Transcript L08- Stress Testing Banks and Financial Institutions

Macro-Prudential Policy and
Financial Stability:
Issues and Challenges
16-18 December 2013
Intercontinental Hotel
Amman, Jordan
Stress Testing Banks and Financial
Institutions
•
•
•
Good practices
Stress testing models and scenarios
Practical considerations
Mr. Keith Pooley
METAC Banking Supervision Technical Expert
GOOD PRACTICES
WHAT DOES BASEL SAY ABOUT STRESS TESTING?
The importance of stress testing
Stress Testing and Risk Governance
“Stress testing plays an important role in:
“Stress testing should form an integral part of the
overall governance and risk management culture of
the bank.






providing forward-looking assessments of
risk;
overcoming limitations of models and
historical data;
supporting internal and external
communication;
feeding into capital and liquidity planning
procedures;
informing the setting of a banks’ risk
tolerance;
and facilitating the development of risk
mitigation or contingency plans across a
range of stressed conditions.”
Stress testing should be actionable, with the results
from stress testing analyses impacting decision
making at the appropriate management level,
including strategic business decisions of the board
and senior management.
Board and senior management involvement in the
stress testing programme is essential for its effective
operation.”
Stress Testing and Capital Plans
“Supervisors should consider the results of
sensitivity analyses and stress tests conducted by
the institution and how these results relate to
capital plans.”
3
STRESS TESTING GOOD PRACTICES
The following slide provides an example of
good practice in relation to the Governance
Framework
The process is Board-led, with the Board:




proposing some of the scenarios to be
run
challenging approaches, scenarios and
outputs
approving approaches, scenarios and
outputs
receiving sufficient training to support
these objectives
GOOD PRACTICES - GOVERNANCE
Under leading practice, stress testing needs a set of governance arrangements driven
primarily by the Board and an integrated operating framework.
GOVERNANCE FRAMEWORK
Board and senior management engagement
Regular review of
programme to assess its
effectiveness
Governance of stress
testing and use
Outputs are actionable
and inform strategic
decision making
Clear responsibilities, allocated resources, and written policies and procedures
Embedded into risk management processes and supported with an appropriate risk infrastructure
GOOD PRACTICES – OPERATING FRAMEWORK
The following slide provides an example of
good practice in relation to the accompanying
Operating Framework
The framework encompasses:
 macroeconomic stress tests,
 scenario / ‘what-if’ analysis and sensitivity
analysis - appropriate for the scale and
complexity of the asset classes
 assumptions are sufficiently challenged and
tested
Outputs inform:
 management actions and gross and net
impacts
 Pre-emptive responses to manage
vulnerabilities
 Strategy and capital/liquidity requirements
GOOD PRACTICES – OPERATING FRAMEWORK
OPERATING FRAMEWORK
Firm strategy
Capital planning
•Identifying capital required to support growth
•Link to ICAAP
Stress Testing
Quantitative
Elements
(e.g. GDP, CPI)
Define
scenarios
Identify consequence
Assess impact
(quantitative &
qualitative)
7
TYPES OF STRESS TESTS
TYPES OF STRESS TESTS
Sensitivity analysis




Defined by shift in sensitivity
variable
Relatively easy to define and
implement – often used at trading
desk and business line level
Exposure
Shifts in several variables have to
be used in order to simulate
historical events
Correct use of stressed
correlations between risk types is
essential
Rating class
SENSITIVITY AND SCENARIO ANALYSIS APPLIED TO
CREDIT RISK AND MARKET RISK
Sensitivity analysis
Credit risk
Market risk
(incl ALM)
What if all ratings worsen by one grade?
What if default rates double in portfolios
X,Y,Z?
What if all correlations on our credit
portfolio model increase by 20%?
What is our collateral recovery rates are
systematically 25% lower for real estate
collateral?
What is the EUR/USD rate changes by X%?
Scenario analysis
What if there was an emerging market crisis
(contagion effects)?
GDP down 6-8%?
Deflationary vs. inflationary?
What if the previous historical boom and bubble in
house prices leads to a long-lasting retail mortgage
credit crisis?
Parallel shift of yield curve by X basis
points?
What if exchanges rates or interest rates
behave as they did historically during
period X?
Increased in implied volatility of European
stocks of Y%
What if interbank liquidity dries up during
next economic shock?
10
THE CLASSIFICATION OF GLOBAL STRESS SCENARIOS
Global stress scenarios can be classified accordiing to the construction of the scenarios
Moderate scenarios
Challenging scenarios
Extreme scenarios
TOP DOWN APPROACH TO DEFINING STRESS SCENARIO
Competitive
expectation
Senior management
Macro-economic
expectation
Strategic planning
department
Business & risk strategy
Research department incl.
chief economist
Translation of strategy into targets as well as business, risk and capital planning requires planning assumptions
• What makes the business case
profitable?
• What could endanger our
business case?
• Where are the considerable
concentrations (products,
customers, markets) ?
• What are the biggest threats of
our strategy?
• “What if” strategic decisions fail
(e.g. acquisition, capital increase)
12
Definition of planning assumptions
Best case planning scenario
Normal case planning scenario
Stress case planning scenario 1
Stress case planning scenario 2
Stress case planning scenario 3
• On which economic drivers does
the business case depend?
• What are macro-economic
expectations?
• What about business cycle
effects?
• What are unlikely by possible
economic crisis?
• How would counterparties,
customers and competitors be
effected by a crisis?
Risk controlling Moderation of scenario definition and portfolio input
HIERARCHY OF STRESS TESTS
Global scenarios
Scenario analysis
Sensitivity analysis
LINKING MACROECONOMIC EVENTS WITH THE TARGET METRIC
Process for thinking through linkage between macroeconomic events and target metric in
scenario based layered stress-testing
Portfolio
characteristics
Macroeconomic
scenarios
Macroeconomic
factors
Historical:
 GDP
 Score distribution
 PD
 Earthquake
 Interest rates
 Segment distribution
 LGD
 Recessions
 Exchange rate
 …
 EAD
 Credit crises
 Debt ratio
 Utilization
Hypothetical:
 Asset prices
 Attrition
 Extended recession
 Consumer
sentiment
 Response
 …
 Process costs
 Correlation of
events
 ...
Risk drivers
 Delinquency
 …
Target metric
 Portfolio loss
(Accounting)
 Portfolio loss
(Economic loss)
 Provisioning/Writeoffs
 Cap. Requirements
 Reduced margins
 Breakdown of new
business
 Fall in share price
 …
14
REVERSE STRESS TESTING – TEST TO DESTRUCTION
Reverse stress testing
Reverse stress testing should be modelled
through various stresses and consider
factors relating specifically to the firm as
well as the external environment.
Considerations when devising tests:
 Proportionality
 Basis of scenarios: hypothetical v
historical
 Internal v external shocks
 Idiosyncratic v market wide stress
 Rapid crystallisation versus protracted
impact
 Solo v Group
Scenarios that
could lead to
business failure
Why do it?
■ Explore business model
vulnerabilities
■ Engage senior management
■ Confront possibility of failure
■ Make decisions that better
integrate risk management
Mitigating actions or
triggers for future action
■ Improve contingency
planning
■ Inform stress testing
framework
It should be used as a risk management tool
– not a means of directly increasing capital
requirements
15
EMBEDDING THE STRESS TESTING PROCESS
Scenario Generation
Credit Risk
Department
 Suggests credit risk relevant scenarios
covering bank wide, regional, portfolio
or product specific credit risks

Market Risk
Department
 Suggests scenarios in ALM, market
and liquidity risk
Operational Risk
Department
Prioritisation
First Pass
Quantification
 Workshop o better
understand scenarios and
underlying risks
 Risk acts as moderator
and supplies basic
portfolio parameters in
order to allow
quantification
 Discussion based
quantification to extract
insights before more
rigorous first pass
quantification by central
team
 Suggests scenarios on external, fraud,
litigation, IT or compliance risks
Management /
Business

 Suggest strategic, bank-wide and
business case specific scenarios
16
Review of first
pass
quantification
 Central project team
 All scenarios are assigned a
rough probability and impact
Quantification
Review of
selection
 Reviews selection of
pilot scenarios with
representatives from
(Sen.) Management,
Business, Risk
Prioritisation –
Pilot Scenarios
 Select based on probability and
impact
 Pilot scenarios based also on variety
 Additional second tier risks to be
covered later or at less accuracy
Senior Management bears ultimate responsibility
Review of
Quantification
 Risk functions review
results and provide
comments, in particular
for their own risks
 Group risk is ultimately
responsible for scenario
analysis
 Results reported to
Management and
Business
Final Scenario
Quantification
 For each scenario estimates for
probability and impact are justified by
using various sources
 Macro model (recessions, avian flu, real
estate), benchmarks (most), theoretical
model / distribution analytics (ALM,
market, credit concentrations), external
data (mis-selling), think tank (ship sunk)
STRESS TESTING DESIGN INVOLVES A LOT OF REFLECTIONS
• Under which scenarios is my business model seriously endangered?
Design and strategy
• What will cause my business model to fail?
• What happens to my business in the next crisis?
• Which risk drivers are relevant to my business?
• In which scenarios would my largest sensitivities lead to major losses?
• Whom in my organisation should I involve in the design, modelling,
parameterisation and evaluation of stress scenarios
Methodology
• How do I translate (top-down) scenarios to relevant risk drivers, and what
severity do apply?
• What level of sophistication in describing and modelling of scenarios do I need
to apply to capture their essence?
• Which parameters should be shocked (how)? Over what time horizon?
Implementation
and evaluation
• How are second order effects and feedback incorporated?
• What do I do with the results of stress testing?
• To what degree may I anticipate help from the outside (investors, central bank,
government)?
17
KEY SUCCESS FACTORS WHEN SETTING-UP A
STRESS TESTING FRAMEWORK
Clearly define the
perimeter, i.e. the
group on a
consolidated level
Adequately involve
business and senior
management at all
levels of stress testing
18
Clearly identify the
target metric, e.g. the
P&L, capital, external
rating,…
Validate stress testing
results, question and
refine scenarios
across time
Create a real risk
culture across the
organization – get
stress testing out of
the risk function
Identify the variables
that have the most
effect on target
variables
Link stress testing
results to action,
e.g. outright
acceptance,
evaluate
contingency
planning
THE IMPORTANCE OF STRESS TESTING LIQUIDITY
“Many banks despite adequate capital levels experienced difficulties because they not manage their
liquidity in a prudent manner……The rapid reversal in market conditions illustrated how quickly
liquidity can evaporate and that illiquidity can last for an extended period of time…..” Basel 3
30 July 2007
17 August 2007
15 September 2008
13 October 2008
National bail-outs: Germany - €470bn, France - €340bn, USA
- $250bn bank nationalization, Spain - €100bn
17 August 2007
15 September 2008
17 February 2008
14 March 2008
16 September 2008
15-31 October 2008
Tapped national bail-out funds
RBOS, HBOS, Lloyds, Commerzbank, Bank of America, JPMorgan Chase,
Citigroup, Wells Fargo, Goldman Sachs, Morgan Stanley and more
LIQUIDITY RISK STRESS TESTING
Basel 3 prescribes stress testing should consider the two types of stress tests below and combinations of both




Institution specific
Market wide
(idiosyncratic)
(systemic)
Name-specific rumours
and bad news.
Wholesale funding
Downgrade and negative
outlook.
Retail funding
Retail run?
Marketable assets
10 risk
drivers
Funding
concentration
Wholesale run?
Short term market
disruption

Long term erosion in
funding sources due to
Off-balance sheet
ongoing market
tightening
Franchise
viability
I

Non-marketable
assets
Intra-day liquidity
Intra-group liquidity
Cross
currency
Combined stress test

Name specific concerns
- restricting recovery
LIQUIDITY RISK STRESS TESTING
Idiosyncratic Impact
First two weeks
Out to 3 months

Inability to roll over wholesale secured and
unsecured funding
Sustained leakage of funds
Sizeable retail outflow
Sustained outflow







Reduced intra-day credit provided by firm’s
settlement bank
Increase in payments withheld to a clearer
Prefunding for all payments
Closure of FX markets
Intra-group deposits repaid at maturity,
intra- group loans treated as evergreen
Multiple downgrade of long-term rating
Gradual return to normality
Systemic Impact




Uncertainty of accuracy of valuation of assets and
those of counterparties.
Inability to realise or realise particular classes of
assets only at excessive cost.
Risk aversion in funding markets.
Uncertainty as to the ability of a significant number
of banks to meet liabilities
EXAMPLE OF A CAPITAL STRESS TESTING
Economic variable
Year
US
GDP
2013
2.4%
-1.8%
-2.1%
Unemployment
2014
7.2%
13.4%
13.5%
House Price Index
2015
4.5%
-13.1%
-14.7%
GDP
2013
2.2%
-1.6%
-3.9%
Unemployment
2014
7.1%
13.1%
15.2%
House
600.0 price indices
2015
-2%
-18.2%
19.1%
Euro
Base Case
Regulatory mandatory
test
Eurozone Sovereign
18%
Pillar 1 RWAs (Stress)
16%
500.0
Pillar 1 RWAs (Base)
£ mns
RWA £bn
14%
400.0
12%
Minimum Core Tier 1 %
= 6% (Base)
10%
300.0
8%
200.0
6%
4%
100.0
2%
0.0
0%
Based
Stress
2011
2009
Based
Stress
2012
2010
Based
Stress
2013
2011
Based
Stress
2014
2012
Based
Stress
2015
2013
Minimum Core Tier 1 %
= 4% (Stress)
CoreInsuracne
Tier 1 as- %
of
Post
Core
RWAs
(base)
Tier 1 as % P1 RWA
(Base)
CoreInsurance
Tier 1 as- %
of
Post
Core
RWAs
Tier
1 as(Stressed)
% P1 RWA
(Stress)