financial stability, early warning indicators and the …€¦ · restricted the financial cycle is...
TRANSCRIPT
Restricted
Financial Stability, Early Warning
Indicators and the Countercyclical Capital
Buffer
Mathias Drehmann1
Bank for International Settlements
Seminar on Financial Stability and Macroprudential Policy
Hosted by the Central Bank of Brazil and the Inter-American
Development Bank, Rio de Janeiro, 12 December 2014
1 The views presented are those of the author and do not necessarily represent those of the Bank for
International Settlements or those of the Basel Committee
Restricted
Operationalising macroprudential policies
CGFS (2012) report sets out 3 high-level criteria that are key in
determining instrument selection and application in practice
The ability to determine the appropriate timing for the
activation or deactivation of the instrument
The effectiveness of the macroprudential instruments (MPIs)
in achieving the stated objective
The efficiency of the instrument in terms of a cost-benefit
assessment
Report ends with 9 high level policy questions and answers that
help operationalising macroprudential instruments
2
Restricted
Increase resilience(Main objective)
Imp
act on
the
cred
it cycle
(side
be
nefit)
↑ lending spreads
dividend and bonuses
Undertake SEOs1
credit demand
Options to address shortfall
Asset prices
Loan market
Incr
eas
eca
pit
al r
eq
uir
em
en
ts
credit supply
Voluntary buffers
Arbitrage away
Leakages to non-banks
Expectation channel
Reprice loans
assets, especially with
high RWA
↑ Loss Absorbency
Tighter risk management
Transmission map for raising CCB requirements (CGFS (2012))
Restricted 4
Early warning indicators for
macroprudential instruments
Restricted 5
What do we know about the financial cycle?
Credit booms tend to go hand in hand with surges in asset and
particularly property prices. At the late stage of the boom, debt
service burdens rise substantially.
Need to pick-up in real time
Once the boom ends, financial crises often occur, or at least
periods of severe macroeconomic stress.
During the bust, asset prices and debt service burdens fall and
output growth is lower than during normal recessions.
See eg Kindleberger (2000), Minsky (1982), Mendoza and Terrones (2008,
2012), Reinhart and Rogoff (2009), Claessens et al (2009, 2011), Aikman et
al (2010), Drehmann et al (2012), Jorda et al (2013, 2014)
Restricted
The financial cycle is different from the business cycle
6
The financial and business cycles in the United States
1 The financial cycle as measured by frequency-based (bandpass) filters capturing medium-term cycles in real credit, the credit-to-GDP
ratio and real house prices. 2 The business cycle as measured by a frequency-based (bandpass) filter capturing fluctuations in real GDP
over a period from one to eight years.
Source: M Drehmann, C Borio and K Tsatsaronis, “Characterising the financial cycle: don’t lose sight of the medium term!”, BIS Working
Papers, no 380, June 2012.
Financial cycle is measured by credit and property price developments
Peaks of the financial cycle coincide with financial crises
Restricted
Which criteria do early warning indicators need to
satisfy?
Criterion 1: Captures boom in the financial cycle in real time
Ideal EWI needs to signal crisis early enough, likely to be 1-
2 year implementation lag
Ideal EWI signal crises not too early as this may undermine
effectiveness
Criterion 2: Is stable as policymakers adjust policy stance
gradually when information is noisy
Criterion 3: Is understood as practitioners value sensibility of
forecasts more than accuracy
Purely statistical approaches are not suitable for policy
purposes and communication
7
Restricted 8
A horse race
Given these criteria, Drehmann and Juselius (2014) assess
Credit measures: (total) credit-to-GDP gap and real (total)
credit growth
Asset prices: Real property and equity price gaps and real
property and equity price growth
None-core bank liabilities (Hahm, Shin, and Shin (2012)):
claims on banks held by foreign creditors relative to M2
GDP growth
History
Debt service ratio (DSR):𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡 𝑝𝑎𝑦𝑚𝑒𝑛𝑡𝑠+𝑎𝑚𝑜𝑟𝑡𝑖𝑧𝑎𝑟𝑡𝑖𝑜𝑛𝑠
𝑖𝑛𝑐𝑜𝑚𝑒
Restricted 9
The ROC curve as measure of signaling quality
Signal quality and policymakers’ preferences Graph A
Fully informative signal Noisy signal Uninformative signal
The red line denotes the ROC curve. The dotted lines denote preferences of a policy maker who weights the expected costs and expected
benefits of macroprudential interventions linearly. The blue (green) line indicates high (low) costs relative to benefits.
Source: Based on Drehmann and Juselius (2013).
1
1False positive rate
True
positive
rate
1
1False positive rate
True
positive
rate
1
1False positive rate
True
positive
rate
Restricted 10
The signalling quality of different EWIs
AUCs for different forecast horizons Graph 2
Credit-to-GDP gap Debt-service-ratio (DSR) Non-core liabilities
Credit growth Property price growth GDP growth
The horizontal-axis denotes the forecast horizons in quarters before crises. The vertical-axis denotes AUC. The horizontal line at 0.5
highlights the value of an uninformative indicator. A solid blue line indicates that the specific variable for the given horizon is statistically
different from an uninformative indicator, while a dashed blue line indicates the opposite. A hollow blue circle shows that the signal is
stable in the sense that it does not reverse direction within the forecast horizon until the crisis. Red diamonds highlight that the specific
variable is statistically the best indicator for this particular horizon. Other indicators that are not statistically different from best-performing
indicator are marked by solid blue circles.
Source: Drehmann and Juselius (2013).
Restricted 11
The credit-to-GDP gap: typical questions
Critique 1: The gap is not an equilibrium notion of credit
Correct, but this is not the objective of the CCB
Question 2: The gap moves countercyclical with GDP growth
Statistically not correct in a cross-country sample
Objective is not to manage GDP cycle but to increase
resilience
Clearly, diverging business cycles and financial cycles
present challenges
Drehmann and Tsatsaronis (2014) discuss questions and critiques in detail
Restricted 12
Is the credit-to-GDP gap a good indicator for EMEs?
Critique 3:
The use of the credit-to-GDP gap hinders financial
deepening
- If financial deepening occurs at a steady pace, gradual and
persistent growth of credit will be embedded in the trend of
the credit-to-GDP ratio.
- Much of the run-up in credit-to-GDP ratios in advanced
economies can be explained by falling interest rates
Critique 4:
Analysis has been undertaken using mainly advanced
economy data.
This is correct, but the credit-to-GDP gap is a good indicator
for EMEs.
Restricted 13
The forecast performance of the credit-to-GDP gap for different samples
AUCs for different forecast horizons
The horizontal axis denotes the forecast horizons in quarters before crises. The vertical axis denotes the AUC. The horizontal line at 0.5
highlights the value of an uninformative indicator.
1 As baseline, credit-to-GDP gaps are calculated by the standard methodology and the full sample of all countries is used. 2 95%
confidence interval for the AUC using the baseline model. 3 EMEs according to World Bank classification. 4 Countries with a credit-to-
GDP ratio below 100%.
Restricted 14
Measurement problems and the credit-to-GDP gap
Does the trend change when new data becomes available?
By design not
Using a 2-sided filter not possible for policy makers and
gives worse EWI results
There is a “start point” problem
An analysis shows that at least 10 years of data for the
credit-to-GDP ratio are needed, before trend calculations
become stable
The gap can also be influenced by structural breaks
Restricted 15
Combining indicators improves EWI
Given asset and credit booms go often hand in hand, useful to
jointly assess credit and asset prices
.2.3
.4.5
.6.7
.8.9
1
AU
C
-20 -15 -10 -5 0Horizon
DSR alone DSR & Credit/GDP gap
.2.3
.4.5
.6.7
.8.9
1
AU
C
-20 -15 -10 -5 0Horizon
DSR alone DSR & prop. gr.
.2.3
.4.5
.6.7
.8.9
1
AU
C
-20 -15 -10 -5 0Horizon
Credit/GDP gap alone DSR & Credit/GDP gap
.2.3
.4.5
.6.7
.8.9
1
AU
C
-20 -15 -10 -5 0Horizon
Credit/GDP gap alone Credit/GDP gap & prop. gr.
Restricted 16
Why do EWIs work?
What do we know?
Credit and asset price developments at the core of the
financial cycle
Credit-to-GDP gap and DSR best individual EWI
Is there an underlying steady state relationship?
Sustainable level of credit at the micro level determined by
Net worth, ie credit relative to assets
Debt service burdens, ie interest payments + amortizations
relative to income
Do these relationships hold at the macro level?
Juselius and Drehmann (2015)
Restricted 17
Net worth and DSR determine the equilibrium level of
the credit-to-GDP ratio: the case of the US
Net worth DSR
Restricted 18
Deviations affect credit and expenditure
-20
-10
01
02
0
1985q1 1990q1 1995q1 2000q1 2005q1 2010q1 2015q1time
Net worth DSR
Restricted 19
Deviations could have revealed that the US economy is
on an unsustainable path already in the mid 2000s
Credit
growth
Expenditure
growth
Restricted 20
Net-worth and debt service deviations key drivers of
credit and expenditure
• Simple framework captures
endogenous boom bust cycles
• System is highly fragile as both
deviations interact
Credit
Expenditure
Lending
rates
Deviations
Restricted 21
What about the credit-to-GDP gap?
-20
-10
01
02
0
1985q1 1990q1 1995q1 2000q1 2005q1 2010q1 2015q1
Credit/GDP gap Financial cycleNet worth DSR
Credit-to-GDP gap very good starting point for discussions
about CCB
Restricted 22
The role of judgement
There are no models\indicators that can deliver effective rule-
based counter cyclical instruments
Wrong signals can be issued
Date problems may hamper the analysis
Authorities need to take account of a broad set of information
State of business cycle, e.g. GDP growth
Market based indicators, e.g. credit spreads
State of the banking sector, e.g. profitability
→ Macroprudential monitoring capacity needs to be enhanced
→ Transparent communication is crucial
Restricted 24
References
Committee on the Global Financial System (CGFS), 2010, Macroprudential
instruments and frameworks: A stocktaking of issues and experiences. CGFS
Publications No 38
Drehmann, M, C Borio and K Tsatsaronis (2012): “Characterising the financial
cycle: don’t lose sight of the medium term!", BIS Working Papers, 380
Drehmann, M, and M Juselius (2014): "Evaluating early warning indicators of
banking crises: Satisfying policy requirements", International Journal of
Forecasting, No 30
Drehmann, M, and K Tsatsaronis (2014): "The credit-to-GDP gap and
countercyclical capital buffers: Questions and answers", BIS Quarterly Review
Juselius, M and M Drehmann (2014): “Answering the Queen’s question:
deviations of the credit-to-GDP ratio from its steady-state”, forthcoming