in net income in a wide range of market rate changes. I like the rate ramps for the majority of testing environments because it does take the guess work out of when the actual rates will change and by how much.

Where our discussion must lead is to another type of risk, which is Yield Curve Risk. Yield Curve Risk is a type of interest rate risk that is the potential for loss of earnings or economic value of equity; caused by a changing of the shape of yield curves. It does not take a lot of research to see the change in the Treasury Yield Curve over the last three years. In the summer of 2005 the shape of the treasury curve was still normal (sloped). Then the yield curve increased while it flatten out with significant increases in short term rates. In the summer of 2006 we also started to see an inverted yield curve as short term rates increased. Just recently we are starting to see some normalcy in the yield curve.

When you test your Income at Risk, it would be prudent to include a rate environment that changes the shape of the Treasury Yield Curve. How much income at risk do I have in that scenario? But how do I accomplish this in Foresight. It’s easy, let’s review how.

How to create a non parallel rate environment: 1. From the ‘All Tasks’ tab select ‘Update External Assumptions’, then select ‘Edit Rate Environments’

where you can see and review all the rate environments in the model. 2. Click on ‘Insert New Item’, enter a name and ‘Continue’; in my case I called it US Non Parallel: 3 yr

forecast. 3. Change the Type in the upper right to ‘User Supplied’, click OK and select the US Treasury Curve. 4. Next click on the ‘Edit X’ and insert periods in the future that you would like to specify the curve

rates. In the example I have selected to enter rates on the curve 12 months, 24 months and 36 months into the future.

5. Enter the rates in the future, on as many of the points as you can, the model will interpolate between rates entered both horizontal and vertical.


Congratulations, you have created a non parallel rate environment. If you need a professional projection consider the Global Insight rates that can be downloaded into the model. How to create a IAR Rate Set: Now that you have created the rate environment, you should create a rate set so this can be selected when running reports. On the Edit Rate Environments page or on the Data Source & Dates tab of the actual report select ‘Edit Rate Sets’. Select ‘Insert New Set’, enter a name and click on Done. The new set will appear in your list, now you need to drag and drop the rate environments from the right into the new set. See the sample created called Test IAR. Reports: Now you can test the sensitivity of your income at risk running your normal forecast reports in selected rate environments. A very helpful report to run is the Forecast Decision Matrix. This summary report can be run with one forecast selected or multiple forecasts for comparison analysis.


With just one Forecast selected the Decision Matrix shows me key financial information in selected rate environments. I can see that net income will decrease with market rates dropping and with the Non Parallel environment I created. Now the question that needs to be answered is how much risk am I willing to take if I believe rates will be increasing in the future. I might be putting myself in to much risk because even in a slight decrease of rates my income and ROA drops significantly. If I believe rates will be decreasing in the future I need to make some major changes to my strategic plan.

With two Forecasts selected the Decision Matrix can compare two different strategies in the selected rate environments. Here the income at risk is less and more balanced in the What If strategy.


Chapter 6

FORESIGHT Internal Assumptions Forecasting is the very center piece of ALM. Testing multiple future scenarios allows modelers to identify gaps and weaknesses by stress testing a model in various rate environments and growth scenarios. Once the user has modified their Chart of Accounts, adjusted the Download, loaded balances and verified cash flows by viewing the Current Balance Sheet and Current Gap report - they can start to project the institution out into the future. How will the institution grow in the future? How will the mix change? What rates are expected? The user answers these questions with Pricing and Targets. Without a process, setting future growth Targets and Pricing rules can be a daunting task. Armed with a coherent strategy, and a clear vision of the future, forecasting the institution into the future can be as easy as filling in the blanks. 6.a. Where are You Going? In Chapter 1 the user was introduced to a complete process that management should employ to identify future growth and pricing patterns. Though it is not the only process, we feel it is the most complete process. Answering the question, “What will my institution look like in three years?” is not an easy task, but failing to answer that question makes forecasting impossible. Refusing to identify how an individual instrument’s cash flows should react in the future is a sure course to failure. Many Support Line phone calls begin with the user suggesting, “The model is not working right.” Or “This instrument is not doing what it should in a forecast.” The first question the Support Staff asks is, “What should it do… IN REAL LIFE?” At this point there is often a pause. There is a pause because there is often a disconnect between real world events and modeling. Even when one does not clearly understand their own expectations of the future, they believe the model should produce the ‘right’ numbers. This is impossible. The user needs to ask themselves a string of questions to clarify their institution’s future growth pattern. For FORESIGHT to provide adequate forecasts, the user must know how each chart item (balance sheet and income statement) will grow and change in the future. Initial questions include:

• What are my total assets today? What do I expect them to be at the end of year one, two and three?

• What is my EA/AA today? What do I expect it to be at the end of year one, two and three? • What is my Loans/EA today? What do I expect it to be at the end of year one, two and three? • What is my current loan mix? What do I expect it to be at the end of year one, two and three? • Do I plan to grow investments a specific way, or are all investments instruments of opportunity? • What is my current investment mix? What do I expect it to be at the end of year one, two and

three? • What is my Deposit mix? What do I expect it to be at the end of year one, two and three? • Do I have any FHLB borrowing restrictions or preferences? • What is my current Non-Interest Income and Expense mix? How will it change over three years?

Users must answer these questions before they start setting pricing and targets, but these are not the only points of interest. Certainly, once the projected loan mix has been identified the user would then try to break that down further. For example, the user may want to see 25% growth in mortgage loans. The user then needs to clearly decide what types of mortgage loans they hope to grow. Will it be 25% across the board? 50% growth in ARM’s with a reduction in Fixed? The user must also try to identify growth patterns over the three year period. Will a majority of the changes take place in one year? Users that cannot clearly identify these numbers should refer back to Chapter 1.


The next thing the user must consider is how to input the various “ending points” or possible scenarios. Certainly there is no one set of answers to the questions above. If there were, institutions would have no reason to invest in models. The strength of modeling is that it should enable institutions to compare various forecast paths and choose the best one. In the FORESIGHT model each forecast path (each set of unique answers to the questions above) represents a Forecast/Budget. Specifically, a Forecast/Budget is a combination of Pricing and Targets that represent a specific future scenario. It is often easiest to jot notes down on a sheet of paper relating to the questions outlined above (for all Forecasts/Budgets), before trying to enter numbers into the model. 6.b. What is a Forecast/Budget? Institutions need to carry various Forecasts/Budgets in their model. Having multiple Forecasts/Budgets allows the user to test their ‘best guess’ assumptions, what we refer to as Base Forecast/Budget, against other possible outcomes, what we refer to as alternative strategies. There are two types of alternative Forecasts/Budgets. One type of alternative Forecast/Budget represents a drastically different ending point. Using this type of Forecast/Budget allows the institution to decide between numerous possible future growth patterns available during budgeting season. For example, an institution could decide to be a $500 million mortgage shop in three years or a $390 million auto lending shop. Inserting both possible ending points as unique Forecasts/Budgets will allow the institution to test which ending point provides the best returns. This type allows the institution to compare various business models or business polices. The differences in the Forecasts/Budgets tend to be very large in size. The second type of alternative Forecast/Budget is a marginal Forecast/Budget. The institution may have decided that they will be a $500 million mortgage shop in three years. The marginal questions they need to solve are borrowing versus CD specials or fixed rate mortgages versus ARMS. Adding additional Forecasts/Budgets to “test” these differences is the center piece of FORESIGHT. The differences between those Forecasts/Budgets tend to be small in size. Users often find the concept of multiple polices confusing. When thinking about Forecasts/Budgets it is often useful to think about vacation itineraries. The map shows our starting point in Madison, WI. We can go on vacation anywhere we would like, but in the initial decision making process we have narrowed down our choices to California, Texas and Florida. We have narrowed the choices due to our likes and dislikes. This is similar to the initial forecast/budget setting process. Using an understanding of their strengths and weaknesses, firms start by setting an initial business plan. Each vacation destination will require a unique travel itinerary. Certainly plane tickets bought for California will not get us to Florida. The successful traveler will plan to have enough money to buy a plane ticket ALL THE WAY to their destination. Over shooting your destination and ending up in Cuba instead of Florida may really dampen your vacation plans. Undershooting and ending up in Denver instead of on the beach in California would be equally disheartening. Remember, most people would not rent a car in Texas, book hotel rooms in Florida and then buy plane tickets to California.


Forecasts/Budgets are like travel itinerary, they get an institution some specific place in the future. Just like each vacation destination required a clear and concise travel plan, every possible future “ending place” for an institution will require a clear and concise game plan. That game plan is called a Forecast/Budget. Just like vacations, setting pricing to get you to ‘Florida’, targets to get you a car in ‘Texas’ and sales to book a hotel room in ‘California’ is sheer disaster. Setting Pricing to become a $500 million dollar ARM shop, while setting Targets to simulate a $350 million Auto shop, while forcing Sales to remove all of your Fixed rate mortgages will not create a useful simulation. It would be an adventure in modeling nonsense. Therefore, each Forecast/Budget should represent a unique combination of Pricing, Targets, and Sales. This combination should be chosen to reach a specific “end point” some time in the future. Separately, a Forecast and Budget are similar. They both require a lot of detailed pricing and target information. Typically a Forecast tends to look three to five years in the future and rolls forward each month with actual data, while a Budget is a one year forecast set in stone. To set Internal Assumptions, the user will select the Update Account Assumptions link from the All Tasks tab. 6.c. Pricing There are currently two different Pricing Methods available in FORESIGHT: Fixed Rates and Offset from Rate Driver.

The Fixed method should be used when an instrument does not change rates when there are shocks to the market. For example, Overdraft Protection, Credit Cards, etc. Users will not adjust the rates if market rates change. Credit Cards and some stock investments will not change rates with market adjustments. Before setting a Pricing Rule on any account, the user MUST verify that they are looking at the correct Forecast and select the FLAT rate environment.


6.c.i. Fixed Rates To set fixed rate pricing, the user needs to select the Pricing Tab on the Account Assumption Screen and the Method will Fixed Rates. The Inherit assumptions from option will not used in the Pricing screen. Since the pricing will be FIXED for this Forecast/Budget, the user will not need to select market drivers. Market rate drivers (found in the upper right) will be employed in the Offset Pricing method. The user must now decide if their origination rate will be constant (i.e. 9%) or if it will vary (i.e. 9% until June then jumping to 10%). To set the rate for all forecast periods the user will select the ‘Use Same Values’ check box. In the yellow Driver Rate box, the user will insert the interest rate. Selecting a new cell will copy the rate to all forecast periods. To set rates that change (i.e., some type of projected rate increase), the user must uncheck the ‘Use Same Values’. Now the entire Driver line is shown in yellow. The user now inserts the rates at the pivot points. A pivot point is a point where rates change. In this example we want FLAT 9% until June then a FLAT 10% from there on. The user will need to enter in 9% in January and June then insert 10% in July. To fill in between the spaces, the user selects each range (from pivot point to pivot point) and right clicks on their mouse to select various fill options (i.e. fill linear, fill constant etc). 6.c.ii. Offset from rate driver The Offset method is most often used to set a specific pricing rule in FORESIGHT. The offset rules reproduce an institution’s current origination and reprice rates as a function of a market driver and an offset. Now as market rates climb or fall (rate shocks in a forecast) FORESIGHT will price every instrument according to the Pricing Rule – market driver after shock plus current offset. The Origination Rate Used row must represent the rate that the bank is currently issuing that instrument. The reprice rate used row must represent the rate that the institution is currently repricing instruments at. Setting an account to Offset from Rate Driver requires the user to complete a two-step process for all fixed rate instruments and a four-step process for all variable rate instruments. Users must update their interest rates (see Chapter 5) before they set their pricing as current interest rates are the basis of every pricing rule. Users should also check and verify their pricing each month prior to running any forecast reports. To create a pricing rule for fixed rate instruments, users must first choose a market driver as the base of the rule. In the second step the user inputs their current origination rate into the model. FORESIGHT will calculate the rule’s offset from the driver and current rate. For variable rate instruments, users will still choose a market driver for originations. In the second step the user inputs their current origination rate into the model. Just as with the fixed rate mortgages, FORESIGHT will calculate the rule’s origination offset from the chosen driver. The third step is to choose a repricing driver.


These may be different for instruments such as a 5/1 and 7/1 ARMS where originations may be based off one treasury and the repricing is done off another. Fourth, the users inputs the reprice rate into FORESIGHT. The model will calculate the reprice offset from the chosen driver and required rate. As a reminder, we recognize that the actual rate that a client receives or pays heavily relies on each unique customer. For example, institutions have a wide range of rates and points for mortgages. It would be inconvenient to create an account in FORESIGHT for every combination. In fact, since institutions do not utilize customer credit as a download field, money is loaded into accounts according to loan characteristics. That means the money in a 30 Year Mortgage account in FORESIGHT, is the aggregate of all of the users 30 Year Mortgage money. Therefore it represents an aggregate of the institutions 30 Year Mortgage credit risk and points. Since the money is lumped together, the Pricing rule rates should represent the users average rates. When setting Pricing, the user must keep this in mind. Choosing a Market Driver (Origination Rate and Reprice Rate Drivers) A Pricing Driver is a market instrument that will be used as the starting point for a FORESIGHT account’s pricing rule. For example, most Commercial Loans are priced using Prime as a driver and many Mortgage Loans are priced using Treasuries or Agency Securities (FNMA, GNMA, etc.) as drivers. There are several ways for a user to choose a Pricing Driver; Following the Pricing Committee’s Rules, Comparing Historical Market Rates with Historical Bank Rates, or Trial and Error. Following the pricing rules set forth by your Pricing Committee is an easy way to set Account Pricing in FORESIGHT. This is often made difficult by the reality that the Pricing Committee sets very detailed pricing in some balance sheet areas (Mortgage Loans and CD’s) and very vague pricing in others (Commercial Loans). This inconsistency will create the need for the user to employ the Trial and Error or comparison method for some of their accounts. Comparing Historical Market Rates to Historical Bank Rates is very work intensive, but it will provide the user with a very good idea of which market instruments closely track their bank rates. When the user finds a market instrument that tends to change rates in a similar fashion as a bank instrument, and that market instrument seems like a logical choice, the user should pick that market instrument as a driver. There is a danger of ‘data mining’ while using this method. The user may find instruments that seem to be related, but in reality their similarity is completely accidental. As mentioned above, even if the instruments seem to move in tandem, the choice must still be logical. This hurdle will reduce the danger of ‘data mining’. To use this method, the user will have to track detailed historical data of numerous market instruments. The user will compare the trends of the independent market instruments to the trends and the coupon rates of their unique instruments. Managing and tracking this much data may be difficult and time consuming.


Using Trial and Error is the simplest method. You start out by choosing drivers according to a best guess. Commercial and Consumer loans would be tied to Prime. Investments and Mortgages would be tied to Treasuries or MBS’s. CDs and Other Borrowings would be tied to Treasuries. Each month the user would pay attention to the Offset from the chosen bogey and the actual account Origination Rate for that period. If the user notices large swings between negative and positive offsets, they can assume they have chosen the wrong driver. They would then need to pick a different market instrument that would act as a better driver. If the offset stayed relatively stable from period to period, the user could assume that they have chosen the correct market instrument to Price their FORESIGHT accounts. With the best origination and repricing drivers chosen the user creates pricing rules (at the account or parent level) by following the steps above. REMEMBER, the user should be setting pricing rules (initially) in a flat rate environment. Create the pricing rule according to the way you are pricing TODAY, let the model shock the rates up an down employing the user-defined rate shocks (See Chapter VII). 6.c.iii Formula The method ‘Formula’ allows the user to create a pricing rule that is connected to another account. A example is that your Jumbo CD interest rate on new money is 25 basis points higher than the like term regular CD. Set the pricing screen for your regular account as you would normally and use a Formula on the Jumbo CD, i.e. PE.ACCOUNT(“CD – 1 Yr”, NewRate, 0) + .25 would be the formula to have the Jumbo CD – 1 Yr look at the CD – 1 Yr rate and increase that by 25 basis points. A more complicated example used for a non maturing deposit might be: If Fed Funds rate is < 3% then pay .25%, if Fed Funds rate between 2.99% and 5% then pay .5% and if Fed Funds rate >= 5% than pay .75%. The formula for this would look like: PE.LESSTHAN(PE.MARKETRATE(“Fed Funds”,Ending,0),3.0)*.25 + PE.BETWEEN(PE.MARKETRATE(“Fed Funds”,Ending,0),3.0,4.99)*.5 + PE.GREATERTHANOREQUAL(PE.MARKETRATE(“Fed Funds”,Ending,0),5.0)*.75 6.c.iv. Lagging Drivers When using an offset rate driver on the pricing screen, you can use lag periods to specify when changes in the driver rate are actually applied. The lag periods are specified at the bottom of the pricing screen and can vary for increases or decreases in the driver rate. By default, or when the increases and decreases are specified as “Immediately”, changes in the driver rate are applied immediately when calculating the new instrument rate for that account. If changes are to be lagged, the “Lag as follows” option is chosen, and a number of periods to lag is then specified, with a default of 1 lag period. For each lag period, you specify the percentage of the change in the driver rate to apply in that period. Examples:

- Lag an increase in rates by waiting 3 periods, and then apply all of the increase. In this case specify Lag Periods of 3, with periods 1 and 2 at zero, and period 3 at 100%.


- Only apply 50% of any increase in rates, but apply it immediately. In this case specify lag periods of 1, and apply 50% in that period.

- Apply an increase in rates gradually, over 12 months. Specify 12 lag periods, and 1/12 (8.3%), 2/12 (16.7%), 3/12 (25%), … as the percentage to apply.

Notes:

- Choosing “Immediately” is equivalent to a lag period of 1 set to 100%. - You cannot reduce a percentage in a subsequent lag period. If you have applied 50% of a change in

one period, the following period must be set to 50% or greater. - If you change the lag values, click Calculate to see the results in the pricing screen. The displayed

rates used will account for the lag and represent the effective new pricing rates used per period. How Lag is Applied: In a shock rate environment, there is only one rate change; either up or down and equal to the shock in basis points. In these cases, the lag is applied by simply calculating the rate shock times the percentage applied specified. Once the number of lag periods is reached, the entire shock is applied (up to the highest percentage specified, which could be less than 100%) and the driver rate then stays level. In a yield curve driver rate environment, which may be gradual and may involve both increases and decreases, a potential rate change is calculated for each period. The lag information is then applied to that change and the change to apply for the next period is calculated. If a subsequent, same direction rate change occurs in a following period, and there is more than one lag period, any remaining changes to apply from the first change are applied, then the lag calculations are done again for the new change. The changes and the lags of those changes are thus applied cumulatively. However, if the rate change reverses, from say an increase to a decrease, any remaining changes to apply are discarded, and the rules that apply for the new direction take over. Those rules could be very different, so no attempt is made to accumulate prior changes in that event. 6.d. Targets Users can forecast changes in the size, shape and mix of their balance sheet and income statement by setting Targets. Before attempting to assign targets, the user must be 100% confident about what each account should do in the future for each Forecast they are adjusting. The user should refer back to Chapter 6.a for more discussion on “projecting” the future. The target METHOD’s currently available in FORESIGHT are:

1. Enter Values 2. Formula 3. Historical Trend 4. Related Account

In some cases Targets can be set at the parent level and forced down to the accounts, or set at the account level. The target method setting process is similar for both cases. Setting Targets Setting Targets in FORESIGHT requires the user to consider two questions.


First, the user must decide how an account’s balance will change over a forecast period (1-60 months).1 Targets could consist of very basic assumptions, such as ‘the balances will grow 5% a year for the entire forecast period’. The user could also choose more detailed growth paths, for example, ‘originate $1,000K a month for the first six months then increase by $250K for the next thirty months’. Second, the user must choose the best FORESIGHT Target Method and Method Option to achieve the prescribed future account balances. Remember accounts do not grow in a vacuum. The user must first know how the entire bank will change in the future. Then the user can identify what each account will do to get there. The first section will outline several methods modelers can use to estimate/calculate their account balances out into the future. The second section will describe the different Target methods available in FORESIGHT; when and how they should be used. To aid in the decision making process, we have separated the FORESIGHT Target Methods into groups; Most Common, Historical/ Trend, and Advanced Methods. The third section will outline the conflicts between Targets, Callables and Sales. The final section will provide systematic methods to test the validity of your assumptions. Estimating Account Balances Projecting the balances for your balance sheet or non-interest income/ expense accounts can be a very difficult task. Users should refer to Chapter 6.a for a discussion of forecasting methods. Most institutions have processes in place to create a budget so users should look first to existing budgets for growth information. If the budgets are incomplete or not available, modelers have other forecast estimation alternatives available. Modelers, if necessary can glean target information from history balances and trial and error methods. Using Budgets The easiest way to estimate future account balances is to use the organization’s budget as a guide. As discussed above, we suggest that users Target more than one year into the future. FORESIGHT allows users to create their own time period set for Targeting into the future. Be aware that many budgets only have a one year horizon which makes targeting beyond that point difficult. Users faced with this dilemma can extrapolate long term growth patterns from the one-year budget projections. This too can be problematic if the one-year budget contains several unique growth patterns. Below are two examples of this.

Example 1 – In your one-year budget 30-year mortgages grow from 1 million to 1.1 million. This is a growth rate of 10%. One could extrapolate a simple three-year growth rate of 30% (ending balance of 1.3 million) or a compound three-year growth rate of 33.1% (ending balance of 1.331 million). In this example, using the budget to project three year Targets is very useful. To input these projections into Foresight, users could choose the Growth Rates or Balance Target Method.

Example 2 - In your one-year budget, the bank decides to start issuing 28 month CDs. Over the first year, balances are expected to grow from 100 to 500,000. Since this is a new account, trying to extrapolate this growth rate into the future could be very misleading.

1 Typically, we suggest looking at a Three-Year simulation. Estimating balances three years into the future can be very difficult. Because of this, many of our users only plan one year into the future. The choice of time frames is up to the user, but they must be able to draw out the path of their balances throughout the time frame, which they choose.


Using History Another useful tool to estimate future balances is to look at historical trends. The user can employ historic numbers indirectly by manually calculating growth patterns or projecting dollar amounts. Extrapolating historical trends can lead the user into similar difficulties as using the unique budget method. Using Trial and Error Once the targets have been set, the user can employ the trial and error method to fine tune their assumptions. Fed Funds Sold & Purchased Method The Fed Funds Method is also often employed to find errors in your Target settings. Before one can rely on this method, users need to understand the theory behind it. Accounting Theory constrains us so that Assets = Liabilities + Equity. ALWAYS! FORESIGHT has been designed with two user defined system balancing accounts to make sure that the accounting formula holds. If you Target your Liabilities to grow too quickly, you will need to “make up” additional funds on the Assets side. FORESIGHT will add enough to Fed Funds Sold (balancing account on the asset side) to deal with the surplus money. If the Assets are targeted to grow too much, FORESIGHT will fund the growth with Fed Funds Purchased (balancing account on the liability side). Because of the accounting rule, institutions can use the size of their Fed Funds accounts in the forecast to fine-tune Targets. Retained Earnings Method The Retained Earnings Method is often used to find errors on your income statement. Often incorrect targets for the Non-Interest Income and Non-Interest Expense accounts cause major errors in your Retained Earnings. Incorrectly set Pricing could also cause runaway Retained Earnings. Excessive Retained Earnings may affect your Fed Funds position. For example, excessive earnings will increase Capital. All other things held constant, increased Capital requires an increase in an Asset account. FORESIGHT will default that required increase into the Fed Funds account. Users should employ Retained Earnings as a gauge to fine-tune their assumptions about Targets and Pricing. With the first question answered, (How will this account change over the next 3-years?), the user must now choose the most efficient target method.

6.d.i. Choosing a Target Method FORESIGHT allows the user to choose different Target methods according to growth expectations, described above. For example, users may project an ending dollar balance (Balances), users may have basic growth ideas (Growth Rates or Growth Offset), users may be able to estimate originations (Originations), users may want an account to run off the books naturally (Originations set to zero), or the user may want to use historical information to target balances into the future. A list of the Targets available in FORESIGHT can be found on the Target screen

Method: Enter Values Method Options: Balances – The values entered here will be the Ending Balances of the account unless the FASB Type of the account is either Available for Sale (FASB115 Mark to Market) or Trading Account. In that case, the balances entered will be the Amortized Cost balances and the Ending Balances will be determined by a present value calculation. For income/expense accounts, the values entered will be the actual income/expense for that period.


Growth Rates – The values entered here will be deannualized and used as a growth rate from the previous period, therefore actual results will be a compounded rate. Growth Offset – This method allows one account to grow at the same rate as a related account, plus or minus a percentage rate offset. Another account, the “Growth Offset” is selected. For Balance Sheet accounts the growth rate on the selected account’s Average Balances is used for as a base. This is the data that shows in the “Growth Rate” in the grid on the Target screen. Then the rate entered on the growth offset is added to the base growth rate. The growth offset is entered as a percent. For example, 4.00 is entered for 4.00%. This annual growth rate is deannualized and added to the Ending Balance Growth rate of the “Growth Offset” account. Originations – For balance sheet accounts, the values entered here will be used as the new dollars in the account as though new instruments with similar characteristics were put on the books in that period. These new dollar amounts are added to the remaining principal balances left after amortization maturity, or prepayment to determine each period’s ending balance. This method can not be used with non interest earning assets. If the user wishes to freeze the balances they would use this method with origination entered with zeros.

Method: Related Account Method Options:

Percentage of – The Percentage of Option allows the user to target a balance in an account as a percent of another account in the chart of accounts. An example would be to target a Fee Income account as a percentage of a loan accounts ending balance or it’s newly originated dollars. There are numerous Balances To Use with the percentage ranging from Ending Balance, Average Balance, Originations, Maturing Amounts, Prepayments, Debits, Credits etc.

Addition To ($) +/- - The Addition To ($) +/- Option allows the user to target additional balances coming into an account from another account on the chart of accounts. An example would be to have all maturing amounts from a Special Certificate of Deposit to roll into a selected deposit account. There are numerous Balances To Use as listed in the Percentage Of Option and the Target Is can be Ending Balance or Originations.

Method: Historical Trend Method Option:

Historical Average – A simple average bases on the number of historic periods specified. That one-time calculated number is then carried through the forecast. Moving Average – Calculated by averaging the most recent specified number of periods. Straight Line – Formula is Trend = a+(b*Time). The model uses the formula to fit a straight-line curve, through a specified range of historical data points. It then extends the straight-line trend out into the forecast, using least squares techniques to generate future balances. Exponential – Formula is Trend = a*(b^Time). The model uses the formula to fit a geometric exponential curve, through a specified range of historical data points. It then extends the geometric exponential curve out into the forecast, using least squares techniques to generate future balances.


Related Account Target the Allowance. Provision calculation: current months Allowance – previous months Allowance

Logarithmic – Formula is Trend = a+(b*LGN(Time)). The model uses the formula to fit a log transform curve, through a specified range of historical data points. It then extends the log transform curve out into the forecast, using least squares techniques to generate future balances. Parabolic – Formula is Trend = a+(b*Time^2). The model uses the above formula to fit a parabolic curve, through a specified range of historical data points. It then extends the parabolic curve out into the forecast, using lease squares techniques to generate future balances. The user can review all the historical trend options by selecting the Best Fit Analysis button and also display the forecasted dollars in a graph by selecting the graph option. A ‘Seasonally Adjust’ option is also available with targeting by Historical Trend.

Method: Formula Method Option: n/a The Formula method can be used when setting balance targets to specify a calculation that is more complex than available with the other methods. If, for example, you are targeting originations by rolling of maturing balances from another account, you can use the related account method. But if you would like originations to be equal to maturing balances in more than one account, you would need to use the formula method to calculate that number. The target formula method is also useful in calculations such as those needed for a reserve for loan loss account where the forecasted balances in that account will potentially depend on several other accounts including provisions, charge offs and recoveries. The formula method can also be used to target growth rates or originations by rate environment. For additional examples click on the Help button in the upper right of the screen and use the search function by typing “Formula Targeting”. To use the method, choose the Formula option in the Method option box. An Edit Formula button will appear which takes you to a new form that allows creating and editing of target formulas.


Additional Formula Target Balance Methods The Formula method can be used when setting balance targets to specify a calculation that is more complex than available with the other methods. If, for example, you are targeting originations by rolling of maturing balances from another account, you can use the related account method. But if you would like originations to be equal to maturing balances in more than one account, you would need to use the formula method to calculate that number. The target formula method is also useful in calculations such as those needed for a reserve for loan loss account where the forecasted balances in that account will potentially depend on several other accounts including provisions, chargeoffs and recoveries. To use the method, choose the Formula option in the Method option box. An Edit Formula button will appear which takes you to a new form that allows creating and editing of target formulas. The screen below shows an example of a formula that calculates a CD account's originations: The formula says that originations in this account will be equal to 2,000,000 plus 1/2 the maturing amounts in the "15 Mo Add On" account. The formula syntax is similar to Excel formulas (without the starting = sign), but there are many built-in functions that are specific to Foresight. In this example, we are using a function called PE.Account to retrieve the maturing amounts from another account. Notice that when a formula needs to refer to another account, you simply use the name of that account, in quotes, exactly as spelled in the chart of accounts. When filling in account names into formulas, you can drag and drop from the chart of the accounts on the right into the cursor location in the formula test. For interest bearing accounts, Foresight is smart enough to know whether it needs to lookup the balance sheet or income statement version of an account name. Another example of adding roll off from an account to another while keeping the receiving account at a zero growth rate. The result would have the account growing only by the amount of the roll off from the selected account. The following formula keeps a MBS account at a zero growth rate (balance staying flat) but adds the maturing and prepay amounts from a CMO account, i.e. MBS will grow only by the roll off from the CMO account. PE.GROWTHRATE(grt_Given, 0) + PE.ACCOUNT(PE.BSA("CMOs - HTM"), MaturingAmount, 0) + PE.ACCOUNT(PE.BSA("CMOs - HTM"), Prepayments, 0) Other Examples Here are some additional examples of formulas you can cut and paste into a formula edit box. Note that you will have to type or drop your chart's actual account names in place of the accounts referenced in these examples (unless the names are identical).


Calculate a reserve for loan loss. Note that the chargeoffs and recoveries are off balance sheet accounts, and use the PE.OBS function to get their values.

PE.ACCOUNT("Reserve For Loan Loss", AmortizedCost, -1) - PE.ACCOUNT("Provision for Loan Loss", FlowData, 0) + PE.ACCOUNT(PE.OBS("Charge Offs"), FlowData, 0) - PE.ACCOUNT(PE.OBS("Recoveries"), FlowData, 0) Note: that the sign in front of "Provision for Loan Loss" will depend on the characteristics of the account (Contra) and the sign of the entered data. Dollar Target Provision for Loan Loss (Method Option: Balances)

Reserve For Loan Loss

PE.ACCOUNT(PE.BSA("Total Loans"), AmortizedCost, 0) * - 0.015 Alternative Method: Related Account Target Method.

Provision for Loan Loss

PE.MIN(PE.ACCOUNT(PE.BSA("Reserve for Loan Loss"), AmortizedCost, 0) - PE.ACCOUNT(PE.BSA("Reserve for Loan Loss"), AmortizedCost, -1), 0) * - 1 Note: This formula will have to be adjusted for the sign of the "Reserve For Loan Loss". If "Reserve For Loan Loss" is a contra account then: PE.MAX(PE.ACCOUNT(PE.BSA("Reserve for Loan Loss"), AmortizedCost, 0) - PE.ACCOUNT(PE.BSA("Reserve for Loan Loss"), AmortizedCost, -1), 0)

Federal Reserve Formula

(PE.GREATERTHAN(PE.ACCOUNT(PE.BSA("Total Demand Deposits"), AmortizedCost, 0), 7000000) * PE.MINMAX(PE.ACCOUNT(PE.BSA("Total Demand Deposits"), AmortizedCost, 0) - 7000000, 7000000, 47600000 - 7000000) * 0.03) + (PE.GREATERTHAN(PE.ACCOUNT(PE.BSA("Total Demand Deposits"), AmortizedCost, 0), 47600000)) * (PE.ACCOUNT(PE.BSA("Total Demand Deposits"), AmortizedCost, 0) - 47600000) * 0.10

Calculate an expense account as a percentage of orginations in a balance sheet account.

PE.Account("Loan Account", Originations, 0) * .05 Alternative Option: Related Account Method

Calculate a number based on a rate environment

Growth Rates by Rate Environment (Amortized Cost): PE.GROWTHRATE(grt_SimpleDeannualized,PE.RATECHANGEGRIDMINUSPLUS100(1,0,10,20)Originations by Rate Environment (Originations): PE.RATECHANGEGRIDMINUSPLUS100(1,100000,50000,25000) or (100000 * PE.LESSTHAN(PE.RATECHANGE(1), 0)) + (200000 * PE.EQUAL(PE.RATECHANGE(1), 0)) + (300000 * PE.GREATERTHAN(PE.RATECHANGE(1), 0)) The RATECHANGE function returns the general change in the current rate environment as a decimal number. For shock rate environments, the function will always return the shock value where 100bp is 1.0 and a -300bp would return as -3.0. For gradual or yield curve environments it returns the change-to-date in the forecast based on the yield curve point specified as the parameter. In the example, point 1 is used, which would be the first (earliest) instrument on that yield curve. For parallel/relative rate environment shifts it does not matter what point you specify. So if your treasury yield curve is designated to go up 200bp in 1 year, then this function will return 0.167 in the first forecast period, 0.333 in the second forecast period, and so on. For non-parallel gradual rate environments, the value returned will depend on exactly how much that point of the yield curve changes from period to period.The LESSTHAN function returns 1 if the first expression (PE.RATECHANGE(1)) is less than the second one (0). The EQUAL and GREATERTHAN functions work the same way; returning 1 if the expression is true and 0 (zero) if not.


Calculate a number based on a Market Rate

PE.GREATERTHAN(PE.MARKETRATE("Prime",Ending,0),5.0) * 100000) + PE.LESSTHANOREQUAL(PE.MARKETRATE("Prime",Ending,0),5.0) * 200000)

Calculate market value change in a trading account to add to an income/expense account

PE.ACCOUNT(PE.BSA("Trading Account"), MarketValue, 0) - PE.ACCOUNT(PE.BSA("Trading Account"), MarketValue, -1)

Calculate FDIC Deposit Ins Expense

The first two months of a forecast the expense will be half of normal. PE.LESSTHANOREQUAL(PE.FORECASTPERIODNUMBER(), 2) * ((PE.ACCOUNT(PE.BSA("Total Deposits"), AmortizedCost, 0) * .0005) * .5) + PE.GREATERTHANOREQUAL(PE.FORECASTPERIODNUMBER(), 3) * (PE.ACCOUNT(PE.BSA("Total Deposits"), AmortizedCost, 0) * .0005)

FORESIGHT users will need to set Targets for all balance sheet accounts and all non-interest income and non-interest expense accounts. These must be reviewed/updated for each and every account, for each and every Forecast/Policy – each and every period. Setting of the Targets is a step in the Monthly Process. For discussion on the Monthly Process, please refer to Appendix – Model Rolling Checklist. 6.d.ii. Suggestions for Target Setting The best tool available to test your Target assumptions is the Forecast Balance Sheet. There are cases when the balances displayed in the report will be quite different from your expectations. Solving the ‘Incorrect Forecast Balance Sheet’ problem is not always intuitive. To understand why this happens, one must first understand the order of calculations performed by FORESIGHT. Below are parts of three different Line Item Simulation reports. The first was run with the Freeze method, the second 10% Growth, and the third is using Originations of $500,000 per month. FORESIGHT starts each month with a Starting Balance. From that the following items are either added or subtracted: Matured or Amortized dollars, Prepayment or Decay dollars, Premium or Discount dollars, Balloon payments, Called money, and new dollar originations. At this point FORESIGHT has a trial ending balance. The model will also examine all data for repricing opportunities, indexes, offsets, caps, floors, call prices and strike prices so it can accurately calculate a forecast. If you choose Freeze (left), FORESIGHT lets the money outlined above roll off and does not replace it. If you choose Origination (center) FORESIGHT will let all of the money roll off and replace it with that Targeted amount of Originations. In these cases, the Ending Balance is a function of money rolling off and the amount of new Originations. If you choose Dollars (right) the Ending Balance IS known. These Target methods tell FORESIGHT what you expect to have in any account at the end of a specific period. Therefore, in the right example the only piece that is unknown is the amount needed to Originate. Unlike the Freeze method, this is an example where new money is a function of what rolls off and the ending balance –FORESIGHT calculates amount originated rather than the ending balance2.

2 This is a very subtle point, but it is also very important. The user must recognize which numbers they control and which they have authorized FORESIGHT to calculate.


Problems can occur when you mix certain Target methods with Callables. FORESIGHT will call off the money, but then to reach the Targets set, FORESIGHT will Originate it all right back. A solution to this problem is to utilize other Target methods, such as Freeze (Origination with zero dollars chosen). Users should set pricing rules and growth targets in the flat rate environment in “Base Forecast/Policy”3 6.d.iii. Calculate An excellent way of checking your assumptions before running an institution wide forecast report is by using the Calculate tab on the Targets screen. For the account chosen, the Calculate function will display all the forecast cash flow data. This function can also be selected for a category, which calculates all data for the accounts in the category and displays the total data on the category level. Line Item Calculation Options When running a Line Item Calculation the following options are available: Calculate current only? This option lets the financial engine run the last relative period of history’s current book of business off. It doesn’t calculate any of the targets and therefore it doesn’t run off any of the new dollars. In effect, the user is presented with a database that replicates an end of history gap type report. Calculate Market Value. This option calculates market values, using the default rate environment, on all accounts. This allows the user to run a balance sheet showing the market values, in addition to a balance sheet showing actual values, book values or average values. The calculation of market values on all accounts may take a significant amount of extra time. Display calculation detail. This option turns on the calculator trace and after the line item simulation it will be displayed on the Messages tab of the Account Assumptions screen. Examination of the calculator trace may often answer questions concerning how, why and in what order things happened. The calculator trace shows the order of calculation, details about each account’s characteristics, calculated values, the amount of time each account took to compute and the total runtime of the line item calculation. Generating a Calculator Trace takes a small amount of extra time. Always recalculate all data. The financial engine saves time by determining what needs to be calculated based on when changes were made and when an account was last calculated relative to the time its dependencies where last calculated. If an account's dependencies were last calculated before the account itself and there were no changes made in the intervening time period the financial engine will simply return the last calculated values to the user. But if this box is checked, the line item simulation will always recalculate this account and all of its dependencies.

3 Base Policy is a concept – your best guess going 3-years out.


6.e. Other Tabs

6.e.i. Sales Tab The Sales tab can be used to sell existing cash flows in a forecast. To sell 100% of the line item, use the option selections at the bottom of the screen. First select the forecast date or periods in the future you would like the transaction to happen and then click the ‘Sell All’ button. The ‘Delete All Sales’ button will clear out all sales information on all cash flows. The ‘Delete Selected Sales’ will clear out only the sales information on the cash flow(s) that are checked in the ‘Selected’ column. To sell individual cash flows click on the Sell button on the cash flow line and the pop up box will have three selection tabs. The Amount & Price tab allows the choice of a sale in percent or dollars, selling price as an amount other than 100.0 and changes by rate environments. Any gain or loss from entering a Price other than the defaulted 100.0 will run through the Gain/Loss on Sale account assigned as a Special Account on the Institution Settings screen. Use the When tab to select the forecast sale date by date or periods in the future. The Conditional If tab allow you to build a rule similar to a call rule for the sale (if the rule is true than the sale will happen).

6.e.ii. Market Value Tab Market value setting on an account level are set in this tab and the Calculate button will display the Market Value calculation. Chapter 7 discusses Market Value in detail.

6.e.iii. Gap Tab

When a user is on the Gap tab and clicks on the Calculate button the model will calculate the repricing gap structure on the line item. The user may choose the time period set.

6.e.iv. Messages Tab The Messages tab displays a detailed outline of the assumptions and cash flow information after the user calculates from the Targets tab.

6.e.v. Comments Report Will produce a report with the comments entered on the assumption pages.

6.f. Forecast and Budgets In Foresight a Forecast/Budget is a combination of targets and pricing representing a unique future scenario. Each scenario has a unique ending point which requires a specific combination of targets and pricing for each item in the chart of accounts. Therefore each scenario will require a Forecast/Budget. The user can either Edit an existing Forecast/Budget or Create an entirely new one.

6.f.i. Create New Forecast/Budget To create a new Forecast or Budget the user will select Edit Forecast/Budgets from the All Tasks screen and then Insert New Item. The Create Forecast/Budget screen will appear for the user to complete.


1. When the user selects Clone From an existing policy they will start with a copy of that policy which includes all the assumptions. If the user does not select a policy to Clone From no assumptions will be set in the new policy. Our recommendation is to clone another forecast and make changes to assumptions rather than starting with nothing set and having to enter all internal assumptions.

2. Choose an appropriate name that clearly describes the new scenario and check the type of plan (Forecast vs. Budget) you are creating.

3. Verify the Start of Forecast date. This is the first month following the actual As Of Date of the model. If appropriate verify the Start of Budget date, which would be the first month of your fiscal year.

4. Choose a Time Period Set from the drop down box or create a new set. For a Forecast we recommend using a time set between 36 -60 months and for a Budget the time period needs to be long enough to cover the start of forecast through the end of the budget year.

5. Choose a Default Rate Environment. In normal circumstances we recommend setting this to the Flat Rate Environment for a Forecast. For a Budget the user will choose the appropriate rate environment expected for the budget year.

When the user has completed steps one through five, click on Done and the new Forecast/Budget will be produced and placed in the list in alphabetical order. When a Budget is finalized you will want to Lock it to prevent you from changing assumptions or rolling into the next year in error. Lock the Budget on the Edit Forecast/Budget screen; select the Budget and click the Lock selection. If the model requires you to run a forecast a pop up box will appear telling you this; just run a forecast balance sheet on the Reports screen in the proper forecast and with this done go back to the Edit Forecast/Budget screen and check the Lock box.

6.f.ii. Edit Forecast/Budgets To review all the institutions Forecast/Budgets in the model select Edit Forecast/Budgets from All Tasks menu screen. From this screen the user can: 1. Insert New Item: Use this option to create a new Forecast/Budget (see 6.d.i). 2. Delete Item: This option deletes an item and all its assumptions and results. 3. Update Forecast: Use this option to update a forecast after recent history is updated. This option

will be discussed in detail in the Rolling the model forward section. 4. Verify set up information and dates. For a Budget check the “Locked” box so the budget will not

roll forward or have the assumptions changed in error.


6.f.iii. Edit Time Period Sets Users may select a time period set for each Forecast created. Very seldom will the time period set be different between forecasts. To create a new set, edit an existing set, or review existing sets click on the green “Edit Time Period Sets” link. From here the user can view all sets, insert a new set or delete a existing set. When a user selects Insert New Set they will be prompted to give it a name and the number of periods. The user can customize the Length, Period Type and Display Name.

The example to the right shows a time period set with Length and Period Type set with a selection other than 1 and month.


6.g. Set up for Dividends 1. Setup a Dividends account as a Statistical Flow account, moved it so it is right under Net Income. 2. Setup a Footing Category called Change in Retained Earnings or Net Income after Dividends, move it right under Dividends. Edit the footing Category formula as: Net Income - Dividends. This account will show the amount of income that is being transferred to Retained Earnings on the balance sheet. 3. Go to the Institution Screen and open the Update the settings for this Institution screen and under the Special Accounts area add your Dividends account to the list next to the Dividends special account. 4. Go to Update Account Assumptions and enter your target for dividends. You can set dollar targets or formula targets off of number of shares. Using number of shares: 5. Setup a Common Stock Shares account as a Statistical Flow account in the memo area. Enter the number of shares in historical balance and/or in the account assumptions. The formula for Dividends then should look like this:

PE.ACCOUNT("Common Stock Shares", FlowData, 0) * .05 This would calculate a five cent per share dividend.

or PE.QUARTERPATTERN(0, 0, 100000) This would give a quarterly dividend with no dividend for the first two months of the forecast then $100,000 for the third month.


Chapter 7

Foresight Market Valuation 7.a. Net Portfolio Value and Valuation The value of a financial instrument and its underlying cash flows can fluctuate over time. For a financial institution it is important to understand how these changes can affect both available capital and the long term profitability of the firm. It is essential to recognize that the book value of an institution’s assets and liabilities may be drastically undervalued or overvalued due to changing market rates, market expectations, or asset impairments. Calculating Net Portfolio Value (NPV) provides institutions with a systematic understanding of their sensitivities to market changes. Net Portfolio Value is a term defined to represent the difference between the market value of assets and the market value of liabilities, in other words, the market value of the capital account. FORESIGHT provides users with a dynamic tool that estimates, in great detail, the long term effects that specific market shocks will have on the overall value of an institution. Modelers can simulate a number of dynamic ‘what if’ scenarios to determine the viability of their institution given changes in market rates and portfolio compositions. These simulations will provide the user with a solid understanding of their institution’s income-at-risk and value-at-risk. To understand the benefits of valuation, let’s first look at the accounting formula. Assets must equal liabilities plus equity (A = L + E). Regulators view capital as a safety net since gains and losses in a bank’s assets or liabilities are reflected in the capital account. The danger of using book capital as a measure of safety is that assets and liabilities are stated in accounting reports at book value. Since we know that the market values (actual values) of assets and liabilities can be different from their book values, we must assume that the NPV of an institution can be radically different from book capital. NPV is equal to the market value (i.e. present value) of expected cash inflows from current assets less the market value of expected cash outflows from existing liabilities.

NPV = MV (Assets) - MV (Liabilities) Or

MV(Assets) = MV(Liabilities) + NPV This formula, MV(Assets ) = MV(Liabilities) + NPV, is similar to Assets = Liabilities + Equity. From this we can see that NPV is a more accurate assessment of an institution’s safety than book capital. Investigating NPV can provide management with an insight into the soundness and profitability of their institution in various rate environments. It allows management to model “what if” scenarios to determine how the institution fairs in relationship to interest rate volatility. Market value provides an accurate glimpse of the value of firm’s instruments with regards to any imbedded options such as prepayment options, caps, or floors. Although market valuation can give us some idea of the liquidation value of the firm, it should be used with some discretion. Market valuation is not intended to determine a selling price or the total value of an institution, a single branch or even a portfolio of assets. Because there are many non-cash flow related issues that must be assessed for any of the latter to occur, an investment banker may be needed to determine the appropriate value of such things as goodwill, market share, or the soundness of an established customer base.


Amortization Table 30 Yr FRM, 7% Coupon

Principal Payment

Interest Payment

$0

$100

$200

$300

$400

$500

$600

$700

1 12 24 36 48 60 120 180 240 300 360

Principal Payment Interest Payment

7.a.i. The Valuation Approach Determining the present value of a financial instrument can be accomplished through the discounted cash flow model (DCF). Basic financial theory tells us that we can calculate the present value of any financial instrument given certain characteristics of the instrument. Two things are required to employ the DCF Model.

1. The size and timing of all future cash flows related to an instrument. 2. A reasonable discount rate for each instrument.

These two components are necessary to determine the present value of any financial instrument. After these two components are identified, the DCF model discounts all future expected cash flows by the respective discount rate(s) and then sums the present value of all the instrument’s cash flows to determine the net present value of the instrument. The arithmetic formula is as follows:

Present Value = ∑( Future Value of Cash Flowt) (1+i)t

Where: Future Value = Future Value of Cash Flow expected at time t

i = discount rate at time t t = time Before looking further into the DCF model and the valuation process for specific instruments, it is important to clearly understand cash flow characteristics and discount rates. Having a definite understanding of these two components is a necessity to the valuation process. 7.a.ii. Cash Flows Identifying the cash flows of an instrument under valuation is imperative in the valuation process. As defined in the DCF model above, each and every cash flow is valued independently. In most cases, the size and timing of an instrument’s cash flows can be readily identified. For example, the amortization schedule of a 30-year fixed rate mortgage with a coupon rate of 10% is straight forward and can be easily calculated. As depicted below, we can reconstruct the size, timing, and compositions of each and every cash flow associated with this instrument. Initially, for purposes of simplicity, let’s assume there are no prepayments associated with this particular instrument.

Year 1 Year 2 Year 3 Year 4 Year 5 Year 10 Year 15 Year 20 Year 25 Year 30 Interest Payment $577.92 $571.60 $564.83 $557.56 $549.77 $501.53 $433.13 $336.17 $198.72 $3.86 Principal Payment $87.38 $93.70 $100.48 $107.74 $115.53 $163.77 $232.17 $329.13 $466.59 $661.44 Total Payment $665.30 $665.30 $665.30 $665.30 $665.30 $665.30 $665.30 $665.30 $665.30 $10.50


The cash flows of this instrument can be seen both in the graph and amortization table above. Here we have a mortgage that amortizes over 30 years. Though there are monthly payments of $665.30, the portion of each payment that is applied towards principal reduction increase over time and the portion of each payment that is applied towards interest decreases as time progresses. It should be noted that the additions of prepayments would increase the principle pay down shown in the graph. Cash flow break-outs can be done manually for any instrument on the balance sheet in a similar fashion. The principal pay down schedule of every instrument on the balance sheet are displayed in a GAP report also. For more information on cash flows of other types of instruments refer to Appendix A and B. It is important to note that options can have a material affect on both the size and timing of any future cash flows. In the discussion above we established the importance of accurate cash flow projections. Because prepayments shorten the life of amortizing instruments, they too must be a factor in cash flow projections. FORESIGHT has the capability to apply a number of prepayment assumptions to existing cash flows on the balance sheet. FORESIGHT will also model other options such as reprice, calls, and cap/floor options. 7.a.iii. Discounting Once the size and timing of future cash flows are projected, the next step in the DCF process is to determine an appropriate discount rate. Much like accurate cash flow projections, the estimation of the discount rate becomes very important in valuing a financial instrument. Essentially, the discount rate is the required rate of return the market demands given the cash flow timing, risks, and costs associated with holding a particular instrument. In order to clearly understand the discounting process, we can think of the discount rate, or required rate of return, as being composed of two main segments: a time factor, or Interest Rate Risk factor; and the Credit Risk/Option Risk factor. The time factor, or Interest Rate Risk component, simply represents the value associated with holding a particular instrument for a given period of time. The second component, Credit Risk represents the default risk or option risk associated with the instrument. Credit Risk/Option Risk is often very difficult to identify. With a general understanding of DCF, let’s look at a specific example: A 30-Year Mortgage with a 7% coupon. We can easily project the future mortgage payments, - giving us the value for Future Value of Cash Flow and t in the present value formula. What should we use as a discount rate? Since FNMA rate is 6.25%, perhaps we could discount all of the payments by the discount rate of 6.25%. Because we know that the discount rate is nothing more then the current required rate of return that market demands for a given set of cash flows, we should then be able to apply this current market rate of 6.25%. But wait, does this seem feasible. Can we assume a flat discount rate over the life of the instrument? For example, if a customer has $1000 they would like to invest in a CD, they have a number of options on how they can invest their money: 6 months, 1year, 2 years, 3 years, etc. Without being a financier the investor knows that if they give their money to the bank for one year they will get a rate of let’s say 1.25%, if it is two years the rate will be slightly higher, and if it is three years it will be even higher. These investors will earn a higher rate of return. This relationship will continue, as the holding period increases so does the required rate of return. These customers are recognizing that if they give up their money for longer periods of time they must be compensated more heavily for their commitment. This relationship is also apparent under normal conditions of the Treasury Yield Curve. The Treasury Yield Curve below suggests, as time to maturity increases so does the market’s required rate of return for a given set of cash flows, holding all else equal.


Treasury Yield Curve

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

3.0%

3.5%

4.0%

4.5%

5.0%

5.5%

1 Mo. 3 Mo. 6 Mo. 1 Yr 3 Yr 5 Yr 10 Yr 15 Yr 30 Yr

Term

Yiel

d

In order to properly VALUE future payments, one must use larger and larger discount rates as the timing of the cash flow extends further into the futures. These higher rates correspond to the higher rates of return investors require to invest their money for a given term. With the bank CD and treasury yield curve examples we can establish that the market demands higher compensation as the time to maturity of any financial instrument increases. By using a flat discount rate, we would be ignoring the time value of money, a concept that is a fundamental theory of finance and would lead us to drastically inaccurate valuation projections. By using this convention we would essentially be saying that we could pay one rate on all of our CDs regardless of term of the instrument. Without using a dynamic set of discount rates that reflect the true, underlying values of the cash flows at each incremental point in time, the value of the instrument would be drastically misstated. Now that this constant discount rate method proves to be flawed, what do we do? We must determine another method that is capable of achieving multiple discount rates (i.e. incorporating the concept of term structure outline above). If there was a Mortgage Yield Curve actively traded in the market, we would use that curve to discount this mortgage instrument; however, there is no such curve. Instead, we could take an existing curve that reflects instruments that are readily available and easily valued, and modify that curve for this mortgage discount curve or create a synthetic curve. This leads us to the concept of an alternative portfolio. Alternative portfolio theory in finance says that if there is a similar set of cash flows in the market that are already being valued, we can use that valuation approach and apply it to the similar cash flows that are not currently being valued by the market. Is there anything in the market currently being valued that we can use to help us value the various cash flows of our 30-year fixed rate mortgage? At first glance this might seem difficult. Instead of viewing our mortgage as a single item, let’s think of a mortgage as 360 unrelated payments, each payment containing part principal and part interest. Could we recreate a portfolio of zero coupon strips that paid off similar to a 30-year mortgage? Yes. Could we easily value this portfolio? Yes. The present value of all the Treasury Strips in this “alternative” portfolio would equate to the amount that we would purchase the portfolio for today.


Another alternative would be to create a FRM Synthetic curve using the TBA 15 year and 30 year market rates. The duration points could be plotted on an agency curve and interpolated between known points. Since US Treasuries are one of the most widely traded instruments in the market and have terms ranging from 1 month to 30 years, this may be a solid foundation for building a discount rate curve. The current value of each treasury instrument can be determined from the treasury yield curve. The treasury yield curve represents the yield the market demands for a given treasury of a given term. By looking at the treasury yield curve we can determine the appropriate discount rate needed to value the individual strips. Since the mortgage cash flows match those of strips, we can use the FORESIGHT discount rate for both; the treasury yield curve. However, if we stop here and only use the Treasury Yield Curve as a measure of the discount rate we would be ignoring another fundamental component of finance. When we defined the two components of the discount rate, we established that one component was the time component, or interest rate risk component and the other was a credit risk component. By using the treasury yield curve absent of any credit risk offset, we are assuming that each individual mortgage payment has no option or default risk associated with it and that its overall level of risk correlates to that of the Treasury Yield Curve. This is fundamentally unsound. In order to accurately reflect the value of the instrument the credit risk associated with the instrument must also be quantified. How shall we identify the credit risk spread as it relates to the Treasury Yield Curve? Regulators and the finance world suggest a parallel shift to the Treasury Yield Curve. We will follow this methodology, but it is first helpful to recognize that a parallel shift is not necessarily an accurate assumption of credit risk over the life of an instrument. We can use the example of two different homeowners to make this point. First, consider a new homeowner that is just out of school carrying student debt. This homeowner is more likely to default on their loan because they are new to the work force and have little equity in the home (See Chart Below). However, as the loan becomes more seasoned, option and default risks will decrease as the mortgage approaches maturity, where the risk will be nearly zero.

Credit Risk (As Seen By Borrow er)

00.05

0.10.15

0.20.25

0.30.35

0.40.45

0.5

1 15 30

Years

Ris

k O

ffse

t

Treasey Yield Curve Risk as Seen by Borrow er


Credit Risk (As Seen By Lender)

00.1

0.20.30.4

0.50.6

0.70.8

1 15 30

Years

Ris

k O

ffse

t

Treasury Yield Curve Credit Risk as Seen by Lender

On the flip side of this scenario, one could also argue that at the time the loan is made, the borrower is financially sound enough or creditworthy enough for the institution to lend them the money (See Chart Below); however, as time progresses the individual could lose his/her job and default on the loan. These are all unpredictable circumstances from the stand point of the bank when underwriting the loan.

Following these methods, determining the actual credit risk of each and every loan in an institution would be a long and tedious task, with limited returns. Fortunately it is industry standard to assume a parallel shift over the treasury yield curve. How should we quantify this option spread. To estimate option risk we will need to make one final assumption. We will assume no institution will have a mortgage at a rate that does not cover the cost of making it. The two major costs (risks) banks face when making a loan are Interest Rate Risk (IRR) and Credit Risk (CR). Therefore:

Market Rate (MR) ≥ IRR + CR

The appropriate market rate can be easily identified. For actively traded instruments (FNMA, FHLMC) one can look to Wall Street. For other instruments one can estimate the current market rate by looking at survey data or by researching their own market. With the market rate identified one needs to estimate IRR in order to find a value for CR. How can one estimate the IRR of an instrument? First, let’s start by agreeing that the Treasury Yield Curve is a direct measure of IRR. In fact, there are no other factors driving the slope of the Treasury Yield Curve other than expectations of IRR. This means that if we could compare our mortgage to a point on the Treasury Yield Curve, we could identify its IRR value. Given MR and the IRR, one could easily solve for CR. The difference would be the spread that reflects the credit risk offset. This final leap, comparing the 30-year mortgage to a point on the Treasury Yield Curve is not an easy one. However, this step is much easier with the concept of Duration. Duration is a measure of IRR. For example, an instrument –ANY INSTRUMENT - with a duration of 6 has the same sensitivity to a small rate shock as does the 6-year zero coupon treasury. An instrument –ANY INSTRUMENT- with duration of 4 has the same sensitivity to a small rate shock as a 4-year zero coupon treasury.


Treasury Yield Curve

0

1

2

3

4

5

6

7

8

9

1 Mo 3 Mo 6 Mo 1 Yr 3 Yr 5 Yr 10 Yr 15 Yr 20 Yr 25 Yr 30 Yr

Term in Years

Yeild

%

Yeild 30-Yr FRM Rate Parallel Shift

Duration, despite what some may think, can be relatively easy to understand. Duration, as defined by Frank Fabozzi, is the measure of the appropriate sensitively of a bond’s value to rate changes. In other words, duration tells us the interest rate sensitivity of an instrument due to a given change in interest rates.

Duration = Sum of (Time * Present Values) Price

Therefore, if we know the duration, we know what treasury instrument to compare our 30-year Fixed Mortgage. This comparison provides us with the Interest Rate Risk measure. Simple math leads us to CR = MR – IRR. This Credit Risk would then be used as the offset to the Treasury Yield Curve at the various cash flow points. Now that it is established that a parallel shift in the overall yield is adequate for quantifying and reflecting the credit risk associated with a particular instrument or portfolio of instruments, the offset or overall shift must be quantified. There is one final note regarding the estimation of an instrument’s duration; when estimating the duration of any instrument, it is important to recognize that based on the assumptions used and the specific characteristics of the market the instrument is issued in, the measure can vary widely. Typically, it is agreed that the duration of a 30-year mortgage is somewhere between 5 and 9 years, depending on the current market expectations and coupon rates. It is important to use a duration that accurately reflects the characteristics of the instruments as it relates specifically to the institution. Let’s again go back to our hypothetical 30-year fixed rate mortgage that was booked at 7%. If we assume that current market rates have declined to 6.25% (FNMA Rate) we can then measure the difference between the treasury yield curve and the current market rate of 6.25% to determine the appropriate credit risk offset. For the case of our mortgage let’s assume that it has duration of 7. This means that our 30-year fixed rate mortgage has the same interest rate risk characteristics as a 7-year zero coupon treasury instrument. With this being said we can then proceed to quantify the appropriate risk offset to value our FRM. Because the difference between the 7 year point on our yield curve (4.25%) and the current market rate of 6.25% is approximately 200 basis points (6.25% - 4.25% = 2.00%) we can assume that the appropriate risk offset for our instrument is 200 basis points over the seven year point on the treasury yield curve.


After determining the appropriate discount rate offset from the treasury yield curve and identifying the instrument’s duration point, we can identify the appropriate discount rate for each and every cash flow associated with that instrument. This allows us to apply this offset across all points of the Treasury Yield Curve. This approach assumes a uniform credit risk over the life of the instrument but each payment is adjusted for the interest rate or term risk which is measured by the Treasury Yield Curve. The method followed above can be used to value any fixed rate instrument. This same approach can be extended to not only fixed rate instruments but also repricing instruments, such as a 1-1 ARM, or a 3-1 ARM. However, these instruments are somewhat unique in that they can reprice. When performing market valuation for repricable instruments, the reprice frequency must be a factor. In the case of a 1-1 ARM, the appropriate market bogy, or spot on the Treasure Yield Curve, is the 1-year spot because it is repricing on an annual basis. However, for other hybrid instruments such as the 5-1 or 7-1 ARMS, they may not even remain outstanding by the time they hit their reprice date. So in such a case it may be appropriate to assume the same methodology we used for valuing a fixed rate instrument. Since instruments can be valued as either fixed or variable, the user may want to value a hybrid once as a fixed instrument and once as a pure variable instrument. Theses two estimations will provide a range of possible values but hybrid’s actual value will fall somewhere in between. Valuation of liabilities or deposits is a much simpler venture, when compared to the valuation of assets. Because there is no external credit risk associated with liabilities, we can look at the cost of our alternative source of funds. The alternative source of funds represents the next best resource for obtaining funding. If we compare the cost of our alternative source of funding to our current funding costs we can determine how effectively or efficiently the institution is able to raise funding sources. If a bank were to need additional funding sources because it was unable to raise the appropriate funding sources through its deposits, it would turn to the Federal Home Loan Bank as their alternative source of funding. Therefore, the easiest way to determine the value of a liability is to discount it with its alternative source of funds – the FHLB curve with no offset. Note: Effective March 31, 2005 Farin & Associates has completed an initial phase of creating synthetic curves to use with MV calculations. See 7.b.

7.a.iv. Setting Up Valuation in FORESIGHT As outlined in the theory above, one needs both future cash flows and discount factors in order to calculate an instrument’s market value. FORESIGHT will use the cash flows and characteristics for each account in the model and the rate environment selected. The user needs to decide on the correct discount factors to use. The steps to estimate the discount rate offset are outlined below.

I. Asset Valuation a. Fixed Rate Assets

i. What instrument are you trying to value? Think about all of the characteristics of the instrument’s cash flows in the real world. Make sure that you have inputted those characteristics into FORESIGHT correctly.

ii. Estimate the TRUE life of the instrument. A 30-year mortgage has a duration of something much closer to 7 years. What is the duration of the instrument listed in (i)? Is it 7 or 8 years?

iii. Find a point on the Treasury Yield Curve that is similar to the term estimated in (ii). If the duration of the instrument is 7 or 8 years, one will have to choose a treasury instrument as close to that as possible. What is the current rate of the treasury instrument chosen?


iv. Find the MARKET rate of the instrument outlined in (i). Do not simply use the coupon you are currently charging. You want to find the TRUE MARKET rate.

v. The discount rate offset = (iv) – (iii). You should input this number on the Market Value screen in FORESIGHT in the Offset box.

vi. The discount rate driver in FORESIGHT should be the Treasury Yield Curve. b. Variable Rate Assets

i. What instrument are you trying to value? Think about all of the characteristics of that instrument’s cash flows in the real world. Make sure that you have inputted those characteristics into FORESIGHT

ii. Estimate the repricing frequency of this instrument. A 1/1 ARM should be seen as a frequency of 1 year. A 3/3 ARM has a frequency of three years.

iii. Find a point on the Treasury Curve that is similar to the repricing frequency in (ii). What is the current rate of the treasury instrument chosen? The examples above would require you to chose the 1 year or the 3 year Treasury as the base to calculate the offset.

iv. Find the MARKET rate of the instrument outlined in (i). Do not simply use the coupon you are currently charging. You want to find the TRUE MARKET rate.

v. The discount rate offset = (iv) – (iii). You should input this number on the Market Value screen in FORESIGHT in the Offset box.

vi. The discount rate tie in FORESIGHT should be the Treasury instrument chosen in (iii).

II. Liability Valuation All deposits are basically valued off of the LIBOR curvei. The OTS chose to switch to this curve in March 2001 to provide better valuations figures. Deposits require an additional Future Value Cost offset. This cost reflects that fact that it costs the bank money to have tellers, passbooks, process checks, etc. Different types of deposits have different future value costs that need to be entered into FORESIGHT. FORESIGHT adds these costs to the coupon paid on the instrument, lowering the estimated Deposit Intangible. Other borrowings are valued similar to the assets outlined above.

a. Non Maturity Deposits NMD’s do not have guaranteed cash flows like CD’s or loans. Institutions need to estimate when dollars will roll out. The OTS provides acceptable Decay Rates for institutions to use when estimating their NMD roll off. Institutions can also calculate their own in house numbers or utilize external providers such as McGuire Solutions. We provide a set of archived newsletters dealing with methods to create in house Decay Rates, at: http://www.fpwrestle.com. FORESIGHT uses the decay rates to estimate cash flows for the NMD’s. The cash flows are discounted similar to the assets.

i. All deposits are tied to/ discounted by the LIBOR Curve. One can simulate a LIBOR Curve by adding an offset to the Treasury Yield Curve. This offset is always changing. Please refer to our web page to get the most up to date number: http://www.farin.com/extras/ex_otsdecay.htm.

ii. Once the user has either selected, or simulated, the LIBOR curve they must input a discount rate offset calculated by the OTS decay formula. This offset is always changing. Please refer to our web page to get the most up to date number: http://www.farin.com/extras/ex_otsdecay.htm. If you are simulating the Treasury Curve, you will need to add BOTH offsets together.

iii. You will need to add in the Cost Adjustment, the number can be found at the very bottom of our Decay Rates web page. This rate is updated annually.

b. CDs i. All deposits are tied to/ discounted by the LIBOR Curve. One can simulate a LIBOR

Curve by adding an offset to the Treasury Yield Curve. This offset is always changing. Please refer to our web page to get the most up to date number: http://www.farin.com/extras/ex_otsdecay.htm.


ii. You will need to add in the Cost Adjustment, the number can be found at the very bottom of our Decay Rates web page. This rate is updated annually.

c. Other Liabilities i. Other FHLB borrowings should be valued in a similar fashion to the assets using the

FHLB advance curve. This curve can be simulated from the Treasury Yield curve by adding an offset. Typically this offset will range from 60 to 75 BP depending on the FHLB district. Fixed amortizing instruments should be valued by a Yield curve. Variable rate and bullet instruments should be valued by specific treasury instruments plus the spread to the FHLB Advance curve.

Examples


7.b. An Alternative Discount Rate – Synthetic Curves When selecting a Discount Rate the user should select a security or debt instrument as close as possible to the line item being valued. If the market instrument is a less than perfect, adjustments must be made to offset for Interest Rate Risk, Option Risk, Credit Risk and Servicing cost. The closer your wholesale match is to the instrument being valued the easier it is to come up with the discount rate. Farin Foresight will assist the valuation process by supplying market rates with the majority of them in yield curves. However, some market yield curves are incomplete. So we also make a number of synthetic yield curves available to be used in the market value calculations, i.e. Fixed-Rate Mortgage Backed Securities curve, Balloon MBS curve, Automobile Asset-Backed Securities curve and Equipment Asset-Backed Securities curve. A synthetic curve is built by starting with an appropriate base curve, plot data points from available securities and interpolating between the available points. General Rule #1: In valuing a fixed-rate financial instrument, use a complete yield curve whenever possible. General Rule #2: In valuing an adjustable-rate financial instrument, use a single discount rate with a duration equal to the instruments reset frequency. Examples REM – 30 Year Fixed


REM – 5/30 Balloon Used the synthetic Balloon MBS curve as the discount rate. These rates reflect both interest rate risk and option risk. Used OTS industry standard cost adjustment for mortgages (20bps). The 5 bp offset to the curve is an adjustment for credit risk based on our own loss experience on 5/30 yr balloon mortgages plus25 bp to offset market securitization. REM – 30 Year 1/1 ARM Used the GNMA 11 ARMS as the discount rate. These rates reflect both interest rate risk and option risk. Used OTS industry standard cost adjustment for mortgages (38bps). The 5 bp offset to the curve is an adjustment for credit risk based on our own loss experience on adjustable rate mortgages plus 7 bp to offset market securitization. Commercial RE – 15 Year Fixed Used the synthetic MBS curve as the discount rate. These rates reflect both interest rate risk and option risk. Used OTS industry standard cost adjustment for mortgages (20bps). The 10 bp offset to the curve is an adjustment for credit risk based on our own loss experience on 15 yr fixed rate mortgages plus 25 bp to offset market securitization. Commercial RE – 5/15 Balloon Used the synthetic Balloon MBS curve as the discount rate. These rates reflect both interest rate risk and option risk. Used OTS industry standard cost adjustment for mortgages (20bps). The 10 bp offset to the curve is an adjustment for credit risk based on our own loss experience on 5/30 yr balloon mortgages plus25 bp to offset market securitization. Commercial RE – 15 Year Variable Used the GNMA 11 ARMS as the discount rate. These rates reflect both interest rate risk and option risk. Used OTS industry standard cost adjustment for mortgages (38bps). The 10 bp offset to the curve is an adjustment for credit risk based on our own loss experience on adjustable rate mortgages plus 7 bp to offset market securitization. Commercial – 10 Year Fixed Used the Indexed B Corporate Bond curve as the discount rate. These rates reflect both interest rate risk and option risk. Used OTS industry standard cost adjustment for commercials (20bps). The 15 bp offset to the curve is an adjustment for credit risk based on our own loss experience on these types of loans. Commercial – 10 Year Variable Used the US CMT 1 month point as the discount rate (duration). This rate reflects interest rate risk. Used OTS industry standard cost adjustment (20bps). The 15 bp offset to the curve is an adjustment for credit risk based on our own loss experience on these types of loans plus 150 bp to offset additional option risk. (current market rates are 150 bps higher than the 1 Mo CMT) Commercial – 5/10 Balloon Used the US CMT curve as the discount rate. This rate reflects interest rate risk. Used OTS industry standard cost adjustment (20bps). The 15 bp offset to the curve is an adjustment for credit risk based on our own loss experience on these types of loans plus 200 bp to offset additional option risk. (current market rates are 200 bps higher than the 5 Yr CMT) Auto – 5 Year Fixed Used the AAA Auto Index curve as the discount rate. These rates reflect both interest rate risk and option risk. Used OTS industry standard cost adjustment for mortgages (20bps). The 20 bp offset to the curve is an adjustment for credit risk based on our own loss experience on auto loans plus 50 bp to offset market securitization.


Personal – 3 Year Fixed Used the US CMT curve as the discount rate. This rate reflects interest rate risk. Used OTS industry standard cost adjustment (20bps). The 20 bp offset to the curve is an adjustment for credit risk based on our own loss experience plus 200 bp to offset additional option risk. (current market rates are 200 bps higher than the 3 Yr CMT) HELOC - Variable Used the US CMT 1 month point as the discount rate (duration). This rate reflects interest rate risk. Used OTS industry standard cost adjustment (20bps). The 20 bp offset to the curve is an adjustment for credit risk based on our own loss experience plus 300 bp to offset additional option risk. (current market rates are 300 bps higher than the 1 Mo CMT) Credit Cards Used the US CMT curve as the discount rate. This rate reflects interest rate risk. Used OTS industry standard cost adjustment (100bps). The 30 bp offset to the curve is an adjustment for credit risk based on our own loss experience plus 300 bp to offset additional option risk. (current market rates are 300 bps higher than the CMT) Deposits - All Use the Average FHLB ADV curve plus the appropriate Cost Adjustment for the type of product. Cost adjustments are listed on our website. 7.c. Line Item Calculations Foresight computes present value/duration numbers and statistics when it calculates a line item on each account. It does the same calculations whether a line item market value is requested, or when doing a market value report. This topic summarizes the calculations and methods for present value, weighted average maturity, duration, convexity and interest rate elasticity. Summary Report An example of the summary portion of a line item market value report is presented below.

Present Value Report As Of 08/31/2006 Sample National Bank Forecast: New Base Policy Rate Environment: I & P 0 Account: Comml REM - 5 Yr Fixed Annualization Technique: Actual/365 Summary Beginning Balance: 6877531.46 100.00 Maturing Balance: 5730624.39 83.32 Ballooning Amount: 0.00 0.00 Calling Amount: 0.00 0.00 Prepaying Balance: 1146907.07 16.68 Sales: 0.00 0.00 Prem/Disc Balance: 0.00 0.00 Total: 6877531.46 100.00


Present Value: 6998985.26 101.77 WAM (Years) 1.33 Macaulay Duration (Days) 421.02 Macaulay Duration (Months) 13.86 Macaulay Duration (Years) 1.15 Modified Duration -1.10 Convexity 0.01 Interest Rate Elasticity -1.08 Discount Rate (YTM) 5.21

The first part of the summary portion shows the dollar and percentage break out of how the balances were distributed among maturities, balloons, calls, prepayments, premium and discount dollar amounts and percentages of the total are shown. If the total amount and percentage doesn’t equal the beginning balance then something was missed. If this occurs, the messages tab should be checked for any pertinent information concerning the problem. Present Value The present value of the beginning balance is shown next. The percentage number shown on the same line is the market to book price. In this example there is a 1.77% premium. Present value is calculated as: PV = CF * (1 / ((1 + i) ^ n), where the discount factor = 1 / ((1 + i) ^ n), remembering that i is a deannualized daily rate and that n is stated in days. The daily rate is calculated based on the account’s annualization method. Weighted Average Maturity Next shown is the weighted average maturity (WAM). This calculated value is unaffected by repricing. WAM = Sum(CF * n) / Beginning Balance, where n is days into the forecast. Macaulay Duration 'p / p = -D 'r / (1 + 'r), where D 'r = Macaulay Duration Modified Duration Modified Duration is the first derivative of the price yield equation. It is computed by using the Macaulay Duration in years divided back thru by (1 + y), where y is the yield to maturity of the instrument. Convexity Convexity is the second derivative of the price yield equation. Convexity Periods = SUM(t(t + 1) * PV(t)) / PV((1 + YTM) ^ 2), where t = number of periods in the investment, PV(t) is the present value of the cash flow in time t, PV is the Cumulative Present Value of the investment. Convexity Days = (Cp / (1 + (YTM / n) ^ 2) * PV, where n is payment periods per year Convexity Years = Convexity Days / (365 ^ 2) Convexity (Price Change Due to Convexity) = 0.5 * Convexity Years * (Price Change ^ 2) * 100, where the Price Change is an industry assumed value of 1% (100 basis points). Interest Elasticity Interest rate elasticity is the sum of the first and second derivates of the price yield equation. Interest Rate Elasticity = Modified Duration + Convexity Discount Rate/IRR If the instrument was discounted off a yield curve, then the discount rate is the computed internal rate of return, derived by setting PV = Sum(DCF). This can be checked by dropping the detail portion of the present value report into Excel and substituting this single discount rate for the yield curve derived discount rates.


Interest Cash Flows Foresight calculates interest cash flows assuming accrual accounting. Each interest cash flow is discounted as of the last date of the period the interest accrues in. This can result in the last interest cash flow having a longer term than the final maturity payment of principal, but only by a number less than 31 days. Repriced Cash Flows For instruments that reprice, as Foresight is primarily a static discounted cash flow model, the present value is calculated until maturity using the repriced interest flows and maturing balances or ‘scenario analysis’. For purposes of Duration thru Interest Rate Elasticity values, the instrument is recalculated until the first repricing. This is because an instrument which reprices to market (a market rate) will reprice to par at that time (ignoring cap, floors can changes to credit quality). Thus duration must be slightly less (not equal to due to it being an interest paying instrument) than the time to the first repricing. 7.d. Adding Off Balance Sheet accounts to Market Value Reports Foresight versions 5.0.9 and later support the ability to report off balance sheet accounts on the market value reports (a report with a Balance Sheet row set and a Market Value Spread column set). It is done by using a new system account to indicate where on the chart of accounts the off balance sheet accounts reside. Normally, the market value report will show Assets and Liabilities at the chart level specified, and then it calculates portfolio equity based on the calculated difference between total assets and total liabilities. To include off balance sheet items in market value, do the following: 1. In the Edit Accounts Dictionary off of the Report Tasks screen, you will see a new system account called “OBS Market Value Category”. Assign this item to a category in the off balance sheet section of your chart that contains the statistical accounts you want to see on a market value report. 2. Statistical accounts are included if they meet these tests:

• they are designated as Statistical Balance account types • they are designated as interest bearing

3. Statistical account balances and present values will be added to calculated portfolio equity unless, in addition to the above, the account is also designated as a contra account. Contra, statistical balance accounts will be subtracted from calculated portfolio equity.

Mortgage Servicing Rights example: Version 5.2.5.11 and later Create a Statistical Balance account in a memo category and assign the category in the Accounts Dictionary to ‘OBS Market Value Category’. Enter the servicing percent in the Cash Flow Characteristics section of the account and the actual downloaded cash flows. The cash flows servicing percent can be hand entered or download from the loan file with the system field ‘Servicing’. Version 5.2.3.5 and prior Create two Statistical Balance accounts in the OBS Market Value Category. Both accounts will include the downloaded cash flows of the serviced mortgages: one will have the actual cash flows, while the second will have the actual cash flows with negative balances and the rates lower by the amount of the servicing. The net of the two market values will be the market value of the servicing rights. To change the balance to a negative add the following transformation on the individual account: EUtil.ColumnValue(“Balance”) * -1. To adjust the interest rate add the following transformation on the individual account: EUtil.ColumnValue(“Rate”) - .25, where .25 is the servicing income.


7.e. Market Value Overrides The model allows two different methods to override Market Values for all I&P rate shocks/rate changes. These methods are ‘Use Account Overrides’ and ‘Use Instrument Overrides’, they can be found by clicking the ‘Market Value’ tab of the ‘Update Account Assumptions’ screen. Once on that screen click the drop down box named ‘Use Method:’ and use either choice based on which of the following descriptions would best meet your needs. When ‘Use Account Overrides’ is chosen, the user must click the ‘Enter Override Values’ link, found directly below the ‘Use Method:’ box, this will open the ‘Market Value Overrides’ popup box. Simply input the values for each of the ‘Rate Change’ values from -300 to 300. You may insert rows for -400 and 400, or delete unneeded rows. When finished, click ‘Done’ and your overrides will appear in a Market Value report, the entire Account will report the ‘Value’ the model finds on this screen, no matter how many instruments or actual individual ‘Cash Flows’ make up this Account. When ‘Use Instrument Overrides’ is chosen, the user must exit the ‘Account Assumptions’ screen and go to ‘Manage Chart’. Next, click the ‘Cash Flows’ tab and go to the end of that screen by moving the scroll bar all the way to the right. Place a check in the ‘Mkt Value’ section and click the ‘X’ in the ‘Edit’ column. The same popup box will appear. Once again, simply input the values for each ‘Rate Change’, the only difference is that these overrides are instrument specific and the process will need to be duplicated for each and every instrument that appears on the cash flow screen. When all overrides are input, the cumulative amount of the overrides for each instrument in each ‘Rate Change’ will appear in a Market Value report.


Appendix – Non Maturing Deposits Oh no! Did someone say Market Value. As you remember, Interest Rate Risk is the

potential for loss of future income or economic value of equity caused by an adverse change in interest rates. If we go one step further, economic value analysis estimates the sensitivity or change in the economic value of equity to changes in interest rates. Depending on your charter this is Economic Value of Equity (EVE), Net Portfolio Value (NPV) or Net Economic Value (NEV) sensitivity. Economic value analysis uses the full cash flows of a financial institution’s assets and liabilities, so it summarizes information about expected earnings over a much longer time period that a 12 to 36 month income forecast. Thus, it is a measure of your long-term earnings capacity.

Because the future is uncertain, no one can consistently predict the future course of interest rates. The correct message from this analysis is that you should position your risk exposures so you can live with a wide range of interest rate changes. Modeling starts with net present value (NPV), which is the current value of all cash flows received in the future; calculated by discounting the cash flows at an appropriate interest rate, which is referred to as the discount rate.

The process of measuring IRR with economic value analysis is relatively straightforward. First determine the economic value of most assets and liabilities, at current market rates, which involves discounting the cash flows to estimate a net present value. The net present value of the assets and liabilities are then used to calculate the economic value of equity. Finally the same process is calculated again with increased or decreased market rates. Interest rate risk is measured by the change in the economic value of equity from the current rate to each of the increased or decreased market rates.

Running a current market value report is like the end of period balance sheet. The economic value analysis measures the risk from assets and liabilities currently on the balance sheet. Economic value of equity is a measure of a financial institution’s earning capacity and the analysis tries to determine how its earning capacity is affected by changes in interest rates. What this means is a current market value report is a static report and does not take into account any future new money or alternative forecast strategies.

With some of the basics understood let’s look at our responsibility in Foresight to model non-maturing deposits.

1. Cash Flows Since non-maturing deposits have no actual maturity dates, the model will need to calculate the duration or life of these accounts (i.e. an artificial maturity). This is accomplished by using a decay table provided by the Office of Thrift and Supervision (OTS), McGuire Performance Solutions (National Averages) or User Supplied. The OTS decay rates are updated quarterly and are loaded into the model on the Prepayment/Decay Tables page. McGuire Performance Solutions (National Averages) are supplied to clients for a small quarterly fee. To verify that the proper decay rate table is assigned to the non-maturing deposit go to Manage Chart of Accounts screen for an account and view the table selected on the Characteristics tab under the Cash Flow Characteristics section. The Prepayment/Decay Tables screen shows the detail on the decay tables by type of non-maturing deposit and rate changes. Viewing a GAP report will show you the spreading of the NMD balances used for the present value calculation in Market Value. 2. Pricing Before market value is calculated the Pricing tab of the Update Account Assumptions screen must also be properly set. This setting is essential in the market value calculation because rate shocks not only adjust the market rate immediately, but also the non-maturing deposit rate. For most, when market rates move there is a lagging effect


in our NMD rates. To verify this setting view the Lag Periods and the Apply% rates in Increases and Decreases of market rates. The OTS does give us guidance in the setting for Reaction to Rate Shocks. They are Savings 33%, NOW 25% and MMDA 50% with each lag taking 6 months. 3. Discount Rate Driver Next we need to use a market curve to discount these future cash flows produced by the decay schedule. On the Market Value tab of the Update Account Assumptions screen you will find a Browse button next to the Discount Rate Driver selection. The driver we will use is the Average FHLB Advance curve since this is an alternative source of funding (i.e. market rate). There will be no additional Offset with this driver; therefore the Offset should be zero. 4. Cost Adjustment You also have to consider and adjust for the cost of servicing these non-maturing deposits. Besides the interest cost these deposits also have additional cost such as: mailing, supplies, personnel, hardware etc. The OTS does provide average servicing cost for NMD’s. As of this writing they are: NIB-Deposits 2.57%, Transaction Accounts (NOW) 1.80%, MMDA’s .86%, Passbook Accounts 1.39% and Certificates of Deposit’s .20%. These cost adjustments are entered on the Market Value tab for each non-maturing deposit.

To make some sense out of these steps lets review three rate environment examples: Dollar amounts at the one year time frame using the OTS decay schedule. Remember MMDA’s have a 50% lag to market rates. Flat Rates (i.e. current rates) Cost of Funds – MMDA Market Driver Rates Current Rate 3.00 Avg FHLB rate 5.00 Cost Adjustment .86 3.86% 5.00% +100bp Rate Shock Cost of Funds – MMDA Market Driver Rates Current Rate 3.50 Avg FHLB rate 6.00 Cost Adjustment .86 4.36% 6.00% -100bp Rate Shock Cost of Funds – MMDA Market Driver Rates Current Rate 2.50 Avg FHLB rate 4.00 Cost Adjustment .86 3.36% 4.00% Looking only at the above examples let’s analyze the effect on our Economic Value of Equity.

In flat rates since our cost are less than market rates the market value of the deposit will be less than book. Remember this is good, as the value of the liability decreases the EVE increases.

With the +100bp rate shock we will increase our EVE even more because the difference between market rates and our rate increases due to the effect of the lag in rates. This is because the market rate increases a full 100bps but our rate increases only 50pbs because of the 50% lag on the account.


In the -100bp rate shock we are still receiving value from the MMDA however the value will be getting closer to book because of the narrowing spread between our rate and market rates. This is because the market rate decreases a full 100bps but our rate decreases only 50bps thereby narrowing the gap between market rates are our rate.

Sample MMDA showing decay table selection.

Sample MMDA showing how the lag is set (both Increases and decreases are the same).


Sample MMDA showing how the MV tab is set with the driver and cost adjustment.


Chapter 8 Foresight Reports

8.a. Overview There are several areas in the model from which the user can produce reports and charts: External Assumptions - Rates screen, Edit Rate Environments screen and the Select a Report Task screen. From the All Tasks screen, the user can get to the Select a Report Task screen by selecting the Reports option. Most of the charts and reports the user will use will be run from the Select a Report Task screen (shown below). In this chapter we will show the user how to run existing reports, how to modify existing reports and how to create new reports. We will also outline the different methods to export/display the reports. The Select a Report Task screen can be separated into two distinct sections. On the left the user can view the existing Reports and the Report Sets in which they are included. On the right the screen provides the user with display option tabs to run Reports, Report Sets and Charts. There are additional report options on the left column under Other Tasks that relate to the reporting function of the model. These will also be discussed in this chapter. Before the discussion of running, modifying, creating or use of Reports/Report Sets/Charts, it is important to understand the differences between the following.

Report: a report is a stand alone document like a Current Balance Sheet or a Forecast Yield Report. All Reports, even the default reports sent with the model, are well-defined. Report Sets: a Report Set is a user–defined combination of Reports. A Report Set allows the user to quickly identify the “right” reports for specific events. For example, the user can create a set of reports for a Board Meeting, a set of reports for a ALCO Meeting, a set of reports for Regulators, a set of Reports for monthly reconciliation etc. Charts: Charts and Graphs can be run of most individual reports.

8.b. View an Existing Report or Chart To view an existing report, the user needs to highlight the single report in the Report Set box on the left. When the user selects the Screen option on the Run tab the report will be created and displayed. The Select a Report Task window allows the user to specify a Forecast/Budget, rate environment (or rate environment set) and dates for a report run. Not all of reports will require the user to choose Forecast/Budget/rate combinations (i.e. history reports & current balance sheets). The changes that can be made to customize the report are listed below. When the user has set the report selections to their specifications and saved those changes you will simply go to the Run tab and select where you want the report to go.


‘Report Type’ tab Report Type and Report Format: this is the settings that tell the model what the underlying structure of the report is i.e. balance sheet verses income statement verses ratio report etc. These selections are used to create new reports and should not be changed on existing reports. Display: a typical change would be from actual to average i.e. forecast balance sheet with average balances. Report Title: this is the actual title of the report that will show on the report once it is run. Year To Date: to be used to produce a forecast income statement with a cumulative YTD income by column. Common Size: to be used to produce a balance sheet with each line as a percent of total assets. Show Summary Column: when checked this will add a column to the end of the report which adds the entire row of numbers. Don’t Check for changed data (faster): with this box checked the model will simply open the selected Report with the data from the previous calculation. This will save the user time, but it should not be used if the user has changed any part of their model since the last report calculation. Without recalculating, the new changes will not be evident in the reports as the calculation for the reports were done PRIOR to the changes. Always recalculate all data: with this box checked the model will calculate everything (forecast and/or history) as if it never did any calculations in the past. **The model recognizes, with no boxes selected, that it must recalculate only the accounts which have had changes since the last report calculation. This is the recommended setting i.e. neither of the first two selections checked.** Apply FAS115 calculations: if the user selects this option with a Forecast report, FORESIGHT will calculate all the accounts that are set as Available for Sale in the FASB Designation in the accounts Characteristics screen. If an institution has AFS accounts there will be the following accounts: FASB115 Gain/Loss (asset), FASB115 Tax Payable (liability), FASB115 Net Gain/Loss (capital). ‘Data Source & Dates’ tab Forecast: select the forecast(s) that you will use to produce the report. Rate Environment: select the forecasted rate environment or set depending on the report you are running. Report Dates: select the starting date and number of periods to show on the report. Verify the ending date. Time Period Size: select from Month, Quarter or Year. ‘Formatting’ tab This tab allows you to customize your report the way you wish to display the numbers. Use dollar signs: will place a dollar sign in front of all numbers. Use commas: will place a comma in the appropriate place in all numbers. Negative numbers in red: all negative numbers will be displayed in the color red. Negative numbers in brackets: all negative numbers will be displayed in brackets. Negative number cells in red: all negative numbers will have the cell highlighted in red. Show History Forecast Line: with this checked the last history column and first forecast column will be labeled. Show subtotal underlines: all category totals will be underlined on the report. Drop Least Significant Digits On Numbers Too Wide: if a number is to large to all fit in a column some digits will be dropped, starting with the cents (a * will be placed at the end to show you which numbers are affected). Show Balance Sheet Out Of Balance Warning: if the balance sheet is not in balance a line will be added to the end of the report stating the out of balance amount. Show Multi Entity Columnar Cross Foot Warning: shows out of balance in the Holding Co. module. Use alternating green bar: select whether you want green bar or plain background. Show Percent Signs for Rates: check if you would like a percent sign after rates on a report. Don’t Show Percent Signs for Percentages: check if you do not want percent signs after percentages. Values in: select dollars, thousands or millions Number of Decimals: this option allows you to change the number of decimals that will show on the report in regards to dollars, percentages and rates.


Display in landscape only: if the reports size allows portrait, it will produce the report in that automatically. If this option is checked it will always be in landscape. Drop date and time printed: if this is check the data will not print in the lower right of the report. Drop Foresight version number: if this is check the data will not print in the lower left of the report. Watermark: whatever is typed in this area will appear as a watermark on the report. Watermark Color: select the color of the watermark entered on the Formatting tab. ‘Fonts & Sizing’ tab Scale report columns to fit: with this checked the model will change the column width to fit contents. Show Inactive Accounts: this will include accounts marked as inactive. Font selection: allows the user to select the font style. Row Title Size: allows the user to change the left column width i.e. chart of accounts column. Left / Right Border Size: used to change the size of the borders. Font Size selection: allows the user to select the font size. ‘Run’ tab Description: this short description about the report shows on the bottom left column of the report screen when you hover over the report. Report Set: displays the set that the report is in. New Report Set: when you create a new report this selection will allow you to create a user defined set. Save to Set: when you make changes to a system report this will save those changes until the next version update. Reporting Depth (0=All): allows you to run a summary report with the levels in you chart of accounts. Chart: select if you wish to produce a chart or graph of the report. After the numbers are calculated the Select Data for Chart screen will display. Export: this option will create a csv document of the report. To PDF File: this option will create a pdf document of the report. Print: this option sends the report directly to your default printer. Screen: this option displays the report on the screen which then allows the user to print, create a pdf or csv file. 8.b.i. Report Tool Bar Features

1. To scroll through the report’s pages, the user can click on the arrows in the upper right of the report.

2. To change the viewing size uncheck the Auto Zoom and adjust the Zoom toggle up or down.

3. To Print select the appropriate button in the upper left of the report and the printer selection window will open.

4. To create a PDF or CSV file select the appropriate button in the upper left of the report and the Select File Path window will apprear to select where you would like to save the report and by what name.


8.b.ii. Graph Features 1. Most reports have underlying graphs and charts available. To produce a chart or graph, the user

should select the report and click on the Chart selection on the Run tab. The Select Data for Chart window opens so the user can select the row and column items (x and y axis) to be displayed on the graph. After the information has been selected click on the orange Graph button in the lower left of the set up screen.

2. With the graph produced the user can print or copy and paste a chart into another Microsoft application.

3. The layout can be adjusted by using the Gallery icon (lets you change the graph style from line to bar for example).

4. The Palette Selector icon can be used to adjust color of the lines and background.

5. The user can shift from 2D to 3D views of their graph and rotate the X and Y axis.

6. The user can adjust the display background of a chart by adding grids etc., data points and legends to the graph by using the Axes Settings, Point Labels, and Data Grid icons.

7. The user can change the tile lines or the placement of the legend box information by right clicking on the item in the chart and making the appropriate selection.

8.c. Change Features of an Existing Report or Chart To edit/change an existing Farin Foresight System Report settings, the user will highlight the report and adjust the items in the five tabs on the right of the screen (i.e. ‘Report Type’, Data Source & Dates’ etc.). If you wish to save these changes click on the ‘Save to Set’ box on the ‘Run’ tab. These changes will be saved on the system reports until the next version of the software is received. To create a user defined report in which the settings will not be changed by any version updates see creating a new report below. 8.d. Creating a New Report Creating a new report requires three steps. The first step is to choose an existing report to copy, then define the changes on the selected tabs (Report Type, Data Source & Dates etc.) as discussed above and finally save the report. To insert a user defined report highlight an existing report you would like to copy. Complete the following steps:

1. On the ‘Report Type’ tab enter the ‘Report Title’.

2. On the ‘Run’ tab enter the ‘Description’.

3. On the ‘Run’ tab select the ‘Report Set’ you would like the report to be placed in. If you would like to create a new set choose ‘New Report Set’ from the drop down box and give the set an appropriate


name on the following ‘New Report Set’ line. 4. Click on the “Save to Set’ option on the ‘Run’ tab and the new report will appear in the list.

Remember when you make changes in the five tabs to customize the report that you should save those changes with the ‘Save to Set’ button. User defined reports will be highlighted in green and will not change in future release.

8.e. Export a Report to a File The model is designed to allow the user to export directly (without viewing) or from the report that is produced and viewed on the screen, to files of varying formats: CSV or PDF. This option allows the user to quickly transform reports into the most useful format. Remember reports can be exported into various formats directly from the report itself. We recommend exporting using the ‘Export’ or ‘To PDF File’ selection on the Run tab only after you have viewed the report and verified its accuracy . 8.f. Print an Existing Report or Chart Though reports can be printed directly from the report itself, FORESIGHT provides the user with the ability to send directly to the printer with the ‘Print’ option on the ‘Run’ tab. This may save the user a step in the report process. However, the user may want to view the report first for accuracy and print the report from the report screen. 8.g. Remove a Report, Chart or Report Set User defined reports that are no longer appropriate can be removed from the list. As outlined in Chapter 2, there is no undo. Prior to deleting reports/charts or report sets, the user will want to make a backup of their model. To delete a report, highlight it and press the Delete button on your keyboard. 8.h. Edit Accounts Dictionary (on side bar) The Chart of Accounts is 100% user defined. With this flexibility comes the responsibility to define in the model which accounts will perfom different accounting tasks. For example, what combination of accounts and categories equal total assets? Which represent total deposits? What represents net income? During normal operations, the user will not need to adjust their Dictionary Definitions. However, if the user is making changes to their chart, they may be required to adjust/modify the dictionary. It is a good rule to contact a Farin support member prior to adjusting this dictionary. 8.i. Edit Policy Limits (on side bar) Policy limits may appear on reports by using the Edit Policy Limits function. When the user chooses this option a number of pre-defined limits appear that may be listed on the reports. Select the appropriate limit description and select the Memo account that has the limits targeted. See 8.l. and 8.m. for examples.


8.j. Calculated Rows (on side bar) The user can review calculations used on reports, i.e. ratios. Select a category, ratio and the formula appears on the right side of the screen. 8.k. Account Database Report – B/S The user can produce an Accounts Report showing the current settings in the model for items in the Chart of Accounts. Select the Account Database Report – Balance Sheet under the Database Reports category. If you would like your model assumptions on the report select the appropriate Forecast on the ‘Data Source & Dates’ tab and check/uncheck the selections you wish to be on the report. The report will list basic characteristic information, instrument count and balances from selected dates along with target, pricing and valuation assumptions. For more detailed characteristics you may create and run a Account Audit Report (Report Type).

8.l. Net Income Limits Report Recent versions introduced new reports for net income and net interest margin variance compared to policy limits. These reports will show change in net income for a forecast relative to the flat rate environment when run with a rate environment set. To use these reports, you must also designate policy limit settings so the report can also show the variance from those settings. In each case the policy limit is expressed as the maximum percentage change in net income in a shock rate environment (usually a negative change).


To setup this report before its first use:

1. Setup a statistical account (of type statistical balance) to hold the policy limits. There are two of them for these reports, a net income change limit and a net interest margin change limit. It does not matter what you name the accounts - we will match them up to the report in a later step.

2. If the policy limit does not vary by rate environment, you can just set a simple balance target and plug in the limit. Remember that when statistical accounts are used in a report, we do use the forecasted balance amount, not any historical balances (unless the forecast method relies on history).

3. If the policy limit does vary by rate environment, you will want to create a formula target that specifies the values for each of the rate environments. Sample formulas for setting targets can be found in chapter six or use the following example: (PE.RATECHANGEGRIDMINUSPLUS400(1,-25,-25,-20,-20,0,-20,-20,-25,-25).

4. Lastly we need to match up/assign the accounts in the Dictionary. Select the Edit Accounts Dictionary on the Select a Report Task and assign your two new accounts to the proper system accounts.

Then each time you run this report, it will check those statistical accounts for the limit settings for the forecast and rate environments selected. If you run the report without setting the limits accounts as described in step 4 above, you will see the following message appear, indicating that it cannot find a statistical account to get the limit settings from:

8.m. ROA Limits Report

1. Setup a statistical account (of type statistical balance) to hold the policy limits. An example of the title might be "ROA Limit".

2. Next you will want to create a formula target that specifies the values for each of the rate environments. A sample formula for setting targets might be (PE.RATECHANGEGRIDMINUSPLUS400(1,.60,.60,.70,.70,.90,.90,.90,.80,.80).

3. Lastly we need to match up/assign the account in the Dictionary. Select the Edit Accounts Dictionary on the Select a Report Task screen and assign your new account to the proper system account.

8.n. Market Value Limits

1. Setup two statistical accounts (of type statistical balance) to hold the limits. i.e. Board Limits and Minimum Post Shock Capital Ratio.

2. Next create a formula target that specifies the values for each of the rate environments. A sample formula for setting targets might be PE.RATECHANGEGRIDMINUSPLUS300(1,6,6,6,6,6,6,6).

3. Lastly assign the accounts in the Dictionary. Select the Edit Accounts Dictionary on the Select a Report Task screen and assign your new accounts to the proper system accounts. The MV reports will then add the necessary lines at the bottom and use the values it looks up in the accounts that are described above.


8.o. Liquidity Report Adjustments 1. Setup three Statistical accounts (of type statistical flow without interest bearing or interest sensitive

checked) which you will be able to set target adjustments. i.e. Adjustment to Inflows, Adjustment to Outflows and External Funding Facility.

2. Next set targets for each of the accounts. Examples of use could be formula (example below), balance or related account targeting. - PE.RATECHANGEGRIDMINUSPLUS300(1, PE.ACCOUNT(PE.BSA("Total Deposits"), AmortizedCost, 0) * 0.10, PE.ACCOUNT(PE.BSA("Total Deposits"), AmortizedCost, 0) * 0.12, PE.ACCOUNT(PE.BSA("Total Deposits"), AmortizedCost, 0) * 0.14, PE.ACCOUNT(PE.BSA("Total Deposits"), AmortizedCost, 0) * 0.20, PE.ACCOUNT(PE.BSA("Total Deposits"), AmortizedCost, 0) * 0.30, PE.ACCOUNT(PE.BSA("Total Deposits"), AmortizedCost, 0) * 0.35, PE.ACCOUNT(PE.BSA("Total Deposits"), AmortizedCost, 0) * 0.40)

3. Lastly assign the accounts in the dictionary. Select the Edit Accounts Dictionary on the Select a Report Task screen and assign your new accounts to the proper system accounts. The Liquidity Report report will then add the necessary lines at the bottom and use the values it looks up in the accounts that are described above.

8.p. Forecast Accuracy Reports The easiest way to create these reports is to copy the Forecast Sources and Uses Report and change the Report Type to ‘Forecast Accuracy Reports’. Then the Display can be changed to: 1. Market Rates: compares current months ending market rates to last months rates using selected rate

environments (i.e. what you forecasted market rates to be the next month). 2. Prepayments: first you will need to bring into the model total credits for your loans either from your

general ledger or loan download files. This report will than compare the forecasted amortization and prepays to actual credits and display actual prepays in selected rate environments.

3. Pricing: compares your weighted ending rate to the forecasted ending rate using a preserved forecast. 4. Targets: compares your actual ending balance to your forecasted ending balance using a preserved

forecast.

foresight user’s manual table of contents - farin user’s manual table of contents 5.2.7.6 1....

Documents