Measuring Financial Stability

1
Measuring Financial Stability

• by Oriol Aspachs-Bracons (LSE), Charles A.E.
  Goodhart (LSE), Miguel Segoviano (IMF), Dimitrios
  Tsomocos (University of Oxford) and Lea Zicchino
  (BoE)

Developing a Framework to Assess Financial
Stability Bank of France 20 February 2008
2
Literature

• Searching for a Metric for Financial Stability
• LSE Financial Markets Group Special Paper
  167
• by Oriol Aspachs-Bracons (LSE), Charles A.E.
  Goodhart (LSE), Miguel Segoviano (IMF), Dimitrios
  Tsomocos (University of Oxford) and Lea Zicchino
  (BoE)
• Towards a Measure of Financial Fragility
• Annals of Finance, 3(1), 37-74, 2007
• by Oriol Aspachs-Bracons (LSE), Charles A.E.
  Goodhart (LSE), Dimitrios Tsomocos (University of
  Oxford) and Lea Zicchino (BoE)

3
Outline of the presentation

• Motivation
• The Model
• Definition of Financial Stability
• Financial stability and banking sectors capital
  adequacy
• Empirical analysis
• Concluding remarks

4
A Framework for Financial Stability
Contrasts between Price and Financial Stability
5
Why a measure of Financial Stability?

• It is easier to define, measure, model, analyse
  and control price stability than to do so for
  financial stability
• In several ways the problem of measurement should
  have priority as we need to compare and analyse
• But in order to measure financial stability, we
  need a formal, model-based definition

6
Alternative definitions of Financial Stability

• Andrew Crockett (1997) financial stability
  requires that the key institutions and the key
  markets are stable
• Mishkin (1994) financial instability occurs when
  shocks to the financial system interfere with
  information flows so that the financial system
  can no longer do its job of channelling funds to
  those with productive investment opportunities
• Haldane et al (2004) financial instability is
  any deviation from the optimal saving-investment
  plan of the econom that is due to imperfections
  in the financial sector
• Issing(2003) and Foot (2003) financial
  instability linked to financial market bubbles,
  or more generally, volatility in financial
  markets

7
Definition of Financial Stability

• Financial fragility is characterised by reduced
• bank profitability and increased aggregate
  default
• An increase in both banking sector vulnerability
  and aggregate default (lower repayment rates) are
  linked to welfare losses (agents utilities)

8
Key features of the model

• General equilibrium model of an exchange economy
  with money and banks
• Incomplete markets (limited participation)
• Heterogeneous investors/consumers and banks
• Liquidity constraints
• Endogenous default

9
Key implications

• Equilibrium is compatible with default (no need
  to resort to multiple equilibria as in models of
  the Diamond-Dybvig type)
• Equilibrium is constrained Pareto inefficient
  (policy matters!)
• Nominal changes affect both prices and quantities
  (money is non-neutral)
• The Central Bank controls the overall liquidity
  of the economy and such liquidity, as well as
  endogenous default risk, determines the interest
  rates

10
Description of the model

• The model has consumers/investors and banks who
  maximise utility (subject to a budget constraint
  the first - and a capital adequacy constraints
  the second)
• It extends over two periods and all uncertainty
  is resolved in the second period
• Trade takes place in both periods in the goods
  and equity markets
• In the first period, agents also borrow from, or
  deposit money with banks, to achieve a preferred
  time path of consumption

11
Description of the model (continued)

• Banks trade among themselves to smooth out their
  individual portfolio positions
• The central bank intervenes in the interbank
  market to change the money supply and thereby
  determines the official interest rate
• Capital adequacy requirements (CARs) on banks are
  set by a regulator
• Penalties on violations of CARs (and on default
  of any borrower) are in force in both periods

12
The time structure of the model
13
Description of the model (continued)

• The Banking Sector it is assumed to operate
  under a perfectly competitive environment

14
Description of the model (continued)

• The Banking Sector
• Since each bank is different (it has different
  risk/return preferences and different initial
  capital), there are more than one market for bank
  loans and bank deposits
• We introduce limited access to consumer credit
  markets, with each household assigned to borrow
  from a predetermined bank
• Therefore, there are different interest rates
  across the banking sector

15
Simplified model for calibration

• The model has two periods and two possible states
  in the second period
• i is the good state, which occurs with
  probability 0.95, and ii is the bad state (with
  probability 0.05)
• Agents
• three banks, d, g and t. Banks d is a net lender,
  while g and t (aggregate of 5 banks) are net
  borrowers in the interbank market
• Households/investors a, b, q, f. Their behaviour
  is modelled via reduced-form equations
• Central bank/Regulator
• Markets commodity, loans, deposits, interbank

16
Banks optimisation problem (1)
subject to
17
Banks optimisation problem (2)
18
Households sector

• Loan demand for each bank is a function of
  expected GDP and the banks lending rate
• Mr f supply of deposits is a function of expected
  GDP and each banks deposit rate relative to all
  other banks deposit rates, adjusted for their
  expected default rates

19

• Households loan repayment rates are a function
  of expected GDP, and the total aggregate credit
  supply
• Finally, GDP is a function of total aggregate
  credit supply

20
Market clearing conditions
21
(No Transcript)
22
Financial Stability with reduced-form household
sector

• Financial fragility is characterised by reduced
• bank profitability and increased aggregate
  default
• An increase in both banking sector vulnerability
  and aggregate default (lower repayment rates) are
  linked to GDP

23
Comparative statics analysis

• A 10 negative shock to M

24
(No Transcript)
25
Comparative statics analysis

• A 10 negative shock to M

26
Transmission mechanism (table 1)
M implies inter-bank rate r
increases Bank d is willing to invest more in
the inter-bank market ds deposit and
lending rates increase Banks g and t
reduce inter-bank borrowing, increase deposit
demand and reduce household lending Less
credit availability induces more household
default To maintain profitability, banks adopt
riskier investment strategies
27
Comparative statics analysis (no CARs)

• A 10 negative shock to M

28
Do capital requirements affect the transmission
of a negative shock to a banks capital?

• Banks have no incentive to raise their capital
  holdings and therefore to raise their
  profitability
• As a result, banks repayment rates are higher
  (equivalently, default rates are lower)
• Less credit by the distressed bank to households
  
• Lower GDP

29
Financial fragility with and without CARs

30
Financial fragility with and without CARs

• When an effective CAR is introduced, a bank
  chooses higher profitability, which it can only
  achieve by taking on more risk and/or raising
  interest rate spreads
• In turn, such higher interest rates, charged to
  borrowers, will cause them to borrow less, which
  reduces GDP in our model, and to take on riskier
  investment (i.e. to plan to default more often)

31
Financial fragility with and without CARs

• The benefit to financial stability of safer
  banks will be offset, to some extent, by both
  banks and bank borrowers selecting riskier
  portfolios, higher interest rates and lower
  output
• However, CARs might still be a net benefit,
  depending on the likelihood of bank contagion,
  the probability of future shocks, etc.
• In practice, this adverse side-effect can be
  mitigated by relating CARs more closely to the
  relative riskiness of assets (as in Basel II) or
  by limiting the allowable rise in interest rates
  (Hellman et al.)

32
Empirical analysis

• We analyse the relationship between a small
  number of macroeconomic variables and the
  probability of default and the banking sector
  equity index of seven industrialised countries
  (reduced-form country-level VARs and panel-VAR)
• Initially tried data from bank accounts, e.g.
  profits, NPLs, write-off, etc but did not work
  well
• - accounting inconsistencies, both between
  countries and over time
• - manipulation and smoothing
• - long lags in reporting, especially write-offs

33
Empirical analysis

• So we turned to market variables equity values
  (as a proxy for profits) and PDs, with the latter
  taken from the IMF
• Note that estimates of PDs incorporate equity
  valuation. So why is it not a sufficient
  statistic in itself?
• Correlation between the change in equity values
  and estimated PD is only -0.32 on a quarterly
  basis

34
Empirical analysis

• Our procedure was to examine relationships
  between GDP, our measure of social welfare,
  inflation, PD of banking sector and equity values
  of banking sector
• Initial exercises indicated that both PD and
  percentage changes in equity values were
  threshold variables they only adversely affected
  GDP if worse than some level
• Threshold level chosen to maximise fit

35
Empirical analysis

• Our procedure was to examine relationships
  between GDP, our measure of social welfare,
  inflation, PD of banking sector and equity values
  of banking sector
• Initial exercises indicated that both PD and
  percentage changes in equity values were
  threshold variables they only adversely affected
  GDP if worse than some level
• Threshold level chosen to maximise fit

36
Data

• Data set includes Finland, Germany, Japan,
  Korea, Norway, Sweden and UK over period 1990Q4
  2004Q4
• PD transformation of the distance to default
  indicator used by the IMF to gauge banking sector
  soundness (confidential data)
• Macroeconomic variables from IFS and OECD
  database
• Residential property prices from the BIS
• Equity data from Bloomberg

37
Results

• Surprisingly good for Panel
• Supportive, but not totally so, for individual
  countries
• PDs have mcuh stronger influence on GDP than
  equity values
• When property prices are included, the effect of
  equity values becomes less important

38
Impulse response functions (PD, GDP, Equity,
Inflation)
39
Impulse response functions (PD, GDP, Equity,
Inflation, Property Prices)
40
Variance decomposition
41
Variance decomposition
42
Can we get a single metric, comparable over time
and across countries for financial
stability? Yes, conditional on- (i) methods
(transformation, etc.) used to estimate
PD (ii) estimation of relative weights of PD and
Eq from empirical exercises What does it look
like in our case? Note that PD much stronger
influence on GDP than Eq, (as might have been
expected), so what determines PD? Further,
separate work by Goodhart, Hofmann and Segoviano,
Default, Credit Growth and Asset Prices, IMF
(2005/6).
43
Welfare effects of POD and Bank equity I. Sweden
44
II. Korea
45
III. UK
46
Welfare indexes of financial fragility
47
Concluding remarks

• We have proposed a model-based definition of
  financial fragility
• We have run comparative statics exercises to
  identify shocks that induce financial fragility
  (with and without CARs for banks)
• We have investigated whether data support our
  claim that banking sector distress (lower
  profitability and higher default) induces welfare
  losses (proxied by GDP)
• Next steps

48
Concluding remarks

• We have proposed a model-based definition of
  financial fragility
• We have run comparative statics exercises to
  identify shocks that induce financial fragility
  (with and without CARs for banks)
• We have investigated whether data support our
  claim that banking sector distress (lower
  profitability and higher default) induces welfare
  losses (proxied by GDP)
• Next steps