Enterprise-wide Scenario Analysis

Transcription

1 Finance and Private Sector Development Forum Washington April 2007 Enterprise-wide Scenario Analysis Jeffrey Carmichael CEO 25 April 2007 Date 1

2 Context Traditional stress testing is useful but limited Regulators becoming more demanding Boards and senior management becoming more demanding Enterprise-wide scenario analysis becoming industry best practice for this type of strategic thinking Benefits to: 1. Business 2. Risk management 3. Regulatory standing 2

3 Building a BIC Scenario Analysis Program 1. Define the scenarios 2. Assess macro-economic impacts 3. Link macro-economic drivers to financial models 4. Reality test the scenario outputs 5. Identify leading indicators of each scenario 6. Decide strategic responses 3

4 1. The Scenarios Should have substance, e.g. USD scenario Substance US trade deficit, foreign ownership etc. Should cover a range of situations, e.g. Global economic Local economic Global/local socio-economic Operational Reputational/Regulatory Be realistic 4

5 2. Macro-economic Impacts Requires serious economic/econometric modelling May need global as well as local models Use existing models Calibrate the outputs to financial model needs 5

6 3. Link Macro Shocks to Financial Models Each financial model has a set of drivers, e.g. Credit Risk: Financial Ratios Scores Weights Financial Score Calibration Ratings Migration EBIT/Tot Interest Debt/Income Debt/Assets Earnings Growth Via Logistic Transformation W% X% Y% Z% % Weight in Overall Score Scores Mapped to PDs New Customer CRSs Need to link macro outputs to financial model drivers 6

7 Scenarios should affect multiple risks Macro Variables: Incomes Prices Sectoral Growth Drivers of the Financial Models Obligor Specific PDs and LGDs Credit Ratings by Sector VaR in Traded Book VaR in Banking Book Operational Risk Impacts Liquidity Concentration Reputation 7

8 A Note on Correlations The bottom-up approach to financial models means With linkages from macro shocks to financial model drivers there should be no problem of estimated correlations The outcome should be a consistent enterprisewide assessment of risk 8

9 4. Reality Test Each scenario should be reality tested before going to the Board Test: Assumptions Overall impact Sectoral impacts Profitability impact E.g. Housing price collapse E.g. Rural price collapse 9

10 5. Identify Leading Indicators Few tail events have no leading indicators Identify the key factors Assign responsibility for monitoring and ranking changes 10

11 6. Strategic Response Protective positioning Contingency planning Combination 11

12 A. Business Benefits Strategic: Scenarios = risks and returns Identify opportunities as well as risks Framework for Risk Appetite, and Capital Allocation Testing ground for strategic plan Insight into and basis for strategic decisions Focus on responsive capabilities 12

13 B. Risk Management Benefits Market Risk: VaR measures risk in normal market conditions Stress testing examines extreme market conditions Credit Risk: PD, LGD and EAD models measure expected loss (EL) Stress testing examines sensitivity of unexpected loss (UL) Operational Risk: Historical data can estimate high frequency, low impact events Stress testing examines low frequency, high impact events 13

14 Original silo Approach Stress Testing Market Risk Scenarios Credit Risk Scenarios Operational Risk Scenarios Market Risk Models Credit Risk Models Operational Risk Models 14

15 Enterprise-wide Approach Scenario Analysis Scenarios Macro economic links (CGE) Risk models Completes the Picture. Scenario analysis goes another level beyond stress testing by integrating the risk drivers and risk types and working through all the flow-on effects of a particular shock 15

16 Improving Risk Models Improves understanding of risk models and their inter-relationships Provides reality test of assumptions and outputs of risk models Improves transparency - weaknesses and gaps in modelling framework documentation is critical Encourages Boards and senior management to dedicate adequate resources to risk modelling and analysis 16

17 C. Regulatory Benefits Basel II Pillar 2: The first key principle: Banks should have a process for assessing overall capital adequacy relative to risks Basel II Pillar 2: The second key principle: Supervisors should review and evaluate banks capital strategies and assessments 17

18 Basel II Framework Coverage of Scenario Analysis Basic Pillar 1 risks Other risks considered under Pillar 2 Other considerations Element 1 Adequacy of Pillar 1 Element 2 Risks not fully captured under Pillar 1 Element 3 External Factors Credit Risk Residual Risk Settlement Risk Reputation Risk Business Risk (earnings and costs) Market Risk Securitization Risk Interest Rate Risk the Banking Book Strategic Risk Economic and Regulatory Environment Operational Model Risk Concentration Risk Liquidity Risk Capital Planning Qualitative Assessments Peer Group Comparisons Supervisory Review Process Individual Capital Guidance Other Measures Internal Governance and Oversight Monitoring and Control 18

19 What have we learned.. First how it can help Good news scenario analysis is a very effective strategic tool in preparing for crisis management Bad news hardly any banks are doing it The problems Resource intensive Needs cooperation Needs strategic initiative Distracts from other priorities Lessons for the World Bank. 19