Firm-Wide Stress Testing Restricted Stress Testing zwischen Granularität und Geschwindigkeit SAS forum Switzerland 2012 Alexandra Hansis May 2012
Why Stress Testing? Experience of the Crisis Severe losses and bank failures Capital eroded very quickly Banks were overconfident of their loss absorption potential Call for more severe stress scenarios ("Reverse Stress") Higher capital requirements, based on stress inputs Severe risk factor moves Beyond scope of risk models Reluctance to statistical models and to historical calibration Emphasis on stress testing as complementary tool Forward looking rather than historical scenarios Severe contagion across risk types and markets "Correlation" between risk types widely underestimated in Economic Capital models Diversification and hedge benefits widely overestimated Emphasis on group wide "holistic" stress testing Integrated frameworks capturing different risk types/exposures 1
Section 1 UBS Combined Stress Test Framework
Three Pillar Stress Testing Framework Reports, Limits, Planning Measure, monitor and control risk under stress conditions Firm-wide stress tests Regularly calculate and assess the potential for loss in a consistent fashion across the entire organization. Based on forward looking multiple macro-economic scenarios calibrated to different levels of severity. In addition to direct effects on reported profit and loss, these scenarios consider implications for risk-weighted assets and capital ratios as well as funding costs and business risks. The results of these aggregate scenarios is included in regular risk reporting and used as part of the overall risk appetite and business planning processes of the firm. Portfolio specific stress tests Stress-based measures focus on the market risk of trading positions and the default risk of credit sensitive positions for specific portfolios within business divisions These control-oriented measures are being monitored against limits. Reverse stress test Work backwards from potential loss amounts to determine the economic or financial scenarios that could give rise to such losses. This element seeks to close the gap related to the risk of under-estimate the severity of stress scenario. Ensure that there is transparency to senior management on the macroeconomic and risk factor movements that would be needed to achieve potential loss levels above those already highlighted by aggregate stress testing 3
The Role of Stress Testing in UBS Credit Risk Market Risk Issuer Risk Investment Risk Funding Risk Structural FX Country Risk Pension Risk Business Risk Operational Risk The purpose of portfolio stress measures is to quantify exposure to extreme market movements and geo-political events. Stress testing has gained increased focus in recent years: Intuitive Forward looking Focus on tail events Complement to statistical measures Stress testing is a core part of business area and group reporting. Stress measures are used: As a key management tool to identify risk profiles and potential concentrations To set risk limits For liquidity, business and capital planning purposes To understand vulnerabilities 4
Scenario-consistent Approach across Risk Types Scenario Portfolio Risk Drivers Stress Calculation Results and Aggregation Macroeconomic view Stress Scenario - Severe Global Recession 2009 2010 2011 Q1'09 Q2'09 Q3'09 Q4'09 t1 t2 t3 t4 t5 t6 t7 t8 Stress Scenario - USD Crash: Macroeconomic Indicators Real GDP growth (%yoy) United States 2009 2010-2.7 2011-2.5 Q1'09 1.0 Q2'09-3.3 Q3'09-3.8 Q4'09-2.6 t1-1.0 t2-1.5 t3-3.0 t4-3.5 t5-2.0 t6-0.5 t7 0.5 t8 1.5 2.5 Stress Scenario - Global Depression/Deflation: Macroeconomic Indicators Japan -5.1-3.6-0.1-8.6-6.0-4.7-1.0-2.5-3.5-4.5-4.0-2.0-0.5 0.5 1.5 Real GDP growth Eurozone (%yoy) -4.0-2.1 0.5-5.0-4.8-4.1-2.0-1.0-1.5-3.0-3.0-1.5 0.0 1.5 2.0 2009 2010 2011 Q1'09 Q2'09 Q3'09 Q4'09 Q1'10 t1 t2 t3 t4 t5 t6 t7 United StatesUnited Kingdom -2.7-1.4-4.8 2.1-3.3-3.3 0.3-3.8-5.2-2.6-5.8-1.0-5.1-1.0-3.0-2.0-2.0-2.0-4.0-0.5-4.0 1.0-3.0 2.0-1.5 2.5-0.5 3.01.0 2.0 Japan Switzerland -5.1-3.1-1.5 0.1-1.9-8.6 0.4-6.0-1.7-4.7-2.3-1.0-1.5-1.5-0.5-2.5-1.0-4.5-2.0-4.0-2.5-2.0-2.0 0.0-1.0 1.0.0 1.51.0 1.5 Real GDP growth (%yoy) Eurozone Germany -4.0-1.6-4.7 0.6-2.1-5.0 0.6-4.8-6.7-4.0-5.8-2.0-4.8 0.0-1.5-0.5-1.0-3.0-2.0-3.0-3.0-1.5-2.5 0.5-1.0 1.50.5 2.01.2 1.6 United States -2.4-3.3-4.5-3.3-3.8-2.6 0.1 1.0-2.0-5.0-7.0-6.0-5.0-4.0-3.0 United Kingdom France -4.8-2.1-2.3 1.4-1.8-5.2 0.5-5.8-3.5-5.1-2.8-3.0-2.3-1.5-0.5-3.0-1.0-2.5-1.5-1.5-2.5 0.0-2.0 1.0-1.0 2.0.5 2.51.0 1.5 Japan -5.2-3.4-4.5-8.4-6.0-4.9-1.4 0.5-2.0-5.0-7.0-6.0-5.0-4.0-3.0 Switzerland Brazil -1.5-1.4-0.2 0.70.1-1.7 2.0-2.3-1.8-1.5-1.7-0.5-1.5 0.04.0-1.0 3.5-2.5-0.5-2.0-1.5-1.0-1.0 0.50.0 1.5 1.5 1.82.5 4.0 Eurozone -4.0-3.6-4.5-5.0-4.9-4.1-2.2-0.5-2.0-5.0-7.0-6.0-5.0-4.0-3.0 Germany Russia -4.7-1.6-8.2 0.8-2.5-6.7 4.3-5.8-9.8-4.8-10.9-1.5-8.9 0.0-3.0-1.0 5.0-3.0-1.0-2.5-7.0-1.0-7.0 0.72.0 1.5.0 2.05.0 5.0 United Kingdom -4.9-3.5-4.5-5.3-5.9-5.3-3.1 0.0-2.0-5.0-7.0-6.0-5.0-4.0-3.0 France China -2.3-1.3 8.1 0.86.1-3.5 6.9-2.86.1-2.37.9-0.5 8.9 0.09.5-0.5 8.0-2.56.5-2.05.5-1.04.5 0.75.0 1.56.0 2.07.5 9.0 Switzerland -1.5-3.4-4.5-2.0-2.4-1.4 0.0 0.5-2.0-5.0-7.0-6.0-5.0-4.0-3.0 Brazil Turkey -0.2 1.0-5.6 2.5-0.3-1.8 1.4-1.7-13.9-1.5-7.1 4.0-3.3 4.02.0 0.5 6.0-0.51.0 0.0-3.0 1.0-5.0 2.5-2.0 2.50.5 4.02.5 4.5 Germany -4.9-3.5-4.5-6.7-5.8-4.8-2.4 0.0-2.0-5.0-7.0-6.0-5.0-4.0-3.0 Russia World -8.2-1.6-2.0 4.9-0.8-9.8 2.1-10.9-3.8-8.9-3.1-3.0-1.9 5.50.8 0.0 1.5-6.0-0.6-6.0-2.2 3.0-1.9 5.0.2 5.51.6 6.02.8 3.8 France -2.2-3.4-4.5-3.4-2.8-2.3-0.3 0.3-2.0-5.0-7.0-6.0-5.0-4.0-3.0 China 8.1 6.8 7.5 6.1 7.9 8.9 9.5 8.5 7.0 6.0 5.5 6.0 6.5 8.0 9.5 Brazil -0.2-2.0-4.0-1.9-1.7-1.4 4.3 4.5-1.0-5.0-6.5-6.0-5.0-3.5-1.5 Turkey Consumer price inflation (%yoy) -5.6 0.5 2.3-13.9-7.1-3.3 2.0 6.5 1.5-2.0-4.0-1.0 1.5 3.5 5.0 Russia -8.4-6.6-9.0-9.8-10.9-8.9-3.8 4.5-4.0-12.0-15.0-15.0-9.0-7.0-5.0 World United States -2.0 0.0-0.3 2.81.2-3.8 1.2-3.10.0-1.9-1.2 0.8-1.6 2.11.5 0.3 2.0-1.41.3-1.11.0 1.00.5 2.50.7 3.41.0 4.21.3 1.6 China 8.5 5.4 2.3 6.2 7.9 9.1 10.7 11.9 5.5 2.5 1.5 1.5 2.0 2.5 3.0 Japan -1.5-1.3-0.6-0.1-1.0-2.2-2.5-1.5-1.5-1.2-1.0-0.8-0.7-0.5-0.3 Turkey -5.2-5.0-8.4-13.4-7.3-3.5 4.1 3.0-3.0-8.0-12.0-11.0-9.0-7.5-6.0 Consumer price Eurozone inflation (%yoy) 0.3 0.1 0.5 1.0 0.2-0.4 0.3 1.0 0.2-0.3-0.5-0.2 0.3 0.7 1.0 World -1.9-2.3-4.1-3.8-3.1-1.9 1.4 3.1-1.0-4.7-6.7-6.0-4.5-3.4-2.3 United StatesUnited Kingdom -0.3 1.92.1 2.71.3 0.0 1.4-1.23.0-1.62.1 1.5 1.5 2.21.7 1.62.0 1.81.3 2.01.0 2.30.8 2.61.0 2.81.2 3.01.6 1.8 Japan Switzerland -1.5-1.9-0.5-1.5-0.5-0.1-0.2-1.00.0-2.2-0.7-2.5-1.0-1.6-0.2-1.8 0.3-2.0-0.5-2.2-0.8-2.0-1.0-1.7-1.0-1.3-0.5-1.00.2 0.5 Consumer price inflation (%yoy) Eurozone Germany 0.3-0.30.3-0.4-0.1 1.0 0.4 0.2 0.9-0.4 0.3 0.3-0.3 0.9 0.4-0.2 0.5-0.8 0.0-1.2-0.5-1.0-0.3-0.6-0.1-0.2 0.2 0.2 0.5 0.8 United States -0.4 1.7-0.4 0.0-1.2-1.6 1.4 2.3 2.3 1.5 0.7 0.0-0.5-0.5-0.4 United Kingdom France 2.1 1.9 0.0 2.90.1 3.0 0.5 2.10.6 1.5-0.2 1.7-0.4 2.20.2 1.5 0.6 1.80.2 2.1-0.3 2.5-0.2 2.80.0 3.00.2 3.20.6 1.2 Japan -1.4-1.9-2.9-0.1-1.0-2.2-2.1-1.4-1.7-2.1-2.5-2.8-3.0-3.0-2.7 Switzerland Brazil -0.5-0.8 4.9-1.03.5 0.0 3.9-0.75.8-1.05.2-0.2 4.4 0.24.2-0.7 4.0-1.23.6-1.53.2-1.43.0-1.13.3-0.83.6-0.54.0 4.5 Eurozone 0.3 0.1-0.9 1.0 0.2-0.4 0.4 0.9 0.3-0.3-0.6-1.0-1.0-0.8-0.6 Germany Russia 0.3-0.611.7-0.6 6.9 0.9 7.9 0.3 13.7-0.3 12.4 0.4 11.4 0.4 9.2-0.4 7.5-1.0 7.0-1.3 6.5-1.1 6.5-0.8 7.0-0.4 7.5 0.0 8.0 9.0 United Kingdom 2.1 2.1 0.0 3.0 2.1 1.5 2.1 3.0 2.5 1.8 1.2 0.5 0.0-0.2-0.2 France China 0.0-0.4-0.8-0.5-1.3 0.6 0.3-0.2-0.6-0.4-1.5 0.2-1.3 0.50.3-0.2-1.0-0.8-1.2-1.2-1.5-1.0-1.5-0.7-1.0-0.30.0 0.20.7 1.3 Switzerland -0.5 0.3-1.2 0.0-0.7-1.0-0.2 1.1 0.4 0.0-0.5-1.0-1.3-1.3-1.2 Brazil Turkey 4.9 3.7 6.3 4.15.8 5.8 5.3 5.28.4 4.45.7 4.2 5.3 4.05.7 3.7 6.5 3.56.0 3.45.5 3.65.0 3.85.0 4.25.2 4.75.4 5.7 Germany 0.4-0.1-1.3 0.9 0.3-0.3 0.4 0.7 0.3-0.5-0.9-1.3-1.4-1.3-1.1 Russia World 11.7 7.1 1.7 8.31.4 13.7 1.8 12.42.5 11.41.4 9.2 0.9 7.52.0 7.2 2.0 6.81.5 7.01.2 7.20.9 7.81.2 8.51.7 9.52.1 2.5 France 0.1 0.5-0.9 0.6-0.2-0.4 0.4 1.3 0.6 0.2-0.3-0.7-1.0-1.0-0.8 China -0.8-0.5 1.7-0.6-1.5-1.3 0.3-0.8-0.5-0.5 0.0 0.8 1.5 2.0 2.5 Brazil 4.9 4.2 3.1 5.8 5.2 4.4 4.2 4.9 4.4 4.0 3.6 3.3 3.0 3.0 3.2 Turkey 6.3 6.0 6.2 8.4 5.7 5.3 5.7 6.5 6.2 5.8 5.5 5.7 6.0 6.3 6.6 Russia 11.7 6.2 4.6 13.7 12.4 11.4 9.2 7.2 6.8 6.0 5.0 4.5 4.5 4.5 4.7 World 1.7 1.7 2.4 2.5 1.4 0.9 2.0 2.1 1.6 1.5 1.5 1.8 2.2 2.6 3.0 China -0.7 2.3 0.7-0.6-1.5-1.3 0.7 2.5 2.8 2.2 1.5 1.0 0.5 0.5 0.7 Turkey 6.2 7.9 5.2 8.4 5.7 5.3 5.7 9.3 8.7 7.5 6.0 5.5 5.0 5.0 5.2 World 1.7 2.2 0.6 2.5 1.4 0.9 2.1 2.9 2.7 2.0 1.3 0.8 0.5 0.5 0.7 Market view FX IR EQ PM Spread Credit Market Business PD LGD EAD Spread IR FX EQ Fee income Trading income Interest income Today MtM PD forecast y Exposure Profile Stress MtM time Stressed risk measures Results drill-down e.g. per country, sector, rating, Aggregate measures across risk categories or business units Impact on P/L, RWA and Tier1 Capital 5
No One Size Fits All One framework Multiple components Metric Analysis Time Horizon Granularity Complexity Portfolio specific stress (LU, RE, Lombard, ) Group wide (all divisions, risk categories, portfolios, ) Reverse Stress Historic Event Analysis, Quantitative Forecasts, Expert Judgment Needs to be flexible enough to run ad-hoc scenarios quickly Annual, quarterly, overnight, Connect shocks with inherent time horizon (e.g. to adjust market risk scenarios for liquidity) Position / Counterparty / Portfolio Level Ideally flexible to aggregate on any desired level Take simplifying assumptions Focus on big items Scenario Must be credible Shock the "right" things not thousands of risk factors 6
Scenarios Scenarios stipulate economic variables, market indicators and risk factors for all major economic and financial variables in major economies. Quarterly "Think Tank" (Research, Business, Risk Control) Set of scenarios Forward-looking view, inspired by historical events Reflection of different time horizons Short-term: Liquidity Adjusted Stress scenarios Longer term: macro-economic variables defined over a 2-year time period Regular updates Reflection of current state of the economy and current outlook Changing environment vs. changing inputs 7
Approach for Credit Risk STEP 1 Sensitivities STEP 2 Application to Stress Scenario STEP 3 Bottom up stress loss calculation Determine how default / loss indicators react to macroeconomic variables Make aggregate forecasts under macroeconomic Stress Scenario Calculate stress results based on different measures and metrics Default Rate ( 0 T T ) T T T T 3m ) 3m 6 ( m 6 m 9 ( m ) 9 m 1 ( y ) ( 1y 1.5 y ) ( 1.5 y 2 y ) Stressed default rate (annualized) RWA vs. Stress Loss Expected Loss, ETL Original PD 10d 3m 6m 9m 1y 1.5y 2y time LGD T T ) T T T T 0 3 ( m ) 3m 6 ( m 6 m 9 ( m ) 9m 1 ( y ) 1y 1.5 ( y ) 1.5 y 2 ( y ) LGD Profile Today's LGD StressPD StressLGD StressEAD 10d 3m 6m 9m 1y 1.5y 2y time Exposure ( 0 T T ) T T T T 3m ) 3m 6 ( m 6 m 9 ( m ) 9 m 1 ( y ) ( 1y 1.5 y ) ( 1.5 y 2 y ) Exposure Profile Traded Products Today's exposure Exposure Profile Mortgage 10d 3m 6m 9m 1y 1.5y 2y time 8
Approach for WMSB Real Estate Financing Client's Equity Key Metric: Liabilities / Assets Client's Affordability Key Metric: Cost vs. Income Assets Real Estate Other assets Liabilities Mortgage Cost Actual Interest Cost House maintenance Other fix cost Income Salary (self-occupied) Rental Income Equity Savings Real estate price change Systematic risk drivers Interest rate changes Unemployment, Rental price declines 9
Liquidity-Adjusted Stress Test for Market Risk Liquidity-adjusted scenarios form the Market Risk element of Combined Stress Testing and are designed to calculate the market risk loss (or gain) before risk can be neutralized. Scenarios are applied across all trading book exposures and positions are "revalued" under the scenario using fair value shocks; resulting change in valuation is then considered as the "Stress loss". Scenarios are "liquidity adjusted" in that they include prolonged holding periods for different risk factors/position types reflecting the varying degrees of difficulties in offsetting/eliminating them: Holding periods are based both on quantitative and qualitative analyses and reflect the time expected to neutralize the exposure to the relevant factor in a given scenario Example: US Govt Bond (shorter holding period), EM Govt Bond Sub-Investment Grade (longer holding period) Scenario-consistent shocks are incorporated as a precursor of longer term macroeconomic events according to the respective scenario specification. 10
Usage of Combined Stress Test Combined Stress Test results are seen as a complement to the Economic Capital model. Results are reported in monthly Risk Reports Overall risk profile monitored on the basis of Stress losses development Focus on developments per risk category (credit, market, investment, operational, ) at Group level Contributions per business division to Stress losses at Group level, incl. breakdown over risk categories Risk Appetite for the bank is defined by a set of criteria, of which some are related to Combined Stress outcomes. Combined Stress is also used to assess impacts of Business Plan, e.g. Business Initiatives. Combined Stress is used as input for capital allocation and capital planning. The methodology is also used in communication with regulators related to risk assessments (Building Blocks Analysis, Loss Potential Analysis). 11
Section 2 Challenges
Challenges Processes and Governance Integration in business decisions Relation to portfolio specific stress tests Definition of scenarios Scenario severity perception Pro-cyclicality of stress scenarios and impact Capture of second-round effects/interdependencies Regulatory requests Higher granularity vs. reducing reporting timelines Implementation Efficient implementation of calculation and reporting process Flexibility Multiplicity of data sources Size of input files 13
High Granularity For largest portfolios calculations are done bottom up Input files with very fine granularity needed Calculations done on millions of positions Calculations done based on several characteristics Aggregation of results to provide multiplicity of views Increasing demand for finer granularity in results Finer breakdown of stress impact Resulting in materially larger input files for following calculations Increasing required number of calculation and simulation steps Rising demand for fast reaction times internally as well as from the regulator Increasing number of scenarios 14
Implementation Unique setup with development, prototyping, production implementation, reporting and analysis within the team of quantitative analysts. Powerful, flexible, and scalable IT platform is essential. IT platform has to allow for integrated implementation of diverse risk categories with a multiplicity of data sources. Large degree of automation needed to ensure sufficient time for development and analysis. IT platform has to enable process conform with Audit and operational risk requirements. 15
Section 3 Q & A
This material is for distribution only for the purpose of this conference and is only for informational purposes. Any opinions expressed in this material are the author's and may differ or be contrary to opinions expressed by other people, business areas or groups of UBS. No representation or warranty is provided in relation to the accuracy, completeness or reliability of the information contained herein. 17
Contact Information Dr. Alexandra Hansis Stockerstrasse 64 Postfach 8098 Zurich +41-44-234 50 69 (internal: 1923-45069) alexandra.hansis@ubs.com www.ubs.com 18