new Basel framework Philipp Rickert Partner Dr. Dominik D. Lambrigger Manager g ETH Riskday, 9 September 2011 What makes a Traders life miserable? Political challenges Uncertainty in US and Europe S i d bt i i Sovereign debt crisis Global economy slow down Market behavior Increased participant mistrust Increased competition (chasing the flow) Client behavior Client behavior Cash is King Low volumes, flight to quality, risk averse R l t Regulatory response Risk taking becomes disproportionally capital intense and expensive Qualitative requirements increase (IT, processes, people) All that makes the Traders life miserable, because less volume and less risk combined with higher volatility and higher production costs results in bad P&L and that means LOWER BONUS! 1 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. LOWER BONUS! Basel III NOT THIS BASEL THIS BASEL! 2 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Why Basel III Regulators Views Why the Financial Crisis Excess of cheap money and procyclicality of globally important banks Excessive leverage in the global financial system Excessive leverage in the global financial system Too little capital of insufficient quality (leverage between 1 to 33 and 1 to 100) Inadequate liquidity buffers Purpose Improve the banking sector's ability to absorb shocks arising from financial and economic stress, whatever the source Improve risk management and governance Strengthen banks' transparency and disclosures g p y Response Quality of capital tangible equity Quality of capital tangible equity Coverage of risk focus on trading book exposures Much higher levels of capital 3 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Global liquidity standards Capital under Basel Requirement vs Eligibility Required Capital .
kOperational Risk Available Capital Basel 2 O p . R i s kOperational Risk k Fixed assets Derivatives Tier 3 Abolished Basel 2 C r e d i t R i s k B a n k i n g
B o o kDerivatives Secured Finance Retail Loans Tier 2 Restrictions Basel 2.5: Trading Book B Commercial Loans: Banks CVA Tier 2 Tier 1 Trading Book R i s k B o o k CRM Securitizations Deductions Basel 3 M a r k e t
R T r a d i n g B Stressed VAR IRC VAR Tier 1 4 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. VAR Paul Embrechts: Paul Embrechts: Always be scientifically critical, Always be scientifically critical, socially honest and adhere to the highest ethical principles, especially in the face of temptation which will come! 5 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Modeling market risk (Value-at-Risk) Key elements Market risk charge = C x VaR 99% (V), C 3 10-day holding period V is the value of the trading book portfolio Risk Weighted Assets CVA B III V = 10-day change of V Challenges Securitization CRM Complexity of the portfolio (and hence V ): need sophisticated front office pricing models Comprehensive scope of risk factors Historical simulation vs. Monte Carlo IRC Stressed VaR B II.5 Historical simulation vs. Monte Carlo Responsiveness of VaR Data quality Complexity of risk infrastructure Credit Risk Op Risk B II p y Backtesting VaR Market Risk 6 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. The Greeks A Taylor expansion of V reduces the problem to the calculation of: The risk factor changes S and their volatilities (for simplicity assume that there is only 1 risk factor and no specific risk) The sensitivities (Greeks) ) ( 1 ) ( 2 2
S V V S V S V ... ) ( 2 ,...) , ( 2
S S S S S V
Delta Vega Gamma
Historical Risk Factor Changes Front Office Pricing Models and Cross-Gamma, Vanna, Volga, etc. g (stochastic) g (deterministic) Risk factor Volatility Value Delta Vega g Delta Gamma Vanna Vega Vanna Volga Gamma Speed Zomma Volga Ultima 7 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. For highly non-linear products Taylor might not be appropriate Challenges in modelling market risk (1/4) Banks tend to use full revaluation of the whole portfolio for VaR calculation and stress testing. Greeks & Taylor Expansion Partial revaluation Full reval C o m p l e x i t y Most banks use historical simulation based on a 1-5 year dataset Looking at the past instead of the future Looking at the past instead of the future Some banks have the crisis in the VaR Others have a strong desire to exclude the crisis from the dataset (due to the introduction of stressed VaR and a double-counting of the crisis) 8 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Challenges in modelling market risk (2/4) How responsive should VaR be in a crisis? Standard implementation of VaR is typically not very responsive to a crisis, especially if the underlying data is unweighted and is based on several years underlying data is unweighted and is based on several years. How can this problem be addressed? Stressed VaR B l 2 5 ( l t ) Responsive VaR S li b k t l tilit Dual approach Basel 2.5 (see later) Scaling by market volatility Weighting of data 9 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Challenges in modelling market risk (3/4) Backtesting: Check performance of VaR by comparing VaR against PnL Check performance of VaR by comparing VaR against PnL But have 10-day VaR and 1-day PnL?! Two options: Scale 10-day VaR down to 1-day VaR: VaR 1-day VaR 10-day Re-calculate 1-day VaR (costly) 10 / Daily adjusted PnL 1 Day VaR (99%) 10 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Source: Credit Suisse Annual Report 2005 Daily adjusted PnL 1-Day VaR (99%) Challenges in modelling market risk (3/4) How many backtesting exceptions per year are acceptable? N b f ti I i th lti li Number of exceptions Increase in the multiplier 4 or less 0.00 5 0.40 6 0.50 7 0.65 k P(N=k) P(Nk) N ~ Bin(n=250, p=0.01) P(N > 4) = 10.8% 7 0.65 8 0.75 9 0.85 10 or more 1.00* Source: FINMA Market Risk Circular 2008/20 0 0.081 0.081 1 0.205 0.286 2 0.257 0.543 3 0.215 0.758 4 0.134 0.892 5 0.067 0.959 * Any shortcomings must be eradicated without delay, since otherwise the conditions for determining capital adequacy requirements according to the model-based approach will be deemed no longer to be fulfilled. 6 0.027 0.986 7 0.010 0.996 8 0.003 0.9989 9 0 0008 0 9997 Credit Suisse Annual Report 2008: Annual Report 10 - CS: 9 0.0008 0.9997 10 0.0002 0.99995 UBS Annual Report 2008: 0 exceptions - UBS: 1 exception 11 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Challenges in modelling market risk (3/4) Alternative backtesting approach: 2 ... ) ( 2 1 ,...) , ( 2 2 2
S S V V S S V S V
Delta Vega Gamma
Historical Risk Factor Changes (stochastic) Front Office Pricing Models (deterministic) Use realized risk factor changes to calculate V Model-based PnL (risk-based PnL) Backtest model-based PnL against realized PnL Pros: Model deficiencies detected on a daily basis (dont have to wait for a VaR backtesting exception) Cons: Non-market risk factors have to be modelled or need good clean PnL data 12 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Challenges in modelling market risk (4/4) Invest in human capital and their quantitative (high quality) education! 13 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. and their quantitative (high-quality) education! Basel 2.5 Incremental Risk Charge (IRC) Key elements Default and credit rating migration of trading book positions 99.9% VaR, 1-year holding period Fixed Income products (mainly plain vanilla bonds and credit default swaps) Risk Weighted Assets CVA B III credit default swaps) Key inputs are ratings/PDs, LGDs, EADs, credit spreads, migration matrix and liquidity horizon Securitization CRM Challenges Migration risk modelling Constant level of risk (and liquidity horizon): d f lt d/ i t d iti h t b b l d IRC Stressed VaR B II.5 defaulted/migrated positions have to be re-balanced Completeness of positions Risk data capturing and enrichment Model risk (imposed by the regulators?) Credit Risk Op Risk B II Model risk (imposed by the regulators?) Structured credits, credit derivatives Market Risk 15 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Stressed VaR Key elements What-if-the-crisis-comes-in-again-VaR Standard VaR model with todays portfolio But risk factors based on 1-year stressed period (e.g. 2008) Risk Weighted Assets CVA B III 99% VaR, 10-day holding period Challenges Securitization CRM Choice of stressed period Methodology and implementation framework Model integration within existing market risk framework Ch t f k f th t d V R i k IRC Stressed VaR B II.5 Change management framework for the stressed VaR risk parameters Credit Risk Op Risk B II Market Risk 16 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Securitizations and Re-securitizations Key elements Capital charge for (re)-securitization positions in trading book Product scope: CDO, CDO^2, RMBS, CMBS, NTD, etc. Standardized charge: very costly Risk Weighted Assets CVA B III Standardized charge: very costly Corporate CDOs and NTDs can be modelled by CRM (see next slide) Specific risks only, general risk through VaR Securitization CRM Challenges Robustness of methodology IRC Stressed VaR B II.5 gy Completeness of scope Completeness of data feeds Data quality Credit Risk Op Risk B II Market Risk 17 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Comprehensive Risk Measure (CRM) Key elements Risk Weighted Assets CVA B III Capital charge for correlation trading Product scope: synthetic CDOs (index tranche and bespoke tranche), nth-to-default and corresponding hedges (e.g. CDS, index CDS and bonds) Securitization CRM , ) All price risk model: default, rating migration, basis risk (index basis, i.e. basis between index and single names, correlation basis, e.g. basis between index tranches and bespoke tranches), FX (?) and interest rate risk (?) IRC Stressed VaR B II.5 bespo e t a c es), ( ) a d te est ate s ( ) Challenges Credit Risk Op Risk B II Very sophisticated models Large model risk Scope definition C l t f d t f d Market Risk Completeness of data feeds 18 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. And beyond market risk? Dont forget Operational Risk Advanced Measurement Approach (AMA): Does having a large Op Risk capital add-on help to reduce the risk or is it just a regulatory capital add-on? After the crisis people stopped talking about Op Risk Weighted Assets CVA B III After the crisis, people stopped talking about Op Risk (since 2010, regulators start to care about Op Risk again). The new challenges are the old challenges: data quality data quality data quality Securitization CRM quality, data quality, data quality. How to combine external, internal data with expert judgment and scenario analysis? How to estimate correlation between business lines (d it k )? IRC Stressed VaR B II.5 (does it make sense)? Does a 99.9% quantile make sense (dont want to speak about 99.97% economic risk capital)? C dit Ri k Credit Risk Op Risk B II Credit Risk A-IRB and Liquidity Risk Market Risk 19 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Basel 3 Credit Valuation Adjustments (CVA) Key elements Risk Weighted Assets CVA B III CVA is the Market Value of Counterparty credit risk (CCR) Not only default of counterparty, but also change in credit worthiness of counterparty has to be addressed CVA is also a lesson learned from Lehmans default as a Securitization CRM CVA is also a lesson learned from Lehman s default as a major counterparty for OTC derivatives If Bank has an internal VaR model: Advanced CVA All other Banks: Standardized CVA IRC Stressed VaR B II.5 Need to dynamically price CCR Very costly (infrastructure, resources and capital) Credit Risk Op Risk B II Challenges Potentially large capital impact Methodology and implementation framework not straight- f d Market Risk forward Securitizing CVA? 21 2011 KPMG AG/SA, a Swiss corporation, is a subsidiary of KPMG Holding AG/SA, which is a subsidiary of KPMG Europe LLP and a member of the KPMG network of independent firms affiliated with KPMG International Cooperative (KPMG International), a Swiss legal entity. All rights reserved. The KPMG name, logo and cutting through complexity are registered trademarks or trademarks of KPMG International. Thank you! Thank you! Philipp Rickert Partner Dr. Dominik D. Lambrigger Manager Manager Audit Financial Services KPMG AG Badenerstrasse 172 CH-8026 Zrich CH 8026 Zrich prickert@kpmg.com dlambrigger@kpmg.com