Transcript Credit cycle and systemic risk

Macroprudential policy – indicators and
tools
Jan Frait
Executive Director
Financial Stability Department
Credit cycle and systemic risk – the case for forwardlooking approaches in conduct of macroprudential policy
• In a credit cycle, systemic risk evolves differently in two stages:
accumulation (build-up) and materialization (manifestation).
• Note the financial (in)stability paradox: a system is most vulnerable
when it looks most robust.
Build-up of systemic risk
period of financial
exuberance
Materialisation of systemic risk
period of financial
distress
period of low
current risk
time
period of high
current risk
time
normal conditions
marginal risk of
financial instability
degree to which risks
materialise as defaults, NPLs
and credit losses
2
Systémové riziko a finanční cyklus
• Systémové riziko vzniká v průběhu finančního cyklu, kdy finanční instituce
vytvářejí rostoucí a vzájemně korelované expozice (typicky hypoteční
úvěry) citlivé na shodné makroekonomické faktory.
• Špatné úvěry vznikají obvykle v dobrých časech, což se ukáže v následující
recesi.
Procyklické chování ekonomiky
70
10
60
8
50
6
Opravné položky a úvěry v selhání (v mld. Kč)
(leden 2001-listopad 2011)
200
180
160
100
-2
80
-10
-4
60
-20
-6
40
-30
-8
20
opravné položky (mld. Kč)
I-11
I-10
I-09
0
I-08
tempo růstu úvěrů (mzr. v %)
tempo růstu cen bytů (mzr. v %)
tempo růstu HDP (mzr. v %)
I-07
I/99 I/00 I/01 I/02 I/03 I/04 I/05 I/06 I/07 I/08 I/09 I/10
I-06
0
I-05
0
10
120
I-04
2
20
140
I-03
30
I-02
4
I-01
40
úvěry v selhání (mld. Kč)
3
Conseptual approach to macroprudential policy
• The key concept describing the time dimension of systemic risk over financial
cycle is leverage (the indebtedness of economic agents, stocks of loans, the
ease of obtaining of external financing, the size of interest rate margins and
credit spreads, etc..).
• The leverage can be approximated by credit-to-GDP ratio:
• increases until the financial cycle turns over; normally the turnover happens
in an orderly way (a case of cycle without a crisis in Chart) without a need
for an extra policy action.
• sometimes the turn is disorderly and presents itself as the eruption of
financial crisis.
• then leverage starts to decline, although in the early phase of the crisis
remains high (given falling nominal GDP it can even rise in the initial postcrisis years).
• The deleveraging phase can therefore last several years, and in the event of
a deep crisis the leverage ratio can, after a time, fall below its long-term
normal value.
• consequently, the leverage ratio adjusts to economic conditions after a
considerable lag, so stock measures have only a limited information value as a
guide for the macroprudential policy response during the financial cycle.
• for this reason, forward-looking variables are needed that can be used to identify
situations where the tolerable limit for systemic risk has been exceeded.
4
Leverage over credit cycle – a slow motion process
The evolution of systemic risk and conduct of macroprudential policy over the financial cycle
Leverage
Good times (accumulation of
systemic risk): phase of
increasing leverage with
excessive optimism
Turning point (or
outbreak of crisis)
Bad times (materialisation
of systemic risk): phase of
deleveraging with
excessive pessimism
Normal leverage level
A case of cycle
without a crisis
Time
Signal to activate
macroprudential policy: forwardlooking indicators credit gap or
property price gap…
Discontinuous change in
marginal risk of financial stability:
e.g. financial market indicators
(credit spreads, CDS spreads) or
market liquidity indicators
Signal to end support policies:
current indicators (default rate,
NPL ratio, provisioning rate,
lending conditions) and financial
market indicators
5
I.
Systemic Risk Indicators
6
Prevention and forward-looking indicators
• The main task of financial stability analysis as regards prevention is timely
identification of the risk of financial instability – the marginal contribution of the
current financial environment to the build-up of risks of a future financial crisis.
• In this phase, macro-prudential analysis must be focused primarily on the
identification of hidden risks to financial stability being generated in the balance
sheets of financial intermediaries and their clients.
• Analytical attention, however, must also be paid to the quality of cash flows, as
financial institutions with structural problems in their balance sheets, weak balancesheet liquidity and long maturity transformations are naturally far more prone to
cash-flow problems.
• Authorities need a set of forward-looking indicators providing information on the
possibility of materialisation of systemic risk in the future as a result of currently
emerging financial imbalances.
• This refers mainly to “gap” indicators based on the assessment of deviations of
factors determining the degree of leverage from their equilibrium or normal values.
• As regards the possibility of using forward-looking indicators to construct earlywarning systems, we feel that their information value and practical applicability
remain limited.
• Beware of financial markets‘ data signals – IMF, ESRB …..
• Analytically, it is also possible to use the FS paradox – extremely good values of
parallel indicators tell us that something strange is going on in the system.
7
Mitigation and identification of discontinuities
• In the systemic risk materialisation phase, the macroprudential policy focus
must be shifted to mitigating the impact of the crisis.
• In this phase it is vital to assess the scale of the risk materialisation problem and
the resilience of the financial system.
• Stress tests of the financial system’s resilience are a suitable analytical instrument
for performing this task.
• Macroprudential analyses must take into account the high degree of
discontinuity in the evolution of systemic risk – the potentially sharp transition
from good to bad times.
• To this end it is necessary to construct indicators characterising the start and end of
the materialisation of financial instability.
• In a small open economy, financial or informational contagion resulting from the
links between an economy and its institutions and the external environment can
be a major source of materialisation of systemic risk and of discontinuity in the
evolution of such risk.
• The analytical approach will differ significantly from country to country (share of
foreign ownership of financial institutions, dominance of subsidiaries or branches of
foreign banks, share of foreign currency loans, net external and foreign exchange
8
position of the banking sector and entire economy...).
Countercyclical capital buffers – the example of policy
• Policy of countercyclical capital buffers should follow the developments of
leverage over financial cycle and associated risk for financial stability.
Credit-to-GDP over financial cycle
period of financial
exuberance
Credit dynamics (e.g. y-o-y growth)
period of financial
exuberance
period of financial
distress
period of financial
distress
CCB set to zero
again
time
long-term „normal“
level of credit-to
GDP
time
CCB set at
maximum 2,5 %
turning point (start of
crisis): credit-to-GDP
still very high, but policy
has to change sharply
CCB set to zero
turning point (start of
crisis): credit growth
falls, lending conditions
tighten
9
How to Tell Normal Times from the Not So Normal Ones
• Financial exuberance period - in addition to the availability of cheap
credit, the emergence of overly optimistic expectations about future
income and asset prices required.
• Financial distress period - in addition to the limiting the availability of
credit, economic agents become over-pessimistic.
• To identify the onset of or exit from not so normal times - the gaps
based on the indicators‘ level relative to their long-term average or
trend reveal the story:
•
•
•
•
•
•
credit growth, credit-to-GDP;
debt-to-income, debt-to-assets dynamics;
money market risk premia, credit spreads and CDS spreads;
financial investors’ lever length;
length of maturity transformation by banks;
lending standards, credit risk gauges, ...
• Bad news: the framework is still more philosophical than operational!
10