Risk management in banks Leoš Souček, Komerční banka MFF UK – March 9, 2012 page 1
Introduction / Czech economic specifics… 1/2 § Private sector indebtedness at the lower end of EU levels ü Indebtedness of corporate sector at 22% of GDP ü Indebtedness of households 30% of GDP ü Dynamic growth of mass retail exposure (above 30% yoy in 20032008) slow-downed in 2010/2011 (6. 6% yoy in 07/2011) § Also public sector indebtedness lower than EU peers § Despite fragile political stability some progress in reforms, which help to secure fiscal sustainability ü CZ at 38. 5% end of 2010 (FR: 82%, DE: 83%, HU: 80%, PL: 55%, SK: 41%, IT: 119%, GR: 143%) ü Budget deficit narrowed to ‑ 4. 7% GDP in 2010 (SK: ‑ 7. 9%, HU: ‑ 4. 3%, GR: ‑ 10. 5%, FR: ‑ 7. 0%, DE: ‑ 3. 3%, AT: ‑ 4. 8%) ü Health care and pension reforms under progress ü S&P raised the LTFC Czech Republic rating to 'AA-' § No concrete plans on EUR adoption § Limited dependence on external financing ü Relative low level of external debt in terms of GDP ü External financing needs covered by FDI and EU funds MFF UK – March 9, 2012 page 2
Introduction / Czech economic specifics… 2/2 § Limited share of private foreign currency debt ü 13. 0% for private sector (in 7/2011, both residents and non residents) ü 17. 4% for corporate, 0. 14% for households (residents only) export-oriented corporates as main users ü 150% coverage ratio by foreign currency deposits (res. only) § Healthy banking sector ü Confirmed by CNB stress tests ü Sustained banking sector profitability since early 2000 s ü Favorable loan-to-deposit ratio at 78% July 2011 ü Strongly capitalized (in end-June, 15. 9% average regulatory capital ratio) § Growth largely relying on external factors … § . . . and on the cyclical industry sectors ü Share of nominal exports on GDP at 80% in 2010 ü 84% of exports concentrated on the EU-27 in 2010 (two predominant partners: Germany 32% and Slovakia 9%) ü Absence of major macroeconomic imbalances - current account worsened to 3. 8% of GDP in 2010 ü Substantial foreign ownership following high FDI inflows ü Industry sectors represents 30% of GDP (vs. 22% in Germany and 12% in France) ü Highly cyclical (predominant car industry and machinery) MFF UK – March 9, 2012 page 3
Introduction / Banking sector As of Zoom on KB: § SG Group member since 2002. § 12/2011 Third largest bank in ČR: ČS (ERSTE Group) ČSOB (KBC Group) § ü Small Business: 20% ü Corporates: 30% ü Municipalities: 40% 783, 3 721, 6 441, 4 483, 5 440, 5 -7, 3 -5, 5 -5, 0 -2, 0 -5, 5 -1, 8 o/w other risks (Gr) -5, 3 0, 0 -3, 2 181 bps (39 bps) 114 bps 36 bps Loan to Deposits 77, 5% 71, 9% 69, 5% LUSR 5, 7% 6, 0% 5, 2% CIR 41. 2% 41, 8% 44, 8% CAR 14, 6% 13, 1% 15, 6% ROE KB’s market share on credit lending: Mortgages: 23% 586, 0 o/w loans Retail developed after 2000. ü 11, 2 12, 3% 18, 2% 17, 3% Cost of Risk (Bn CZK) KB initially corporate bank. § 13, 6 Loans (Bn CZK) about 400 point of sales. 9, 5 Deposits (Bn CZK) about 7 800 employees. ü § KB (SG Group) Cost of Risk (bps) ü Net profit Three key banks (market share at about 70%) MFF UK – March 9, 2012 page 4
Risk Management / Functions & Missions § Credit Risk Management ü ü Individual approach for non-retail (Corporate, Banks, Sovereign) ü § Retail: model based and statistical approach (PD, LGD, EL) Collateral Evaluation (independent on client or distribution channel, on-site visits) Market Risk Management ü § FX, IR, commodity, credit risk, … Monitoring and reporting ü ü § Quality of portfolio / Focus on sensitive sections / Distribution channels / Sensitivity to market (FX, IR, . . ) Back-testing of models Recovery / Collection ü ü § Pre-early collection (-5 DPD - 5 DPD), Soft collection (5 DPD – 90 DPD) Hard recovery (90 DPD +) Operational Risk Management ü Antifraud policy, Insurance, Business continuity plans, estimations of operational losses, … MFF UK – March 9, 2012 page 5
Risk Management / Zoom on KB organization § § Universal Risk Management Function / Op. Risk out of scope. Matrix organization / One of largest Risk Management in the SG Group (330 FTE). SG RISQ KB RISQ A. Viry (L. Souček) Credit Risk Assessment Corporate deal-flow Capital Markets Risks Market risk Capital markets Assets Valuation & Recovery Hard recovery Collateral Evaluation Risk Information Systems Risk databasis Supervision and Measurement Scoring models Monitoring Risk Methodology Credit frauds Functional links with SG RISQ departments MFF UK – March 9, 2012 page 6
Risk Management / History of model development in KB § In-house score-card development since 1998 (IND, SB, Corp, Muni). § SG models used for sovereign and banks since 2002. § KB historical view: ü 1990 – 1997: Score-card developed by analysts (very simple expert models). ü 1997 – 1998: Score-card developed by statisticians (consumer loans, mortgages, corp). ü 2002 – 2003: Models implemented to the central rating system. ü 2001 – 2002: Behavioural scoring model developed (IND, SB). ü 2002 – 2005: Review of models with SG after acquisition. ü 2002 – 2007: Progressive usage of credit bureaus for retail (CBCB, SOLUS). ü 2005 – 2007: Implementation of Ba 2 standards in KB (advanced methods for all credit portfolios). ü 2008 – 2011: Development and implementation of credit fraud prevention. MFF UK – March 9, 2012 page 7
Risk Management / Key risks for the bank CREDIT RISK MONITORING / REPORTING OPERATIONAL RISK 3 DANGERS MARKET RISK RECOVERY NOT INTEGRATED CYCLE LOW UNDERSTANDING MODEL RISK MFF UK – March 9, 2012 page 8
Credit Risk / Key elements PD, LGD, Ea. D § § § Expected Loss (EL) = Ea. D * 1 Y PD * LGD Risk Weighted Assets (RWA) = RW * Ea. D § § § 1 Y PD = Probability of Default during following 1 Y LGD = Loss Given Default Ea. D = Exposure at Default RW = Risk Weight RWA = Risk Weighted Assets RW = Function (PD, LGD, Maturity, Regulatory correlation, Regulatory interval of conf. 99. 9%) Actual Exposure = On B/S + Off B/S Exposure at Default (Ea. D) = On B/S + Off B/S * CCF Off B/S CCF LGD Recovered cash On B/S Non-Default, PD < 100% Default, PD= 100% “ 90 DPD or unlikely to pay” MFF UK – March 9, 2012 Recovery process page 9
Credit Risk / Ability to absorb a loss Bank Capital Market Risk RWA + Credit Risk RWA + Operational Risk RWA FREQUENCY OF LOSSES CAR = § CAR = Capital Adequacy Ratio § Regulatory minimum at 8% § Unexpected Loss as a variation of Expected Loss Probability 99, 9% Expected Loss covered by revenues > 8% SIZE OF LOSSES Unexpected Loss covered by the capital MFF UK – March 9, 2012 Stress Testing Probability 0, 1% Extreme Loss !!! DEFAULT !!! page 10
Credit Risk / Credit portfolios per PD & LGD § Expected Loss (given by PD and LGD) is reflected in pricing. ss Lo d cte pe Ex Client rate: 3, 5% Net margin 1, 0% Expected Loss 0, 5% Cost of funds 2, 0% MFF UK – March 9, 2012 page 11
Credit Risk / Corporates / Rating Model § Individual assessment is prevailing. ü ü § Financial assessment (financial data) Economic assessment (position in market, …) Model rating revised by credit analyst. Financial Rating Behavioural Rating Economic Rating Model Rating Credit analyst Final Obligor Rating MFF UK – March 9, 2012 Rating scales for non-retail Moody’s Aaa Aa 1 Aa 2 Aa 3 A 1 A 2 A 3 Baa 1 Baa 2 Baa 3 Ba 1 Ba 2 Ba 3 B 1 B 2 B 3 Caa 1 Caa 2 Caa 3 Default S&P AAA AA+ AA 1 Y PD 0, 01% 0, 02% Country DE, USA, FR (M) FR (S&P) BE, SI AAA+ A ABBB+ BBBBB+ BB BBB+ B BCCC+ CCC- 0, 03% 0, 04% 0, 06% 0, 13% 0, 26% 0, 50% 1, 10% 2, 12% 3, 26% 4, 61% 7, 76% 11, 42% 14, 33% 20, 44% 27, 25% 100, 00% IT, JP CZ, SK, CN PL RU, BR HU, RO BUL UA GR page 12
Credit Risk / Retail / Granting process § § Maximally automated (95% of approvals) / Maximally parameterized / Maximally centralized. Data collection / Independent verification / Assessment by statistical model. DATA COLLECTION (COMPLEX INFORMATION ABOUT CLIENT) APPLICATION FORM (DEMOGRAPHIC DATA) INTERNAL BANK DATA (BEHAV. DATA) CREDIT REGISTERS (CREDIT HISTORY) TRANSACTION PARAMETERS COLLATERAL CREDIT ANTIFRAUD SYSTÉM (IDENTITY, EMPLOYER, DATA) SCORING MODEL (YES / NO) INSTALLMENT LIMIT (YES / NO) MFF UK – March 9, 2012 COLLATERAL EVALUATION (YES / NO) page 13
Credit Risk / Retail / Scoring Model MAIN DRIVER OF PREDICTING POWER Behavioural Model (data in the bank) Behavioural Models (data in subsidiaries) IN KB SINCE 2002 • Basic Behavioural SM (Bank) • Complex Behavioural SM (Group) • Advanced Behavioural SM • Advanced Application SM IN KB SINCE 2007 Application rating Credit Register Model (data from the register) IN KB SINCE 2006 3 KEY ADVANTAGES J High predicting power. Demographic Model (application form) IN KB SINCE 1998 J Complex assessment. J High flexibility (4 boxes). MFF UK – March 9, 2012 page 14
Credit Risk / Retail / Scoring Model 3 KEY ADVANTAGES 3 KEY RISKS J Fast and easy process L No view of expenditures J High volume of production L Limited assessment J Top quality of production L Change of behaviour 1 200 ths. clients EUR 6 400 M 300 ths. clients EUR 1 200 M 80 ths. clients EUR 32 M 2002 170 ths. clients EUR 400 M 2004 2006 2011 PRODUCTS AO AO, CL, CC MAX LIMIT EUR 400 EUR 2 400 EUR 6 000 EUR 10 000 KB KB GROUP CLIENTS MFF UK – March 9, 2012 page 15
CONTACT RNDr. Ing. Leoš Souček Deputy Head of Risk Management Komerční banka, a. s. Tel: +420 222 435 141 Email: leos_soucek@kb. cz MFF UK – March 9, 2012 page 16
Скачать презентацию Risk management in banks Leoš Souček Komerční banka
8cc25d56d16fd2d2048d3e121bf949ce.ppt