Financial Instability: Theories and Applications

1
Financial InstabilityTheories and Applications

• Jing Yang
• Bank of England

2
The views I express in this presentation are my
own, and do not necessarily represent those of
the Bank of England. The Bank of England does
not accept any liability for misleading or
inaccurate information or omissions in the
information provided.
Financial InstabilityTheories and Applications
3
Bank of England --1997 reform

• Monetary Policy
• -- price stability
• -- output
• MPC monthly
• Inflation Report

• Financial Stability
• -- to prevent systemic risk
• -- publication
• Financial Stability Review

4
Financial Stability A Working Definition

• Financial stability is the condition where the
  financial system is able to withstand shocks
  without impairing the allocation of savings to
  investment opportunities in the economy.

5
Financial Instability in Pictures
The Financial System
The Real Economy
6
Financial Instability in Pictures
The Financial System
The Real Economy
7
Financial Instability in Pictures
The Financial System
The Real Economy
8
Financial Instability in Pictures
The Financial System
The Real Economy
9
Financial Instability in Pictures
The Financial System
The Real Economy
10
The Financial Instability Avalanche

• The Anatomy of a Crisis
• An (endogenous or exogenous) shock hits the
  banking system
• The shock wipes out an initial set of (inherently
  fragile) banks
• The first wave of failures creates a
  chain-reaction among otherwise healthy banks
  (contagion), creating a second wave of failures,
  etc., etc.
• When the dust clears the banking/financial system
  is in melt-down
• The real economy suffers
• Let us explore each step of the chain

11
Informative Surveys of Financial Stability

• A number of excellent surveys of the topic can be
  found
• De Brandt and Hartmann 2000, Systemic Risk A
  Survey, ECB Working Paper 35
• Summer 2002, Banking Regulation and Systemic
  Risk, Bank of Austria Working Paper 57
• Kaufman and Scott 2002, What is Systemic Risk
  and Do Bank Regulators Retard or Contribute to
  It? Loyola University (Chicago) Working Paper
• More definitions of Financial Stability than you
  could possibly want
• Schinasi 2004, Defining Financial Stability,
  IMF Working Paper WP/04/187

12
I. Source of the Fragility of BanksII.
Channels of ContagionIII. Cost of Banking
Crisis
The Financial Instability
13
I. Source of the Fragility of Banks
14
The Fragility of the Banking System

• The Source of Fragility
• Banks take in deposits callable on demand
• but extend (relatively) long term loans
• This process is known as Maturity Transformation
• So, if every depositor wants his money back at
  the same time, it wont be at the bank
• Why do banks do this? The seminal model
• Diamond and Dybvig 1983, Bank Runs, Deposit
  Insurance and Liquidity, Journal of Political
  Economy

15
The Diamond-Dybvig Model

• The Structure of the Model
• The world consists of three periods (t1, t2, t3)
• People wish to consume in t3, but may have to
  consume in t2
• In t1 the bank and depositors can invest in
• a long term technology with a high payoff in t3
  but a negative return if liquidated in t2
• a short term liquid technology with a low
  positive return that can easily be transformed
  into cash in either period without loss
• On their own, risk-averse individuals choose the
  liquid technology as the small probability that
  they need the money in t2 means that the liquid
  technology return exceeds the expected return on
  the long term technology

16
The Diamond-Dybvig Model

• The Role of Banks
• Banks take deposits from everyone and
• Invest most of the money in the long term
  technology
• Hold a portion of deposits in the liquid
  technology to provide for the liquidity needs of
  the depositors who want their money in t2
• Banks offer an interest rate somewhere between
  that offered by the liquid short term technology
  and that of the long term technology
• So, all depositors are better-off (when things go
  well), as they get higher returns then they can
  obtain on their own plus the option of getting
  their money in t2 if they need it
• In essence, banks offer liquidity insurance

17
The Diamond-Dybvig Model

• Banking Fragility
• Suppose that for some random reason a greater
  than expected proportion of depositors seek to
  withdraw their cash in t2
• As withdraws continue, the expected return for
  the patient investors declines (liquidating the
  long-run investment at a loss in t2 makes the pie
  smaller)
• There comes a point where even a patient investor
  is better off getting his money in t2 than
  waiting. Anticipating this, everyone wants to
  get their money out first
• BANK RUN!!!!

18
From Banks to the Banking System

• Limitations of a Single Bank Analysis
• A single bank failure wont have much affect on
  the stability of the financial system
• A Banking System
• The system consists of a number of Diamond/Dybvig
  banks
• Individual banks experience liquidity shocks
• The banks are linked together via the interbank
  loan market (the payment system)
• The banks can therefore insure each other against
  an idiosyncratic liquidity shock by borrowing or
  lending on the interbank market
• The Banking System is one big bank

19
A Banking Crisis

• Normal Times
• If an individual bank experiences a liquidity
  shock, then it can meet that extra liquidity
  demand by borrowing on the interbank market
  rather than liquidate a portion of its long term
  assets (at a loss)
• Individual bank runs become less likely
• Crisis
• Suppose that a liquidity shock hits the economy
  as a whole
• If the shock exceeds the liquid reserves of the
  banking system, then moving liquidity around via
  interbank loans will not suffice
• Individual banks must then liquidate long-term
  assets (at a loss), making the patient investors
  who dont withdraw worse off
• BANK RUNS!!!!

20
Key Papers

• Diamond and Rajan 2001, Liquidity Risk,
  Liquidity Creation, and Financial Fragility A
  Theory of Banking, Journal of Political Economy
  109
• Allen and Gale 2000, Financial Contagion,
  Journal of Political Economy 108
• Chen 1999, Banking Panics The Role of
  First-Come, First-Served Rule and Information
  Externalities, Journal of Political Economy 107

21
II. Channel of Contagion
22
Getting the Snowball RollingChannels of
Contagion

• Bank Runs
• Random liquidity demands trigger bank runs that
  cause system melt-down
• Interbank exposures
• Bank failures spread because failing banks lack
  the resources needed to repay their interbank
  loans to otherwise healthy banks, causing them to
  fail in turn
• The Asset-Price Channel
• An individual shock that causes many banks to
  sell related assets at the same time, overloading
  the market for such assets, pushing prices down
  further
• Exposure to Common Shocks (not really contagion)
• Banks with big exposures to the same risk can all
  be wiped out by a big shock

23
Channel 1 bank runs
24
Bank Runs and Transaction Costs

• Taking all of your money out of the bank isnt
  free
• You will have to find another bank
• You might get robbed
• So, one might think that people would need a very
  good reason to withdraw all of their cash
• If everyone knows that everyone faces a big cost
  for withdrawing all of their cash, sunspot/panic
  driven bank runs are far less likely to occur

25
Bank Runs in the Real World

• Bank runs dont actually happen
• Reviewing almost 3,000 bank failures over the
  period 1865 to 1936 (mostly before deposit
  insurance), OConner 1938 found that runs or
  loss of public confidence was cited as the reason
  for the failure in less than 5 of cases
• Deposit insurance makes runs even less likely

26
Bank Runs Key Empirical Papers

• OConner 1938, The Banking Crisis and Recovery
  Under the Roosevelt Administration, Chicago
  Callaghan and Co.
• Calomiris and Mason 2000, Causes of U.S. Bank
  Distress During the Depression, NBER Working
  Paper no. 7919
• Calomiris and Mason 1997, Contagion and Bank
  Failures During the Great Depression The June
  1932 Chicago Banking Panic, American Economic
  Review 87
• Saunders and Wilson 1996, Contagious Bank
  Runs Evidence from the 1929-33 Period, Journal
  of Financial Intermediation 5
• Benston, Eisenbeis, Horvitz, Kane, and Kaufman
  1986, Perspectives on Safe and Sound Banking,
  Cambridge MIT Press

27
Channel 2 inter-bank linkages
28
Contagion Through Connections The Interbank
Loan Channel

• The interbank loan market is extremely active
• This loans create exposures between banks
• So, if one bank fails, it is at least logically
  possible that it will drag its creditors down
  with it
• People have explored this possibility through
  case studies and through simulations of bank
  failures using the actual (estimated) matrix of
  interbank exposures for many countries

29
Interbank ExposuresA Case Study of Continental
Illinois

• Continental was the 7th largest US bank at time
  of failure in 1984
• Assets in excess of 32 billion
• Continental was extremely active in the interbank
  market
• Weird Illinois banking regulations limited each
  bank to one branch
• Even with a big branch, one is not going to
  become the 7th largest bank in the on the basis
  of local deposits alone
• Continental therefore financed its expansion with
  interbank loans
• At the time of failure, Continental was the
  largest correspondent bank in the US, with either
  loans or deposits from 2300 other banks

30
Continental Illinois What Happened

• In the event, the FDIC bailed all creditors out
• But, the House Banking Committee investigated
  what would have happened without the government
  bailout
• Assuming losses of 60 cents on the dollar for
  Continentals assets
• 27 small banks become insolvent
• 56 additional small banks take a big hit ( gt50
  of capital)
• Total losses to heavily affected banks lt 500
  million

31
Contagion Through Interbank Exposures
Simulation Studies

• Method
• Estimate the matrix of interbank exposures using
  real data for the banking system as a whole
• Suppose that a given bank fails or
• Model overall bank exposures and hit the banking
  system with a shock such that an initial set of
  banks fail
• Elsinger, Lehar, and Summer 2003
• Assume a loss given default on a failed banks
  assets
• Trace the impact of initial and any follow-on
  bank failures through the system using the
  estimated interbank exposure matrix

32
Simulation Studies Results

• Results
• Every study finds the same thing with a large
  value of loss given default, there is limited
  contagion.

33
Simulation Studies Key Articles

• Furfine got the literature rolling, and the idea
  was contagious!
• US Furfine 2003, Interbank Exposures
  Quantifying the Risk of Contagion, Journal of
  Money, Credit, and Banking 35
• Belgium Degryse and Nguyen 2004, Interbank
  Exposures An Empirical Examination of Systemic
  Risk in the Belgian Banking System, National
  Bank of Belgium Working Paper
• UK Wells 2002, UK Interbank Exposures
  Systemic Risk Implications, Financial Stability
  Review

34

• Germany Upper and Worms 2002, Estimating
  Bilateral Exposures in the German Interbank
  Market Is there a Danger of Contagion?,
  Deutsche Bundesbank Discussion Paper 9
• Austria Elsinger, Lehar, and Summer 2003, The
  Risk of Interbank Credits A new Approach to the
  Assessment of Systemic Risk, Bank of Austria
  Working Paper

35
Channel 3 The Asset Price Channel
36
The Asset Price Channel

• Demand curves slope down in securities markets
  too
• Markets are not infinitely deep
• As more people sell a given asset, the natural
  buyers (i) exhaust their demand (ii) hit their
  risk tolerance (iii) run out of money
• The equilibrium price of the asset falls (for at
  least a while)
• Pioneering work Grossman and Miller 1988

37
Assets Price Channel

• The snowball
• A trader knows that he must sell security A, that
  others may have to sell A, and that As demand
  curve has a steep slope at the time
• Everyone who is long in A wants to sell first
• A race for the door ensues (everyone rushes to
  sell)
• The rush to sell causes the panic that all the
  traders were dreading

38

• The mountain down which the snowball falls
• An initial market price fall puts many traders at
  a level close to their loss-limits, encouraging
  them all to close out their positions
• Morris and Shin 2003
• Risk neutral traders who may have to sell next
  period all sell now to avoid the possibility of
  selling during a panic, causing the panic instead
  
• Bernardo and Welch 2003
• Analytically equivalent to a Diamond/Dybvig Bank
  run

39

• Why doesnt the market automatically stabilize?
• Suppose that trader Z knows that other traders
  have to sell
• Z will benefit by starting an asset run
• Prices will fall to a below long-run value level,
  giving Z a chance to buy low (during the panic)
  and sell high (after normal conditions apply)
• Far from stepping in to provide liquidity during
  a panic, then, traders in Zs position help the
  panic along
• The market will not automatically stabilize

40
Asset Price Contagion

• The problem conflict between individual
  incentive and collective incentive
• Runs start because each individual trader ignores
  how his actions affect the probability of a
  crisis
• Individually rational to run, but collectively
  daft
• Privately rational (but collectively inefficient)
  selling then leads to a crash

41
Asset Price Contagion Key Papers

• A new literature
• Grossman and Miller 1988, Liquidity and Market
  Structure, Journal of Finance 43
• Bernado and Welch 2003, Liquidity and
  Financial Market Runs, Forthcoming in the
  Quarterly Journal of Economics
• Morris and Shin 2003, Liquidity Black Holes,
  Forthcoming in Review of Finance
• Schnabel and Shin 2003, Foreshadowing LTCM
  The Crisis of 1763, LSE Working Paper

42
Channel 4 common shock
43
What Does Cause Banks to Fail - common shock
channel

• Individual banks fail due to adverse economic
  conditions
• -- managerial Incompetence/Rogue Trading
  Barings
• Adverse Conditions The most common factors cited
  for bank failure in OConnors 1938 survey of
  why banks failed were local financial distress
  and incompetent management
• One might then think that waves of bank failures
  occur because many banks pursue strategies that
  create exposures to the same adverse shock

44
An Aside Prudential Supervision

• If (unhealthy) banks fail because of macro
  factors, is there any point to prudential
  supervision?
• Yes!
• As banks approach failure, their incentive to
  gamble for resurrection becomes enormous
• If a bank increase the risk in its portfolio, it
  can stick the deposit insurance fund if it loses
  and keep the gains if it wins
• If regulators do not monitor banks to see which
  banks are in this position, and do not act
  promptly to shut-down any bank they find in this
  position, then the cost of a financial crisis can
  rise dramatically

45
Common Shocks Key Papers

• This literature is vast, but here are a couple of
  papers to get one started
• Calomiris and Mason 2000, Causes of U.S. Bank
  Distress During the Depression, NBER Working
  Paper no. 7919
• Demirguc-Kunt and Detragiache 1998, The
  Determinants of Banking Crises in Developing and
  Developed Countries, IMF Staff Papers 45
• Kaufman and Scott 2002, What is Systemic Risk
  and do Bank Regulators Retard or Contribute to
  it?, Loyola University (Chicago) working paper
• Kaufman and Seelig 2002, Post-Resolution
  Treatment of Depositors at Failed Banks and
  Severity of Bank Crises, Systemic Risk, and
  Too-Big-To-Fail, Economic Perspectives (Chicago
  Fed)

46
III. Consequence of Financial Crisis
47
Cost of Financial Instability
48
Counter Example Norwegian

• During the Norwegian banking crisis, banks
  holding 95 of all commercial bank assets became
  insolvent
• Without affecting the health of the Norwegian
  Corporate Sector
• Ongena, Smith, Michalsen 2003
• Norways well developed and well working
  financial markets provided corporates with a
  robust alternative method to acquire financial
  services
• Strong protection for minority shareholders
• Transparent accounting
• Strong public markets

49
Summary

• Financial crises are public badcreate
  externality and cause output losses.
• Source of contagion liquidity shock.
• Channels of contagion bank run inter-bank
  market asset price cascade and common shock.
• Financial market can compliment banking system as
  financial intermediation.
• How does the structure of a financial system
  contribute to its stability?