ERM@Hannover Re Practical experience and key aspects for capital requirements Eberhard Mueller Managing Director Group Risk Management Chief Risk Officer and Chief Actuary JPM Insurance Seminar London, 24 April 2009
Volume-league table WE ARE AMONG THE TOP REINSURERS IN THE WORLD 2007 figures in million USD 1) Rank Group Country GWP NPW 1 Swiss Re CH 30,673 27,872 2 Munich Re D 29,843 28,439 3 Berkshire Hathaway 2) USA 17,952 17,398 4 Hannover Re D 12,327 10,779 5 Lloyd's 3) GB 10,361 7,990 6 SCOR F 7,106 6,501 7 London Re CDN 6,133 5,155 8 RGA Re USA 5,371 4,909 9 Transatlantic Re USA 4,283 3,953 10 Everest Re BDA 4,078 3,919 11 Korean Re ROK 3,890 2,775 12 Partner Re BDA 3,810 3,757 13 XL Re BDA 3,406 2,812 14 Aegon NL 2,462 2,173 15 Odyssey Re USA 2,283 2,089 1 1) Source: A.M. Best 2) GenRe Group; Berkshire Hathaway Re Group (National Indemnity) 3) 66 syndicates (as of 2007)
GENERAL STRUCTURE Internal steering and control system Board Overall responsibility of the board Supported by GCS (Controlling) et.al. steering Unit Businesssegment RISK Risk Committee CRO supported by GRM (Risk Mgmt.) et.al. "controls" Steering / P&L responsibility (incl. controls within processes) "monitors" Risk steering and control (independent risk control function within processes) Process independent control (on behalf of board) Group Auditing 2
CENTRALISED GROUP RISK MANAGEMENT Reporting structure CEO (Chair) CFO CEO Life COO Non-Life Chief Controller CRO Executive Board Group Risk Committee RC Risk hierarchy 1. Reserving risk 2. Exposure risk 3. Mispricing risk 4. Asset risk 5. Operational risk Group Risk Management headed by Chief Risk Officer Quantitative risk management (DFA-Model) Qualitative risk management (Risk Cockpit) Group Auditing / Corporate Developmt Internal control system 3
STARTING POINT: RISK MAP 26 individual sources of risk Liability risk Asset risk Operating risk U/W risk life U/W risk non-life Market risk Credit risk Operational risk Business risk Catastrophe risk Premium risk (incl. mispricing risk) Equity risk Corporate bonds Regulatory and compliance risks Reputational risk Longevity risk & mortality risk Catastrophe risk Interest rate risk Settlement risk Loss of key personnel Balance sheet risk (OCI, collaterals) Morbidity & disability risk Lapse risk Reserving risk Credit risk Spread risk Real-estate risk Insufficient systems (actuarial, controlling, accounting) Loss of infrastructure Strategic risk Emerging risk Reserving risk Foreign exchange risk Insufficient data quality Credit risk... addresses the main risk drivers of Hannover Re 4
Catastrophe risk STRESS TEST EVOLUTION Lessons learned from Katrina/Rita/Wilma in 2005 Limits for catastrophes redefined Maximum annual aggregates for all occurrences - 100y scenario not to exceed (Non-Life EBIT + Delta), Delta set annually by board - 250y scenario not to exceed 1.3 times 100y scenario Capacity cascaded down to occurrences on clear defined RDS* scenarios incl. all lines of business Applications for new capacities have to fulfil 2 conditions - No violation of above mentioned risk appetite - Value creation (after serving bound capital) Measurement against limits Modelled business: Vendor model results with own calibration, if necessary Not modelled business: Upgrossing according to RDS* scenarios 5 * Realistic Disaster Scenario
Catastrophe risk CLEAR BREAKDOWN OF RISK APPETITE Limits exist from group risk model down to treaty departments Global risk appetite for all risks (Group risk DFA model) EBIT risk tolerance for 6 different return periods Maximum loss of EBIT: According to limits Global NatCat (All nat. cat. perils, annual aggregate, net exposure) EBIT risk tolerance for 100y/250y Maximum loss of EBIT: 100y: planned Non-Life EBIT ± Δ 1) 250y: (planned Non-Life EBIT ± Δ 1) ) 1.3 Local NatCat (Selected realistic disaster scenarios, single event, net exposure) EBIT risk tolerance for realistic disaster scenarios Maximum loss of EBIT: RDS 2) : planned Non-Life EBIT ± Δ 1) Capacities per Treaty Department (TD) (For over 250 perils, return periods and maximum liabilities, annual occurrence, gross exposure) Maximum gross loss per Treaty Department Depending on individual peril 6 1) Δ is reviewed and redefined by Executive Board annually based on current market situation 2) Realistic Disaster Scenario
Catastrophe risk 2008 UTILISATION OF EXPOSURE RISK APPETITE Well below limits Planned Non-Life EBIT Planned Group EBIT Utilisation: as percentage absolute EUR 764 m. EUR 995 m. EUR 1,300 m. Global NatCat (All NatCat perils, annual aggregate, net exposure) 100 y 250 y 90% EUR 778 m. Limit: EUR 864 m. 87% Limit: 1,123 m. EUR 977 m. Local NatCat (Selected realistic disaster scenarios, single event, net exposure) 74% California MW 7.8 EQ Limit: EUR 864 m. 46% EUR 642 m. Miami cat. 5 hurricane EUR 395 m. 56% Gulf of Mexico cat. 5 hurricane 35% EUR 484 m. Europe wind EUR 300 m. 96% Greater Tokyo MW 8.0 EQ EUR 829 m. 7
STARTING POINT: RISK MAP 26 individual sources of risk Liability risk Asset risk Operating risk U/W risk life U/W risk non-life Market risk Credit risk Operational risk Business risk Catastrophe risk Premium risk (incl. mispricing risk) Equity risk Corporate bonds Regulatory and compliance risks Reputational risk Longevity risk & mortality risk Catastrophe risk Interest rate risk Settlement risk Loss of key personnel Balance sheet risk (OCI, collaterals) Morbidity & disability risk Lapse risk Reserving risk Credit risk Spread risk Real-estate risk Insufficient systems (actuarial, controlling, accounting) Loss of infrastructure Strategic risk Emerging risk Reserving risk Foreign exchange risk Insufficient data quality Credit risk... addresses the main risk drivers of Hannover Re 8
Market risk STRESS TEST EVOLUTION Lessons learned from financial crisis 2008/2009 Recalibration of economic scenario generator Vendor ESG in use: Barrie & Hibbert (B&H) Current decision: Follow B&H s recalibration Recalibration of correlation assumptions Work in progress Major source of diversification at reinsurers still holds: diversification between insurance risks from natural catastrophes and capital market risks Already in the model before financial crisis: dependency between insurance risk and counterparty default risk (higher combined ratios trigger higher default rates) Adjustments to investment guideline: Significant reduction of exposure per debtor Further diversification of portfolio Covered Bonds/ABS/MBS subject to improved accumulation control 9
Market risk STRESS TESTS FOR TURBULENT MARKETS Portfolio Scenario Change in market value in m. EUR 31 Dec 2008 31 Dec 2007 Equities Prices -10% -2-200 Equities Prices -20% -5-400 Equities Prices -30% -7 - Fixed-income securities Yield increase +50 bps -346-322 Fixed-income securities Yield decline -50 bps +356 +312 Fixed-income securities Yield increase +100 bps -684-622 Fixed-income securities Yield decline -100 bps +720 +645 Unrealised gains/losses (Available for sale) 31 Dec 06 31 Dec 07 31 Dec 08 Fixed income (119) (103) 102 Equities 212 191 1 Total unrealised gains and losses 93 88 103 10
Sensitivity analysis of the Market Consistent Embedded Value (MCEV) 2007 EFFECT ON FORECAST MCEV 1) Domestic operations Before consolidation Foreign operations Total After consolidation Total MCEV (basic) in m. EUR 1,488.1 995.8 2,483.9 1,715.1 Sensitivities to economic assumptions Discount rate +100 bps -1.4% +2.2% +0.1% +0.1% Interest rate environment -100 bps +0.7% -3.1% -0.8% -1.2% Value local currency + 5% 2) -0.8% -0.5% -0.7% -1.0% Sensitivities to non-economic assumptions Expenses +10% -0.4% -1.0% -0.7% -0.9% Lapse +10% -1.0% +1.1% -0.2% -0.2% Mortality/Morbidity +5% -4.8% -12.5% -7.9% -11.4% Mort./Morb. +5%, life/disab. bus. only -5.2% -12.8% -8.3% -12.0% Mort./Morb. +5%, annuity bus. only +0.4% +0.3% +0.4% +0.6% 11 1) After tax, before minorities 2) Treaties written in foreign currencies
Diversification across Non-Life business lines and regions REDUCES CAPITAL REQUIREMENTS BY OVER 40% 5,000 4,500 in m. EUR 784 4,559 4,000 3,500 3,000 927 300 134 46% diversification 2,500 653 2,000 1,500 1,000,500 836 377 201 347 2,445 Effective capital requirement 0 USA Germany Marine Aviation Credit, surety & pol. risks Global cat. Global fac. Structured products Others Non-Life Risk-based capital (RBC) 12 As at December 2008 According to internal risk model 99.5% Value at Risk (VaR)
REQUIRED CAPITAL HR GROUP PER BUSINESS SEGMENT Major changes due to recalibration of credit model 000 000 000 000 000 000 000 0 in m. EUR 1,730 5,952 663 3,559 3,537 Non-Life R/I Life/Health R/I Assets HR Group Rise in required capital mainly due to higher credit and spread risks caused by recalibrations following the financial crisis Now, credit risk has a higher impact on the asset as well as on the liability side; Diversification decreases Operational risk included according to the current Solvency II standard model Risk-based capital (RBC) 13 As at December 2008 According to internal risk model 99.97% Value at Risk (VaR); equivalent to "AA" S&P rating
Capital requirements ACQUISITION OF ING PORTFOLIO WITHOUT MAJOR IMPACT 6 in bn. EUR 6.0 1.7 0.7 Sale of equities reduces required capital, but countered by increased capital for fixed income 6.0 1.7 0.7 Reduction of return on assets reduces available capital Increased required capital due to acquisition of US-life portfolio Because of diversification only slight increase of required capital U.S. Life acquisition utilises available excess capital Attractive opportunities in Non-Life reinsurance Investment strategy focusses on stable returns 3.6 3.6 6.4 3.6 3.5 6.3 Hannover Re's portfolio is conservative and broadly diversified 0 31.12.2007 Dec 07 31 Dec 08 31 Jan 09 Required capital Non-Life Required capital investments Available capital Required capital Life/Health Total required capital after diversification 14
Dynamic Financial Analysis Example CAPITAL MONITOR Internal capital model provides market-consistent answers for steering 9 8 7 6 5 4 3 2 1 0 in bn. EUR 5.3 IFRS 1) 6.4 3.6 Economic capital model 2) 5.3 2.4 Regulator (Solvency I) 8.2 7.2 Rating agency 2) Available capital Required capital Economic capital = Market Value of Assets - Market Value of Liabilities Market consistent: Use market values where available mark to market Use market value models otherwise mark to model 15 1) Policyholders' surplus as of 31 December 2007 2) AA-rating equivalent capital requirements, current assumption for internal model: 99,97% VaR.
LESSONS LEARNED FROM THE CRISIS Stuff happens... (credit crunch, Gustav and Ike in one year) Increased mistrust from non-actuaries in actuarial models Perception of failed models rather than failed parameters Concerns about diversification effects Increased demand for conservatism (safety margins in addition to model results) DFA-models combine different kinds of "randomness" Natural catastrophes: "really random" events, underestimated consequences Economic scenario generators: "assumed random" events, underestimated uniform behaviour of market participants (systemic risk) Different time horizons for decisions (u/w: 1 year, capital market: daily) Necessity to push hard for constructive alternatives going forward Underlining the purpose of models: decision aid, not decision maker Underlining the importance of key assumptions (spread of stocks over bonds???) Underlining the difference between "recalibration" and "remodelling" Broaden the scope of thinking (combination of models and scenario analysis) Follow (and contribute to) best practices (e.g. CRO-Forum) CRO-Forum publications (e.g. liquidity mgmt., financial crisis) 16
SOME COMMENTS ON QUESTIONS RAISED Please note the structural difference between banks and insurers! Undue conservatism for capital requirements through add-ons via stress tests should be avoided. Capital models should try to be "true best estimates". Local regulators should recognise ability of multinational organisations to transfer capital to where it is needed (group support regime) rather requiring local entities to be treated on a standalone basis Current Solvency II "compromise" (postponing group support) is dissatisfying Wrong incentives for prices and domiciles Crisis situations are best managed under the guidance of a lead supervisor Crisis gave reason to recalibrate economic scenario generator and correlations but not to "forget about the models"! Recalibration of return periods for observed scenarios Increased capital requirements for credit and spread risks Asset risk in the crisis is quite different between banks and (re)insurers Liquidity problem mitigated through advance payments of premiums No. 1 risk currently through accounting requirements (OCI) Risks which can hardly be managed by providing capital must be managed through suitable ERM-systems (Pillar II) 17
OUR INVESTOR RELATIONS TEAM Hannover Re Karl-Wiechert-Allee 50 30625 Hannover, Germany Tel. + 49/5 11/56 04-0 (Switchboard) Fax: + 49/5 11/56 04-16 48 (direct line) Stefan Schulz Klaus Paesler Associate Director Senior IR Manager Corporate Communications Corporate Communications +49/5 11/56 04-15 00 +49/5 11/56 04-17 36 stefan.schulz@hannover-re.com klaus.paesler@hannover-re.com Daniela Gissinger Deputy IR Manager Corporate Communications +49/5 11/56 04-15 29 daniela.gissinger@hannover-re.com