FEDERAL RESERVE BANK OF S T
.
LOUIS
REV I EW
SEPTEMBER
/
OCTO B E R
,
PA R T
1 20 0 9 403
Systemic Risk and the Financial Crisis: A Primer
James Bullard, Christopher J. Neely, and David C. Wheelock
How did problems in a relatively small portion of the home mortgage market trigger the most
severe financial crisis in the United States since the Great Depression? Several developments
played a role, including the proliferation o
f complex mortgage-backed securities and derivatives
with highly opaque structures, high leverage, and inadequate risk management. These, in turn,
created systemic risk—that is, the risk that a triggering event, such as the failure of a large financial
firm, will seriously impair financial markets and harm the broader economy. This article examines
the role of systemic risk in the recent financial c
risis. S
ystemic concerns prompted the Federal
Reserve and U.S. Department of the Treasury to act to prevent the bankruptcy of several large
financial firms in 2008. The authors explain why the failures of financial firms are more likely to
pose systemic risks than the failures of nonfinancial firms and discuss possible remedies for such
risks. They conclude that the economy could benefit from reforms
that reduce systemic risks, s
uch
as the creation of an improved regime for resolving failures of large financial firms. (JEL E44, E58,
G01, G21, G28)
Federal Reserve Bank of St. Louis Review, September/October 2009, 91(5, Part 1), pp. 403-17.
International Group (AIG), and Citigroup—kept
financial markets on edge throughout much of
2008 and into 2009. The financial turmoil is
widely considered the primary cause of the eco-
nomic recession that began in late 2007.
As individual firms lurched toward collapse,
market speculation focused on which firms the
government would consider “too big” or “too
connected” to allow to fail. Why should any firm,
large or small, be protected from failure? For f
inan-
cial firms, the answer centers on systemic risk.
Systemic risk refers to the possibility that a trig-
gering event, such as the failure of an individual
firm, will seriously impair other firms or markets
and harm the broader economy.
Systemic risk concerns were at the heart of
the Federal Reserve’s decision to facilitate the
T
he financial crisis of 2008-09—the most
severe since the 1930s—had its origins
in the housing market. After several
years of rapid growth and profitability,
banks and other financial firms began to realize
significant losses on their investments in home
mortgages and related securities in the second
half of 2007. Those losses triggered a full-blown
financial crisis when banks and other lenders
suddenly demanded much higher interes
t r
ates
on loans to risky borrowers, including other
banks, and trading in many financial instruments
declined sharply. A string of failures and near-
failures of major financial institutions—including
Bear Stearns, IndyMac Federal Bank, the Federal
National Mortgage Association (Fannie Mae),
the Federal Home Loan Mortgage Corporation
(Freddie Mac), Lehman Brothers, American
James Bullard is president and chief executive officer of the Federal Reserve Bank of St. Louis. Christopher J. Neely is an assistant vice president
and economist and David C. Wheelock is a vice president and economist at the Federal Reserve Bank of St. Louis. The authors thank Richard
Anderson, Rajdeep Sengupta, and Yi Wen for comments on a previous draft of this article. Craig P. Aubuchon provided
r
esearch assistance.
©
2009, The Federal Reserve Bank of St. Louis. The views expressed in this article are those of the author(s) and do not necessarily reflect the
views of the Federal Reserve System, the Board of Governors, or the regional Federal Reserve Banks. Articles may be reprinted, reproduced,
published, distributed, displayed, and transmitted in their entirety if copyright notice, author n
ame(s), and full citation are included. A
bstracts,
synopses, and other derivative works may be made only with prior written permission of the Federal Reserve Bank of St. Louis.
acquisition of Bear Stearns by JPMorgan Chase
in March 2008 and the U.S. Department of the
Treasury’s decisions to place Fannie Mae and
Freddie Mac into conservatorship
1
and to assume
control of AIG in September 2008. Federal Reserve
Chairman Bernanke (2008b) explained the Fed’s
decision to facilitate the acquisition of Bear Stearns
as follows:
Our analyses persuaded us…that allowing Bear
Stearns to fail so a
bruptly at a time when the
financial markets were already under consider-
able stress would likely have had extremely
adverse implications for the financial system
and for the broader economy. In particular,
Bear Stearns’ failure under those circumstances
would have seriously disrupted certain key
secured funding markets and derivatives mar-
kets and possibly would have led to runs on
other financial fir
ms.
This article describes how the failure of a
single financial firm or market could endanger
the entire U.S. financial system and economy
and how this possibility influenced the response
of policymakers to the recent crisis. Further, we
explain why failures of financial institutions are
more likely to pose systemic risks than failures
of nonfinancial firms and discuss possible reme-
dies for the systemic
r
isks exposed by this par-
ticular financial crisis.
2
A BRIEF GUIDE TO THE
FINANCIAL CRISIS
We begin with a brief review of the evolution
of the financial crisis and its origins in the hous-
Bullard, Neely, Wheelock
404
SEPTEMBER
/
OCTO B E R
,
PA R T
1 200 9
FEDERAL RESERV E B A N K OF ST
.
LOUIS
RE V I EW
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
1
1
1
1
1
1
HPI/CPI (excluding shelter)
HPI/Rent
HPI/Income
Figure 1
U.S. House Prices Relative to the CPI, Rents, and Median Family Income (1995:Q1–2008:Q4)
NOTE: The house price index (HPI) shown in the figure is the S&P/Case-Shiller National Home Price Index; the consumer price index
(CPI) data exclude the shelter component of the index; the rent index is a separate component of the CPI; median family income is
an aggregated monthly series from the National Association of Realtors; and recession dates (vertical gray bars) are from the National
Bureau of Economic Research.
1
A conservatorship is a legal arrangement in which one party is
given control of another party’s legal or financial affairs. In this case,
the Federal Housing Finance Agency was appointed conservator
of Fannie Mae and Freddie Mac by the U.S. Treasury Department
in accordance with the Federal Housing Finance Regulatory Reform
Act of 2008.
2
This article is based on and extends “Systemic Risk and the
Macroeconomy” (see Bullard, 2008).
ing market to understand systemic risk in the
context of this crisis.
U.S. house prices began to rise far above his-
torical values in the late 1990s. Figure 1 shows
the growth in an index of house prices relative to
the consumer price index (CPI), an index of resi-
dential rents, and median family income, all
normalized to equal 1 in the first quarter of 1995.
House prices rose rapidly relative to consu
mer
p
rice inflation, rents, and median family income
between 1998 and 2006. Analysts attribute the
rapid growth in the demand for homes and the
associated rise in house prices to unusually low
interest rates, large capital inflows, rapid income
growth, and innovations in the mortgage market.
3
A rapid rise in the share of nonprime loans,
especially nonprime loans with unconventional
terms, was a key feature o
f the mortgage market
during the housing boom. Nonprime loans
increased from 9 percent of new mortgage origi-
nations in 2001 to 40 percent in 2006 (DiMartino
and Duca, 2007). Most nonprime mortgage loans
were made to homebuyers with weak credit his-
tories, minimal down payments, low income-to-
loan ratios, or other deficiencies that prevented
them from qualifying for a prime loan.
4
Many non-
prime loans a
lso had adjustable interest rates or
other features that kept the initial payments low
but subjected borrowers to risk if interest rates
rose or house prices declined.
The rise in nonprime loans was accompanied
by a sharp increase in the percentage of nonprime
loans that originating lenders sold to banks and
other financial institutions. The practice of selling
conventional prime mortgages has been common
s
ince the 1930s, when the federal government
established Fannie Mae to promote the flow of
capital to the mortgage market.
5
The federal gov-
ernment chartered Freddie Mac in 1970 to com-
pete with Fannie Mae, which had been sold to
private investors in 1968. Both firms purchase
large amounts of prime mortgage loans, which
they finance by selling bonds in the capital mar-
kets. Before the 1990s, Fannie Ma
e, F
reddie
Mac, and other firms rarely purchased nonprime
loans. Instead, the originating lenders held most
nonprime loans, which comprised a relatively
small portion of the mortgage market, until they
matured.
6
When a lender sells a loan rather than holding
it until maturity, the lender has less incentive to
ensure that the borrower is creditworthy. Many
analysts contend that lax underwriting standards
con
tributed t
o the high rate of nonprime loan
delinquencies.
7
Although purchasers of loans do
have an incentive to verify the creditworthiness
of borrowers, many evidently failed to appreciate
or manage the level of risk in their portfolios dur-
ing the recent housing boom (Bernanke, 2008a).
In some instances, investors may have relied
too heavily on the judgments of credit rating
agencies.
8
The banks and other financial institutions
that purchased nonprime mortgage loans typically
created residential mortgage-backed securities
(RMBSs) based on pools of mortgage loans. An
Bullard, Neely, Wheelock
FEDERAL RESERV E B A N K OF ST
.
LOUIS
REV I EW
SEPTEMBER
/
OCTO B E R
,
PA R T
1 200 9 405
3
Bernanke (2005) describes the “global saving glut” and changing
pattern of international capital flows during the 1990s and early
2000s, and Caballero, Farhi, and Gourinchas (2008) discuss the
role of capital inflows in fueling the housing boom. Taylor (2009),
by contrast, blames the housing boom primarily on loose monetary
policy during 2002-05.
4
Mortgage loans are typically classified as prime or nonprime,
depending on the risk that a borrower will default on the loan.
Nonprime loans are further distinguished between “subprime” and
“alternative-A” (Alt-A), again depending on credit risk. Generally,
borrowers qualify for prime mortgages if their credit scores are 660
or higher and the loan-to-value ratio is below 80 percent. Borrowers
with lower credit scores or other financial deficiencies, such as a
previous record of delinquency, foreclosure or bankruptcy, or
higher loan-to-value ratios, are more likely to qualify only for a
nonprime loan. See Sengupta and Emmons (2007) for more infor-
mation about nonprime mortgage lending.
5
Wheelock (2008) discusses the establishment of Fannie Mae and
other agencies and programs to alleviate home mortgage distress
during the Great Depression.
6
Fannie Mae and Freddie Mac are not permitted to purchase loans
that exceed a specific limit (currently $417,000) except in desig-
nated high-cost areas. Further, Fannie Mae and Freddie Mac require
minimum documentation and other standards on the loans they
purchase, and hence they purchase relatively few nonprime loans.
7
Demyanyk and Van Hemert (2008) and Bhardwaj and Sengupta
(2008) provide alternative perspectives on the role of lax under-
writing of nonprime loans.
8
Critics charge that the rating agencies had a conflict of interest
because bond issuers paid for the ratings (New York Times, 2007;
Fons, 2008a,b). In addition, the rating agencies used inadequate
risk models that did not account for a possibility of a serious drop
in housing prices. See Fons (2008a,b).
RMBS redistributes the income stream from the
underlying mortgage pool among bonds that differ
by the seniority of their claim. Sometimes addi-
tional securities, known as collateralized mort-
gage obligations (CMOs) or collateralized debt
obligations, are created by combining multiple
RMBSs (or parts of RMBSs) and then selling por-
tions of the income streams derived from the mort-
gage pool or RMBSs to
investors w
ith different
appetites for risk.
The securities rating agencies assigned high
ratings to many of the mortgage-related securities
created to finance purchases of nonprime loans.
As long as house prices were rising, most non-
prime loans performed well because borrowers
were usually able to refinance or sell their house—
at a higher price—if they were unable to make
their loan payments.
9
When house prices began
to fall, many borrowers found that they owed
more on their house than it was worth. This situ-
ation made it impossible for some borrowers to
repay their loan by selling their house or refinanc-
ing their mortgage, and it also created an incentive
simply to default. Consequently, loan defaults and
foreclosures rose sharply, as shown in Figure 2,
which plots data on the percentage of home mo
rt-
g
ages entering foreclosure in a given quarter and
the year-over-year percentage change in the S&P/
Case-Shiller National Home Price Index.
Rising loan delinquencies caused many RMBSs
and CMOs backed by home mortgage loans to
default, and investment banks and other investors
that held large portfolios of RMBSs and CMOs
experienced substantial losses. Ultimately, the
decline in house prices and the incre
ase i
n mort-
gage loan defaults that began in 2006 were the
root cause of the financial crisis. The following
sections explore how systemic risks caused losses
on nonprime mortgages and mortgage-related
securities to disrupt the entire financial system.
9
Most nonprime loan originations were refinances of existing
mortgages in which borrowers withdrew accumulated equity from
their homes (a phenomenon known as a “cash-out” refinance).
See Bhardwaj and Sengupta (2009).
Bullard, Neely, Wheelock
406
SEPTEMBER
/
OCTO B E R
,
PA R T
1 200 9
FEDERAL RESERV E B A N K OF ST
.
LOUIS
RE V I EW
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
–20.00
–15.00
–10.00
–5.00
0.00
5.00
10.00
15.00
20.00
Foreclosures Started
U.S. House Price Index
New Foreclosures Started (percent)
U.S. House Prices (year/year percent change)
Figure 2
U.S. House Prices and Foreclosures
NOTE: Foreclosures data are from the Mortgage Bankers Association; the house price index (HPI) is the S&P/Case-Shiller National
Home Price Index. Vertical gray bars indicate recessions.
SYSTEMIC RISK
Systemic Risk, Counterparty Risk, and
Asymmetric Information
In the recent financial crisis, the most impor-
tant type of risk to the financial system has been
“counterparty risk,” which is also known as
“default risk.”
10
Counterparty risk is the danger
that a party to a financial contract will fail to live
up to its obligations.
Counterparty risk exists in large part because
of asymmetric info
rmation. Individuals and f
irms
typically know more about their own financial
condition and prospects than do other individuals
and firms. Much of the recent concern about sys-
temic risk has focused on investment banks that
deal in complex financial contracts. Consider the
following example: Suppose Bank A purchases
an option from Bank B to hedge the risk of a
change in the term structure of interest rate
s. I
f
Bank B later fails, perhaps because of bad invest-
ments in home mortgages, then the option sold
by Bank B may lose value or even become worth-
less. Thus, Bank A—which thought it was care-
fully hedging its risk—is adversely affected by
Bank B’s problems in housing markets.
Of course, financial firms can protect them-
selves to some degree in such simple situations.
The logic of self-interested beha
vior c
ombined
with market clearing would lead to an appropriate
pricing of risk; Bank A would have considered
the possibility of the failure of Bank B and taken
this into account in its contingency plan. For
example, Bank A might require Bank B to post
collateral to protect the value of the option in case
Bank B failed. But in actual financial markets,
arrangements are so complex that the nature of
risk th
at firms face might not be obvious. In a
ddi-
tion, the value of collateral fluctuates and thus
even carefully collateralized deals are subject to
some risk.
11
Systemic Risk and Information Cascades
Sophisticated investors and counterparties
will cease to do business with a firm once the
firm’s weak condition becomes known, as they
did with Bear Stearns and Lehman Brothers. How -
ever, the inability to sort p
erfectly among good
and bad risks can lead banks and other investors
to pull away from nearly all lending during a crisis.
The tendency of lenders to seek safe investments
during a crisis explains why trading in risky assets
declined sharply and their market yields rose
relative to yields on federal government debt
during 2007-08.
Sometimes all firms in an industry are “tarred
by the same brush” and one
firm’s
failure leads
investors to shun an entire industry. For example,
before the introduction of federal deposit insur-
ance in 1933, the failure of individual banks some-
times caused the public to shift a large portion
of its funds from bank deposits into cash. Why
should the failure of a single firm cause the public
to suspect an entire industry? Again, the answer
is related to the fact that people h
ave i
mperfect
information. Because depositors lack complete
information about the condition of their bank, the
failure of one bank can trigger mass withdrawals
by depositors of other banks to avoid losses in the
event their own bank fails. Indeed, even if a par-
ticular depositor believed that his bank was fun-
damentally sound, it would still make sense for
him to withdraw his money if he thought that
wit
hdrawals b
y other depositors might cause the
bank to fail. Banking panics are especially danger-
ous because large-scale deposit withdrawals can
make bank failures more likely, as well as cause
banks to reduce their lending in an effort to boost
liquidity. Several severe banking panics during
the nineteenth and early twentieth centuries
resulted in widespread bank failures, financial
distress, and economi
c c
ontractions.
12
Federal deposit insurance has largely ended
the problem of banking panics. When IndyMac
Bank was rumored to be near failure in 2008,
Bullard, Neely, Wheelock
FEDERAL RESERV E B A N K OF ST
.
LOUIS
REV I EW
SEPTEMBER
/
OCTO B E R
,
PA R T
1 200 9 407
10
Taylor (2009) argues that the financial crisis was associated mainly
with an increase in counterparty risk and not a shortage of liquidity.
11
Kiyotaki and Moore (1997) and Pintus and Wen (2008) discuss
how procyclical fluctuations in the value of collateral can exacer-
bate financial booms and busts and contribute to macroeconomic
fluctuations.
12
Calomiris and Gorton (2000), Dwyer and Gilbert (1989), and
Wicker (2000) are among the numerous studies of the causes and
effects of U.S. banking panics. Diamond and Dybvig (1983) pro-
vide an important theoretical analysis of banking panics.
