joining dots

12
Joining the dots - Tackling the Basel II and IFRS debate IFRS – Global Reporting Revolution March 2004

Upload: widayat-wahyu

Post on 28-Sep-2015

10 views

Category:

Documents


4 download

DESCRIPTION

DOTS

TRANSCRIPT

  • Joining the dots - Tackling the Basel II and IFRS debate

    IFRS Global Reporting RevolutionMarch 2004

  • 1 | Joining the dots - Tackling the Basel II and IFRS debate PricewaterhouseCoopers

    Joining the dots - Tackling the Basel II and IFRS debate

    Welcome to the sixth in a series of papers dedicated to discussing International Financial

    Reporting Standards and the impact on the banking industry.

    This paper compares some of the synergies and differences between the Basel II and

    IFRS frameworks*. In particular, it tackles the IFRS approach to loan loss provisioning and

    the Basel II approach to calculating capital requirements.

    I hope that you find this paper thought-provoking and insightful. If you would like to

    discuss any of the issues addressed in more detail, please speak with your usual contact

    at PricewaterhouseCoopers or those listed at the end of this paper, as this helps us to

    ensure that we are addressing the issues that you are most focused on.

    Phil Rivett

    Global Leader, Banking & Capital Markets

  • 2 | Joining the dots - Tackling the Basel II and IFRS debate PricewaterhouseCoopers

    Whilst many of the requirements,and consequently the sourcedata, are extremely similar underboth approaches, banks shouldnot lose sight of the fact that theaims of IFRS and Basel II arefundamentally very different. Theobjective of IFRS is to ensurethat the financial statementsadequately reflect the lossesthat are incurred at the balancesheet date, whilst Basel IIsobjective is to ensure that thelender has sufficient provisionsor capital to support itsexpected losses over the courseof the next 12 months2 andsupport any unexpected creditlosses3. IFRS clearly states thatit is an incurred loss model;Basel II is all about expectedand unexpected losses.

    This distinction has beensomewhat confused by the factthat a number of regulators setrules for the level of provisions infinancial statements. From anaccountants perspective, it isnot a regulators primaryfunction to determine theamount of provisions a bankshould hold, its focus shouldrather be on capital levels. Theregulatory response to a beliefthat a bank should have higherprovisions should not be toimpose further provisions thatmay not comply with Accounting

    Standards but instead to requirethe banks to hold more capital.

    Expected andunexpected losses

    The decision in October 2003by the Basel Committee onBanking Supervision (theCommittee) to removeexpected losses from the riskweight functions in the InternalRatings-Based (IRB) approach(but not from the simplerapproaches) has been driven byits belief that provisions shouldreflect a banks expected creditlosses whereas capital shouldprincipally reflect any unexpectedlosses that may arise.

    However, as described later inthis paper, the IFRS accountingprovision relates strictly toincurred losses which areunlikely to be the same asexpected losses. Since theCommittee views capital asprimarily covering onlyunexpected credit losses thereis a risk of a shortfall betweenincurred and expected creditlosses which are not providedfor by either an accountingprovision or by capital (seeFigure 1 for illustration).However, this has beenrecognised by the Committee inits January 2004 press release

    1Datamonitor estimates that total spend in Europe on Basel II to exceed $6 billion by the end of 2005.

    2The Internal Ratings-Based approach is based upon a long-run average twelve month probability of default and the banks mostconservative estimate of loss given default across an economic cycle.

    3The requirement to distinguish between expected and unexpected losses applies only to those banks that have elected to use theInternal Ratings-Based approach to credit risk.

    Introduction

    There has recently been

    significant debate in the

    banking industry in

    connection with the

    crossovers and linkages

    between the IFRS

    approach to loan loss

    provisioning and the

    Basel II approach to

    calculating capital

    requirements. Many banks

    have expressed their

    desire to utilise models

    that are being developed

    under their Basel II

    implementation plans for

    the purposes of

    performing their

    provisioning calculations.

    Given the investment

    being made by the

    banking sector to achieve

    Basel II compliance1, it is

    only fitting that banks

    should investigate whether

    an IFRS compliant

    approach could be aligned

    to a Basel II methodology.

  • where it suggested that bankswill need to compare the BaselII expected loss calculation withthe total amount of provisionsthey have made. Any shortfalls(where the expected lossexceeds the total provision)must be deducted from capital(50% from Tier 1 and 50% fromTier 2) and any excesses(where the total provisionexceeds the expected loss) maybe eligible as Tier 2 capital.

    It is therefore clear that thecalculation of expected lossesis still relevant to the Basel IRBcapital calculation in order toidentify these shortfalls orexcesses. Unless a bank hasexplicitly captured expectedlosses within its future marginincome and can demonstratethis to be the case, the regulator

    will need to understand theamount of cushion that is inplace to manage expectedlosses - either within capital or aspart of provisions. In theory theregulator should not mind wherethis cushion for expected lossesis positioned - future marginincome, provision or capital - justas long as it is somewhere!

    The incurred versusexpected loss debate

    Whilst the debate aroundincurred versus expected lossmodels has intensified, therevised IAS 39 issued by theInternational AccountingStandards Board (IASB) inDecember 2003 has helped toshed light on some of thedifferences between the incurredand expected loss concepts.

    IAS 39 now clearly states thatthe provisioning model it refers to is an incurred loss model,although this still allows forimpairment provisions to beraised on portfolios of loanswhere there is observable datato suggest that there is adeterioration in the expectedcash flows from these assetssince they were initiallyrecognised. The standardprovides two examples oftriggers for this deterioration:one relates to changes ineconomic conditions and theother to changes in the paymentstatus of borrowers which, onaggregate, may often bereferred to as incurred but notreported losses (see Figure 2).

    PricewaterhouseCoopers Joining the dots - Tackling the Basel II and IFRS debate | 3

    Basel I I

    I F R SCurrent Gaap

    Basel I I

    I F R SCurrent Gaap

    TOTAL LOSSES

    Expected Unexpected

    Capital

    Capital

    Incurred

    Impairment ShortfallTIER 1/2Capital

    General

    Provision

    Provision

    Specific

    Provision

    Figure 1:

  • 4 | Joining the dots - Tackling the Basel II and IFRS debate PricewaterhouseCoopers

    Despite the recent changes inthe Basel II requirements, thedistinction between incurredand expected losses is stillrelevant - an expected loss isnot the same as an incurredone, or even an incurred butnot reported one. An incurredloss at the balance sheet dateis one where the trigger eventthat gives rise to an impairmentloss has already happenedwhereas an expected loss isone that is anticipated,irrespective of whether thetrigger event has taken place atthe balance sheet date. So if, atthe balance sheet date, a bankexpects a certain trigger eventto take place, (for example, arise in unemployment rates) itwould include theconsequences of this triggerevent (in other words, anincrease in losses) under anexpected loss model, but notunder an incurred loss model.

    Consequently there are cleardifferences between theobjectives of the two models. Basel II works on statisticalmodelling of expected losseswhile IFRS, although allowingstatistical models, requires atrigger event to have occurredbefore they can be used. IAS39 specifically states thatlosses that are expected as aresult of future events, nomatter how likely, are notrecognised. This is a clear andfundamental area of differencebetween the two frameworks.

    Although the requirements ofIAS 39 imply that conditionsneed to have changed since theexposure was granted for atrigger event to have takenplace, it introduces the conceptof incurred but not reported.This implies that banks canprovide for losses that have notyet crystallised but are likely,

    based on past experience, tohave been incurred at thebalance sheet date. Somebanks may view this as thebridge between the IFRSincurred loss provision and theBasel II capital charge, but itdoes not follow that all expectedlosses for the following year areincurred at the balance sheetdate. Under IFRS banks willhave to identify the events thathave occurred before thebalance sheet date which willcause impairment and they willrequire empirical evidence tocorrelate these events to a likelylevel of loss. Basel II alsorequires empirical evidence butdoes not require trigger eventsto be specifically identified.

    Key definitions

    In order to identify thesimilarities and differencesbetween the two approaches, it is helpful to look at thedefinitions of default underBasel II and impairment underIFRS. It is clear that a Basel IIdefault and an IFRS impairmentare similar (see Figure 3).

    an expected loss is

    not the same as an

    incurred one

    Observable data indicating there is a measurable decrease

    in cash flows since initial recognition, although decrease cannot be indentified on

    any individual asset due to:

    1. Adverse change in payment status of borrowers2. Changes in national or local economic conditions

    Actual breach of contract (e.g. default or delinquency in interest or principal payments

    Borrower granted concessions due to financial difficulties

    Probability of bankruptcy or financial

    reorganisation

    Disappearance of an active market

    in the asset

    Issuer has significant financial

    difficulty

    Objective evidence of impairment

    Figure 2:

  • PricewaterhouseCoopers Joining the dots - Tackling the Basel II and IFRS debate | 5

    Is it just a matterof timing?

    It is clear from Figure 3 that thesignificant difference betweenthe two definitions relates totiming. Basel II defines defaultas the obligor being 90 dayspast due on the obligation(expanded to 180 days for someproducts) whereas IFRS refers to

    actual breach of contract;technically one missed capital orinterest payment. On the face ofit IFRS is more conservative butin fact Basel II takes intoaccount all defaults that arelikely to occur in the next twelvemonths while IFRS onlyrecognises impairments incurredup to the balance sheet date.

    Figure 3:

    Default definition under Basel II

    It is determined that the obligor is unlikely to pay its debt obligations (principal, interest orfees) in full.

    IFRS impairment indicators

    There is objective evidence of impairment as aresult of one or more events that occurred afterthe initial recognition of the asset (a loss event)and that loss event (or events) has an impact onthe estimated future cash flows of the asset.

    A credit loss event associated with any obligation ofthe obligor, such as a charge-off, specific provisionor distressed restructuring involving the forgivenessor postponement of principal, interest or fees.

    Granting of a concession to the borrower.

    The obligor is past due more than 90 days on anycredit obligation.

    Actual breach of contract (i.e. one missed payment).

    The obligor has filed for bankruptcy or similarprotection from creditors.

    Significant financial difficulty of the borrower orprobability of bankruptcy or other financialreorganisation of the borrower.

    No specific reference to economic conditions as a default trigger, but requires different scenarios to be modelled.

    Observable data indicating there is a measurabledecrease in the estimated cash flows from a groupof assets since their initial recognition due to: adverse changes in the payment status of the

    borrowers in the group; or a deterioration in national or local economic

    conditions that correlate with defaults on theassets in the group.

  • 6 | Joining the dots - Tackling the Basel II and IFRS debate PricewaterhouseCoopers

    formula-based

    approaches or

    statistical methods

    may be used to

    determine impairment

    losses in a group of

    financial assets

    IFRS also allows an economictrigger as an indication thatimpairment may be present in agroup of assets. IAS 39.59(f)(ii)states that an adverse change innational or local economicconditions that correlates withdefaults on the assets in thegroup should be used as a basisfor determining that there is ameasurable decrease in theirestimated cash flows. Althoughthe Basel II guidance in this areais not so clearly defined, neitherregulators nor banks wish to seevolatile movements in the level ofcapital held arising from changesin economic conditions. Basel IItherefore seeks to provide astable level of capital over theeconomic cycle whilst IFRSseeks to reflect economicvolatility in provision levels. Thelevel of capital in place should beable to cover unexpected lossesin adverse economic conditions,in other words, there should beenough of a capital buffer inplace to cover losses arising onworst case scenario economicconditions and there is anexpectation from the regulatorsthat banks, especially thoseadopting the requirements ofPillar 2, will scenario and stresstest their capital assessmentsand capital allocationmechanisms to achieve this.

    Banks will therefore need todefine what they consider to bethe economic drivers that

    impact both their loss rates aswell as the probability that theircustomers will default. Thesemay include economic factorssuch as interest rates,unemployment levels andaverage national or regionalhouse prices. As part of thedata collation exercise that theywill need to undertake in orderto build their historical lossexperience for both capital andprovision calculations, theyshould also seek to factor theimpact of these economicdrivers into their calculations sothat a range of loss experiencecan eventually be built thatcovers a number of economicscenarios. This form of scenarioanalysis to reflect current andfuture economic conditions hasbeen a point of debate betweenthe industry and regulators fora number of years and there iscurrently no industry-wideapproach to stress testing.

    Synergies, similarities anddifferences

    IAS 39.AG92 appears torecognise that there will clearlybe significant synergies betweenthe two approaches by statingthat formula-based approachesor statistical methods may beused to determine impairmentlosses in a group of financialassets although it does go on tospecify that the model should

    incorporate the cash flows forall of the remaining life of theasset, (not only the next year).Further to say that it should notgive rise to an impairment losson initial recognition. It canclearly be inferred that the IASBacknowledges that a modelbased approach may be utilised,and that this model can use datacollected for Basel II purposes.However there are clearly somedefinitional differences that arelikely to prevent banks frominserting Basel II data cleanlyinto the IFRS model.

    As previously discussed, one ofthese significant differences isthe point at which an asset isconsidered to be impaired ordefaulted. Others include theprecise elements of what makesup a loss under Basel II it isdefined as economic loss andwill include the direct andindirect costs associated withcollecting on the exposure, suchas allocations of internaloverheads and other non-cashcosts. Under IFRS animpairment loss is defined as thedifference between the carryingvalue of the loan and the presentvalue of the expected cash flowsdiscounted at the effectiveinterest rate. Clearly non-cashtransactions such as latepayment charges or indirectcosts such as the overheads ofa collections department will notform part of the impairment loss

  • PricewaterhouseCoopers Joining the dots - Tackling the Basel II and IFRS debate | 7

    but would be included in a BaselII loss given default. The lossgiven default requires theinclusion of a cost of capital butIFRS specifically states that thediscount rate to be used is thesame Effective Interest Ratethat is used to recognise incomeon the asset at the outset, i.e.before it was impaired.

    Another key difference ariseswith the use of the Exposure atDefault under the Basel IIcalculation. On a financial assetwith a limit facility (for example acommitted loan facility or anoverdraft), this Exposure atDefault will take into account anexpectation of future drawdownsuntil the default event hasoccurred by utilising a creditconversion factor. Howeverunder IFRS and its incurred lossconcept, it is the loan amountoutstanding at the balancesheet date that is considered inthe calculation and not anyfuture movements anddrawdowns, as this could beconstrued as providing forfuture losses. Future drawdownsto which a bank is committedwould need to be consideredseparately under IAS 37.

    Despite these differences, thereare substantial similaritiesbetween the two models whichmean that the underlying datarequirements of the twoapproaches will also be similar.

    For example, both approachesrequire collateral to be taken intoaccount when estimating theloss that will crystallise. On firstreading, it appears that the BaselII approach requires data onlosses while the datarequirements of IFRS relate tocash flows. However the cashflows that are not received underan IFRS impairment are likely tobe very similar to the loss that iscrystallised on a Basel II default,subject to the differences notedabove. Therefore a model thatvalues cash flows not expectedto be received will equate to onethat values the cash flows stillexpected to be received. Manybanks are therefore proposing toutilise the loss data collated forthe purposes of Basel II in orderto estimate the lost cash flowsarising on a group of impairedassets. In order to use the BaselII loss data as an approximationfor the lost cash flows requiredby the IFRS model, banks mustnot lose sight of the timing ofthese lost cash flows. The Basel IIloss will not necessarily equalthe net present value of the lostcash flows.

    Conclusion

    The IFRS-incurred model maybe summarised as the expectedloss on a loan or portfolio ofloans as a result of a particulartrigger that has already occurred.Consequently, there are clearly

    similarities between the twoframeworks and banks will needto leverage as many synergiesas possible from the massivedata requirements that theimplementation of both willnecessitate. There will besignificant data overlaps and it islikely that the calculation enginegenerating the numbers willneed to be flexible enough to cope with the demands oftwo methodologies with twodifferent objectives.

    Although it would appear thatthe underlying Basel II datacould potentially, with suitableadjustments and further analysis,be used for impairmentprovisioning under IFRS, thework required and therefore thecosts involved in making theseadjustments could still besubstantial. However, the cost ofimplementing IFRS in isolationfrom Basel II would besignificantly more. As mentionedpreviously, the European bankingsector is already likely to have tospend over $6 billion to achievecompliance with Basel II itshould be doing its utmost toavoid spending another smallfortune on achieving compliancewith IFRS.

    4Datamonitor estimates total spend in Europe on Basel II to exceed $6 billion by the end of 2005.

  • PricewaterhouseCoopers

    If you would like to discuss any of the issues raised in this paper, please speak with your usual contact atPricewaterhouseCoopers or one of the contacts listed below:

    This paper was prepared by:

    Harjeet BauraSenior Manager, Banking & Capital MarketsTel: 44 20 7804 7687E-mail: [email protected]

    8 | Joining the dots - Tackling the Basel II and IFRS debate PricewaterhouseCoopers

    Etienne BorisTel: 33 1 5657 1029E-mail: [email protected]

    Lloyd BryceTel: 852 2289 2712E-mail: [email protected]

    Michael CodlingTel: 61 2 8266 3034E-mail: [email protected]

    Paul CunninghamTel: 420 251 15 2012E-mail: [email protected]

    Henry DaubeneyTel: 1 646 471 5193E-mail: [email protected]

    Burkhard EckesTel: 49 30 2636 2222E-mail: [email protected]

    Addison EverettTel: 86 10 8529 008E-mail: [email protected]

    Jeremy FosterTel: 44 20 7212 5249E-mail: [email protected]

    Simon GealyTel: 44 20 7212 3513E-mail: [email protected]

    Bernhard HeinemannTel: 41 1 630 25 77E-mail: [email protected]

    John HitchinsTel: 44 20 7804 2497E-mail: [email protected]

    Edmund HodgeonTel: 34 91 568 5180E-mail: [email protected]

    Olga KucherovaTel: 7 095 967 6371E-mail: [email protected]

    Karen LoonTel: 65 6236 3021E-mail: [email protected]

    Johan MnssonTel: 46 8 555 33044E-mail: [email protected]

    John McDonnellTel: 35 3 1 704 8559E-mail: [email protected]

    Unakorn PhruithithadaTel: 66 2 344 1134E-mail: [email protected]

    Lorenzo Pini PratoTel: 390 6 57025 2480E-mail: [email protected]

    Arno PouwTel: 31 20 400 8622E-mail: [email protected]

    Phil RivettTel: 44 20 72174686E-mail: [email protected]

    Pauline WallaceTel: 44 20 7804 1293E-mail: [email protected]

    IFRS Banking contacts

  • If you would like additional copies of this paper, please contact Kirsty Parker via e-mail at [email protected]

    Global Basel contacts

    Global Erik Musch 32 2 710 9747 [email protected]

    Pan-European Charles Ilako 32 2 710 7121 [email protected] Catherine 33 1 5657 1238 [email protected]

    Australia Peter Trout 61 2 82 66 7620 [email protected]

    Belgium Josy Steenwinckel 32 2 710 7220 [email protected]

    CEE Jim Kernan 48 22 5234 312 [email protected] Wake 36 1 461 9514 [email protected]

    France Guy Flury 33 1 5657 1067 [email protected]

    Germany Gnter Borgel 49 69 9585 2115 [email protected]

    Hong Kong Peter Li 852 2289 2982 [email protected]

    Ireland Alan Merriman 353 1 662 6599 [email protected]

    Italy Elisabetta Caldirola 390 2 778 5380 [email protected] Setola 39 349 49 22 663 [email protected]

    Luxembourg Emmanuelle Henniaux 352 49 4848 2527 [email protected] Sergiel 352 49 4848 2531 [email protected]

    Singapore Chris Matten 65 6236 3878 [email protected]

    Spain Jos Luis Lpez Rodriguez 34 91 568 4400 [email protected]

    Sweden Gran Raspe 46 8 555 330 59 [email protected] Lundblad 46 8 555 336 01 [email protected]

    Switzerland Pascal Portmann 41 1 630 2420 [email protected] Pernollet 41 22 748 5440 [email protected]

    The Netherlands Arno Pouw 31 20 400 8622 [email protected] Mars 31 20 568 4537 [email protected]

    UK John Tattersall 44 20 7212 4689 [email protected] Smith 44 20 7213 4705 [email protected]

    US Bill Lewis 1 202 414 4339 [email protected]

    Copyright 2004 PricewaterhouseCoopers. All rights reserved. PricewaterhouseCoopers refers to the network of member firms ofPricewaterhouseCoopers International Limited, each of which is a separate and independent legal entity. Connected Thinking is a registered trademarkof PricewaterhouseCoopers. Designed by studio ec4 (16401 03/04).

  • www.pwc.com