Credit Market Imperfections, CreditFrictions and Financial Crises

Instructor: Dmytro Hryshko

1 / 23

Outline

Credit Market Imperfections and Consumption.

Asymmetric Information and the Financial Crisis.

Limited Commitment and the Financial Crisis: Collateral.

2 / 23

Credit Market Imperfections and Consumption

Assume that lenders can lend at a lower interest rate, r1,than the one faced by borrowers, r2 (e.g., a higher interestrate as compensation for a bank’s credit risks).

The government borrows and lends at the interest rate thatlenders face, r1.

This implies that Ricardian equivalence does not hold, ingeneral.

3 / 23

Budget constraints

Current-period budget constraint :

c+ s = y − t (1)

Future-period budget constraint :

c′ = y′ − t′ + s(1 + r1) if lender, s ≥0 (2)

c′ = y′ − t′ + s(1 + r2) if borrower, s ≤0 (3)

Equations (2)–(3) can be used to find s, for a saver and aborrower respectively, and then plugged into equation (1) toobtain the lifetime budget constraints for the saver and theborrower.

c+c′

1 + r1= y +

y′

1 + r1− t− t′

1 + r1= we1 if c < y − t

c+c′

1 + r2= y +

y′

1 + r2− t− t′

1 + r2= we2 if c >= y − t

4 / 23

A consumer with different lending and borrowing rates

The consumer’s budgetline is AEF: segmentAE applies if consumeris a lender, andsegment EF if aborrower.

5 / 23

Effects of a tax cut for a consumer with differentlending and borrowing rates

The consumer receives a

current tax cut, with a

wealth-neutral future

increase in taxes; this

shifts the budget

constraint from AE1B to

AE2F. The consumer’s

optimal consumption

bundle shifts from E1 to

E2, and the consumer

spends the entire tax cut.

6 / 23

Lessons

The government is effectively making a low-interest loan(∆t at the rate r1) available to a consumer through atax-cut scheme, which the consumer would willingly take

This is very different from the case with no credit marketimperfections, where the consumer will save the entire taxcut to pay higher future taxes

To the extent that credit market imperfections areimportant in practice, there can be beneficial effects ofpositive government debt

7 / 23

Credit Market Imperfections and the Financial Crisis

Two key credit market frictions: asymmetric informationand limited commitment

Asymmetric information: Would-be borrowers know moreabout their characteristics than do lenders

Limited Commitment: Borrowers may choose todefault—lender can overcome limited commitment withcollateral (e.g., auto loans, mortgages)

8 / 23

Asymmetric Information and the Financial crisis

Asymmetric information may give rise to kinked budgetconstraints

Quality of information in credit markets declinedsignificantly during 2008, interest rate spreads went up,lending and aggregate activity went down

9 / 23

Asymmetric information and interest rate spreads

The difference between the interest rates on prime short-termcorporate paper and short-term Government of Canada debt.The spread was particularly high during the 1974–1975 and2008–2009 recessions.

10 / 23

Asymmetric Information in Credit Markets. A model

Market structure: banks, depositors, and borrowers—goodand bad

Lending carried out through banks, which take depositsand loan them out

Deposit rate at banks is r1, loan rate is r2: r2 > r1

Fraction a of borrowers never defaults, fraction 1− aalways defaults—bank cannot tell the good borrowers fromthe bad ones

All good borrowers identical, borrow the amount L

Bad borrowers mimic the good ones, borrowing the sameamount L

Total amount of deposits L

Banks earn zero profit in equilibrium

11 / 23

Bank’s profit

π = aL(1 + r2)− L(1 + r1) = L[a(1 + r2)− (1 + r1)] = 0.

It follows that

r2 =1 + r1a− 1 =

r1a︸︷︷︸

>r1 if a < 1

+

1

a− 1︸ ︷︷ ︸

>0 if a < 1

.There is a default premium, r2 > r1, when a < 1.

The default premium increases as a decreases. How does itaffect the budget constraint?

12 / 23

Reduction in Quantity of Creditworthy Borrowers, a ↓

During the financial crisis, the average borrower was perceivedto be more likely to default, interest rate spreads increased,lending decreased and current consumption expenditures fell

13 / 23

Effect of a Decrease in the Fraction of CreditworthyBorrowers

Default premium increases—even good borrowers facehigher loan rates

Budget constraint shifts in

Consumption falls for all borrowers

Matches observations from the financial crisis—increase incredit market uncertainty, reduction in lending, decrease inconsumption expenditures

14 / 23

Limited Commitment and the Financial Crisis

Borrowers need incentives not to default on theirdebts—these incentives are typically provided by collateralrequirements

Examples: House is collateral for a mortgage loan, car iscollateral for a car loan

Can be potentially important for macro: a decline incollateral value will lower the quantity of lending and willlead to a drop in current aggregate consumption

15 / 23

Example

H = quantity of housing owned by consumer

p = price of housing

Assume that housing is illiquid—cannot be sold in thecurrent period. However, it is possible to borrow againsthousing wealth, with a collateral constraint

16 / 23

Consumer’s Constraints

Lifetime budget constraint:

c+c′

1 + r= y − t+

y′ − t′ + pH

1 + r= we

Collateral constraint:

−s(1 + r) ≤ pH ⇔ −s ≤ pH

1 + r

Since c+ s = y − t, we can write

c = y − t− s ≤ y − t+pH

1 + r.

What happens if the value of the collateral falls, that is, if p ↓?

17 / 23

Limited commitment with a collateral constraint

Initially the budget constraint is ABD and it shifts to FGH with a

decrease in the price of collateral. An unconstrained consumer will

choose first a bundle of consumption on segment AB, and then on

segment FG smoothing out the fall in her wealth by cutting both

current and future consumption. For a constrained consumer, this

causes no change in future consumption but current consumption

drops by the same amount as the decrease in the value of the

collateral since for her c = y − t+ pH1+r . 18 / 23

Implications for the recent financial crisis?

The price of housing in the US declined by about 33% fromits peak in April 2006 to November 2011

As a large fraction of consumer expenditures has beenfinanced by mortgage debt a 33% drop in the value ofcollateralizable wealth can have large effects on themacroeconomy

19 / 23

House price growth and debt 2133MiAn And Sufi: HoME Equity–BASEd BoRRowingVoL. 101 no. 5

of a link, changes in house prices and homeowner borrowing may be jointly determined by an omitted variable such as a shock to expected income growth (Orazio P. Attanasio and Guglielmo Weber 1994; John Muellbauer and Anthony Murphy 1997). As a result, proper identification of the effect of house prices on bor-rowing requires an exogenous source of variation in house price growth.

Figure 1. Aggregate US Leverage and House Price Patterns

notes: This figure presents aggregate US leverage and house price patterns. Aggregate debt information comes from the Federal Reserve flow of funds data, aggregate income comes from National Income and Product Accounts (NIPA), and aggregate house price index data come from Office of Federal Housing Enterprise Oversight (OFHEO). In the bottom right panel, aggregate debt for 1997 homeowners comes from Equifax data where homeowners are defined to be individuals who have either an existing mortgage account with positive balance as of 1997 or a previ-ous mortgage account.  

0

5,000

10,000

15,000

$ Bi

llions

1975 1980 1985 1990 1995 2000 2005 2008Year

Households Corporations

US outstanding debt: households and corporations

3

4

5

6

7

Cor

pora

te d

ebt-t

o-in

com

e ra

tio

0.8

1

1.2

1.4

1.6

1.8

Hou

seho

ld d

ebt-t

o-in

com

e ra

tio

1975 1980 1985 1990 1995 2000 2005 2008Year

US debt-to-income ratio: households and corporations

120

140

160

180

200

220

OFH

EO h

ousi

ng p

rice

inde

x

2,500

3,000

3,500

4,000

4,500

$ Bi

llions

1998 2000 2002 2004 2006 2008Year

1997 homeowners debtHouse price index

US household debt for 1997 homeowners and house prices

Households Corporations

Source: Mian and Sufi (AER 2011). 20 / 23

increase in the household default rate of 12 percentage points and a decline inhouse prices of 40 percent from the second quarter of 2006 through thesecond quarter of 2009. In contrast, the bottom 10 percent leverage growthcounties experienced a modest increase of 3 percentage points in the defaultrate and a 10 percent increase in house prices.

Auto sales and new housing building permits reveal a similar pattern. Bythe third quarter of 2008, auto sales in the top 10 percent leverage growthcounties declined by almost 40 percent relative to 2005. In contrast, auto

Figure 1. Household Leverage and the U.S. Recession of 2007–09

0.04

0.06

0.08

0.1U

ne

mp

loym

en

t ra

te

2004q1 2005q1 2006q1 2007q1 2008q1 2009q1

Unemployment Rate

-0.05

0

0.05

0.1

An

nu

alize

d g

row

th r

ate

2004q1 2005q1 2006q1 2007q1 2008q1 2009q1

GDP Growth

0.8

1

1.2

1.4

1.6

1.8

De

bt-

to-in

co

me

ra

tio

1978 1983 1988 1993 1998 2003 2008

Household debt-to-income ratio

Note: The top panel plots the unemployment rate according to the Bureau of Labor Statistics,and the middle panel plots GDP growth from the National Income and Product Accounts (NIPA).The bottom panel plots the aggregate household-debt-to-income ratio for the United States from1977 to 2008. Household debt data come from the Federal Reserve’s Flow of Funds, incomerepresents wage and salary payments from the (NIPA).

Atif Mian and Amir Sufi

76

Source: Mian and Sufi (IMF 2010).21 / 23

GDP components during and prior the Great Recession

II. County-Level Data and Summary Statistics

The county-level data set is built from a variety of sources. Information onhousehold debt, default rates, and credit scores comes from Equifax zip codelevel aggregates. Data on house prices come from the Federal HousingFinance Agency (FHFA) MSA level house price indices, which are

Figure 4. What Components of GDP Moved First?

-1

-0.5

0

0.5

Gro

wth

sin

ce 2

005q

4

2004q3 2005q3 2006q3 2007q3 2008q3 2009q2

ResidentialStructures

Fixed Investment Growth

-0.2

-0.1

0

0.1

0.2

Gro

wth

sin

ce 2

005q

4

2004q3 2005q3 2006q3 2007q3 2008q3 2009q2

DurablesServices

Consumption Growth

-0.3

-0.2

-0.1

0

0.1

Gro

wth

sin

ce D

ecem

ber

2005

08/2004 08/2005 08/2006 08/2007 08/2008 06/2009

Motor vehiclesOther

Retail Sales Growth

Equipment and software

Nondurables

Furniture and appliances

Note: The top two panels present investment and consumption data from the National Incomeand Product Accounts. The bottom panel presents monthly retail sales data from the Department ofCommerce. Each series represents the cumulative growth rate since the fourth quarter of 2005.

HOUSEHOLD LEVERAGE AND THE RECESSION OF 2007–09

83

Source: Mian and Sufi (IMF 2010). 22 / 23

Readings

Stephen Williamson. 2013. Macroeconomics. FourthCanadian Edition. Chapter 10, pp. 317–328.

Mian and Sufi’s blog: http://houseofdebt.org/

23 / 23