Risk Based Supervision and the Swiss Solvency Test

1. Risk Based Supervision and the Swiss Solvency Test Federal Office of Private Insurance Philipp Keller, Philipp.Keller@bpv.admin.ch Zurich, 1 December 2006

2. Contents • Global Regulatory Tendencies • The Impact of the Old Supervisory Framework • SST Principles • Risk • Valuation • Risk Management • Internal Models

3. Global Regulatory Tendencies There are costs and risks to a program of action, but they are far less than the long-range risks and costs of comfortable inaction John F. Kennedy In the past, insurance supervisors but also insurance companies often were not sufficiently aware of economic reality • The valuation of assets and liabilities was not adequate for an analysis of risk • The artificial smoothing of results often made companies and supervisors inclined to comfortable inaction • An adequate risk quantification was perceived by some as to be too complex and too onerous The financial crisis of 2000-2002 has shown to all that the insurance industry was more exposed than previously thought and both insurers and regulators saw the need for a more adequate, risk based supervisory framework  many regulators (UK, NL, CH,…) and the EU have started initiatives to develop more risk based supervisory models

4. Purpose of Insurance Regulation Insurance is often a long term contract: A policy bought today can be a promise of the insurer to pay at a random future date - up to 50 years later - a random amount. During the contractual period, it is often difficult to sell the product (e.g. only at a loss, replacing a policy might be impossible due to the health state of insured,…) The world about 50 years ago: Churchill was still premier minister Eisenhower finished his first term as US president No man-made object orbited Earth Peak speed of the fastest computer (MIT TX0) was 83kOPS, which is approx. 100bn times slower than today's fasted computer Market Imperfections: Information asymmetry: Policy holders know less about products than insurers, the products are complex and abstract Lack of transparency: Accounting information is often not very relevant to assess the financial situation of a insurer Products are not freely tradable: Once bought, it is often impossible to sell a policy or only at a large loss

5. Current Regulatory Initiatives Current regulatory initiatives are all based on market consistent valuation, quantification of all relevant risks and the use of internal models While some frameworks (Solvency 2, IFRS) will be implemented only in 2010+, they are already influencing the industry’s business model 2000 2005 2010 ICAS* (UK) Solvency 2 SST* (CH) IAIS Implementation Projected Implementation IFRS ICAS: Internal Capital Adequacy Standards SST: Swiss Solvency Test IAIS: International Association of Insurance Supervisors Basel II

6. New Swiss Insurance Regulation • Main Aims • Policy holder protection from fraud • Policy holder protection from the consequences of insurance failures • Functioning and innovative insurance market • Preconditions • Pervasive responsibility culture • People act with integrity • Existence of a risk culture • Transparency • Rule of Law The strategy of FOPI promotes the necessary preconditions to achieve its main aims: • Risk based supervision • Principles based supervision • Enabling competition within the market • Being professional, efficient and transparent

7. Contents • Global Regulatory Tendencies • The Impact of the Old Supervisory Framework • SST Principles • Risk • Valuation • Risk Management • Internal Models

8. Supervision in the Past: Statutory Valuation “The actuarial convention according to which the composition of the assets determines the size of the liabilities is one of the weirdest emanations of the human mind. It's a metaphor - like saying that the advent of jet planes made the Atlantic narrower - and metaphor has a limited place in finance” Speech given by Martin Taylor to the National Association of Pension Funds conference • Discount rate for liabilities was set with reference to an expected asset profit based on past experience • Implicit - often unknown - prudence in liabilities • No explicit valuation of embedded options and guarantees • Amortized cost for bonds • Solvency 1: No capital requirement for market and credit risk • High risk assets resulted in reduction of liabilities • Sales-forces pushed for adding high guarantees to life policies • Foreclosing of investment opportunities due to amortized cost approach for bonds • Cash flow underwriting • Downward spiral when business contracts • Underwriting cycles are exacerbated

9. Old Supervisory Framework • Statutory valuation and Solvency 1 were not adequate for defining risk adequate capital requirements • Statutory valuation is (barely) adequate in a stable economic environment where interest rates do not move and business neither contracts nor grows • As interest rates fell during the 90s, the situation of life insurers worsened • The steering of companies with regard to statutory valuation only led to uneconomic incentives for management • Solvency 1 and statutory valuation were hiding the true economic situation of insurers, leading to a postponement of necessary changes in strategy, making problems much worse when they finally had to be dealt with when share market crashed in 2001/2002 The new risk assessment is based on a market consistent valuation and an appropriate quantification of risk  Swiss Solvency Test The responsibility for the adequate quantification lies with senior management  principles based approach to supervision

10. Contents • Global Regulatory Tendencies • The Impact of the Old Supervisory Framework • SST Principles • Risk • Valuation • Risk Management • Internal Models

11. Rule- vs. Principles-Based Supervision Underlying most arguments against the free market is a lack of belief in freedom itself, Milton Friedman Regulation: A system of laws, decrees, rules, principles, implicit and explicit conventions and expectations, incentives, rewards and punishments, etc. Rule Based Approach Principles Based Approach A system trying to define and micro-manage the insurance market A system promoting a free and liberal market Regulator Regulator The system needs to promote competition, punish collusion and create a level playing field via risk based capital requirements and transparency The complexity grows over time, the system needs to be adapted continuously to new products Liberal Insurance Market Dirigistic Insurance Market The market decides which companies succeed or fail  The “5-year plan” approach to regulation

12. Principles Based Supervision The more laws and order are made prominent, the more thieves and robbers there will be, Lao-tzu Principles based standards describe the objective sought in general terms and require interpretation according to the circumstance Companies tailor their approach such that a clearly stated objective is attained The objective can be attained if the companies interpret principles faithfully Principles based Objective e.g. company specific risk-based solvency assessment = Rule-based Objective The attained result likely deviates from the objective (e.g. a true company specific solvency requirement) since the rules will not capture the specific situation of the insurer sufficiently well A rule based approach does not allow a truly company specific risk assessment (or the set of rules becomes huge and Byzantine)

13. The SST Principles Those are my principles. If you don't like them I have others, Groucho Marx • All assets and liabilities are valued market consistently • Risks considered are market, credit and insurance risks • Risk-bearing capital is defined as the difference of the market consistent value of assets less the market consistent value of liabilities, plus the market value margin • Target capital is defined as the sum of the Expected Shortfall of change of risk-bearing capital within one year at the 99% confidence level plus the market value margin • The market value margin is approximated by the cost of the present value of future required regulatory capital for the run-off of the portfolio of assets and liabilities • Under the SST, an insurer’s capital adequacy is defined if its target capital is less than its risk bearing capital • The scope of the SST is legal entity and group / conglomerate level domiciled in Switzerland • Scenarios defined by the regulator as well as company specific scenarios have to be evaluated and, if relevant, aggregated within the target capital calculation • All relevant probabilistic states have to be modeled probabilistically • Partial and full internal models can and should be used. If the SST standard model is not applicable, then a partial or full internal model has to be used • The internal model has to be integrated into the core processes within the company • SST Report to supervisor such that a knowledgeable 3rd party can understand the results • Public disclosure of methodology of internal model such that a knowledgeable 3rd party can get a reasonably good impression on methodology and design decisions • Senior Management is responsible for the adherence to principles Defines How-to Defines Output Transparency

14. Contents • Global Regulatory Tendencies • The Impact of the Old Supervisory Framework • SST Principles • Risk • Principle 1: The Economic View • Principle 3: The Economic Balance Sheet • Principle 2 & 4: Risk Quantification • Valuation • Risk Management

15. The Economic View The measurement of risks: Accounting risk or economic risk? • Reported earnings follow the rules and principles of accounting. The results do not always create measures consistent with underlying economics. However, corporate management’s performance is generally measured by accounting income, not underlying economics. Therefore, risk management strategies are directed at accounting, rather than economic performance. • Enron in-house risk-management handbook • For a risk-based solvency system, risks need to be measured objectively and consistently

16. The Economic Balance Sheet The market consistent (economic) balance sheet Assets Liabilities Risk bearing capital Free capital Available Capital Market value of assets SCR: Required capital for 1-year risk Wherever possible, market-consistent valuation is based on observable market prices (marking to market) If such values are not available, a market-consistent value is determined by examining comparable market values, taking into account liquidity and other product-specific features, or marking to model Market-consistent means that up to date values are used for all parameters Market Value Margin Best estimate of liabilities Market consistent value of liabilities Best-estimate = Expected value of liabilities, taking into account all information from financial and insurance market All relevant options and guarantees are valued. No explicit or implicit margins Discounting with risk-free interest rate

17. Risk Quantification Risk quantification via standard models or internal models Available capital changes due to random events Year 0 Year 1 Probability density of the change of available capital Revaluation of liabilities due to new information Available capital Market Value Margin New business during one year Probability < 1% Claims Change in market value of assets Average value of available capital in the 1% ‚bad‘ cases = TailVaR = -SCR Catastrophes Market consistent value of liabilities Market value of assets Best estimate of liabilities Economic balance sheet at t=0 (deterministic) Economic balance sheet at t=1 (stochastic)

18. Risk Quantification • Most capital models consist of two basis components: • A valuation V(.) is a mapping from the space of financial instruments (assets and liabilities) in R: • V: A * L  R, where A * L is the space of all assets and liabilities • A risk measure rm(.) of a random variable (e.g. VaR, TVaR,…) AC(t) = V(A(t))-V(L(t)), t=0,1 SCR = - rm( AC(1) – AC(0) ) Available capital at time t: random variable Available capital at time 0: known Valuation: Market consistent Risk Measure: Expected Shortfall For the SST:

19. Contents • Global Regulatory Tendencies • The Impact of the Old Supervisory Framework • SST Principles • Risk • Valuation • Requirements on the Valuation • Principle 1 & 5: Market Consistent Valuation • Risk Management • Internal Models

20. Requirements on the Valuation 实事求是, We must seek truth from facts 邓小平, Deng Xiaoping • Consistency across assets and liabilities and across products: Without consistency, arbitrage and pure ‘valuation’ volatility will result (e.g. a IFRS phase 2 situation) • Uniqueness: There should not be choice in the sense that one can switch a valuation scheme arbitrarily (e.g. amortized cost or market value for bonds, discounting or not discounting liabilities) • Codifiability: The valuation scheme must be such that it can be codified (e.g. via principles or rules or a mix thereof) • Approximating observable prices: A valuation scheme should not result in prices which are far off observation, at least for a reasonably efficient market • Accepted and used by market participants: Without acceptance by the market (not only by actuaries!), the valuation can not become embedded within the companies For FOPI, market consistent valuation of assets and liabilities optimally satisfies the requirements on an appropriate valuation

21. Market Consistent Valuation Market Consistent Value of Liabilities: Best Estimate (future cash flows discounted with risk free interest rate) + MVM (cost of capital approach) • market value (if it exists); or • value of a replicating portfolio of traded financial instruments + cost of capital margin for remaining basis risk as a proxy for the MVM, For life insurance liabilities: the complexity lies with the projection of future cash flows, which have to take into account all relevant optionalities and guarantees - Options of the company: profit participation features,… - Options of policy holders: lapse options, annuity vs lump sum options,… Review: Supervisors need comfort that the management rules correspond to actual strategy; requirement on technical sophistication of companies increases massively • Optionalities can not easily be replicated using traded instruments, policy holder behavior is often not strictly financially rational • Options of the company imply the formulation and coding of management rules; in case of profit participation features, the modeled future economic (and statutory) performance of the company effects the valuation of liabilities • The long duration of many contracts implies modeling of long-term economic parameters

22. Contents • Global Regulatory Tendencies • The Impact of the Old Supervisory Framework • SST Principles • Risk • Valuation • Risk Management • Principle 11: Risk Management • Principle 11 & 14: Elements of Supervision • Principle 11 & 14: Expectations on the Board • Internal Models

23. Risk Management Warren Buffett‘s three key principles for running a successful insurance business: • They accept only those risks that they are able to properly evaluate (staying within their circle of competence) and that, after they have evaluated all relevant factors including remote loss scenarios, carry the expectancy of profit. These insurers ignore market-share considerations and are sanguine about losing business to competitors that are offering foolish prices or policy conditions. • They limit the business they accept in a manner that guarantees they will suffer no aggregation of losses from a single event or from related events that will threaten their solvency. They ceaselessly search for possible correlation among seemingly-unrelated risks. • They avoid business involving moral risk: No matter what the rate, trying to write good contracts with bad people doesn't work. While most policyholders and clients are honorable and ethical, doing business with the few exceptions is usually expensive, sometimes extraordinarily so. February 28, 2002, Warren E. Buffett • An insurance regulator should set incentives such, that good risk management practices are rewarded: • setting transparent requirements • putting responsibility to the board and senior management • Enforce requirements consistently

24. Senior Management Responsible Actuary Risk Management Internal Audit Elements of Supervision Indirect supervision to ascertain that professional standards are defined and in-line with regulatory expectations Principles based supervision will depend on a web of relationships between the company, professional bodies and the supervisor For a liberal, principles based approach to function, all have to see to it that the system of checks and balances works Supervisor Direct supervision and check that oversight responsibilities are implemented Actuarial Profession Accounting Profession Professional guidance and enforcement of code of conduct Implications for supervision: closer contact and dialogue with the board, professional bodies and all relevant functions within the company Board of Directors

25. Expectations on the Board • The Board of Directors is responsible for: • the governance, guidance and oversight responsibilities that are critical to an effective internal control structure • defining necessary board committees (e.g. audit committee, nomination and compensation committee,…) • The Board as a whole needs to have sufficient technical as well as strategical insurance know-how to be able to supervise and guide the company as well as the necessary stature and mindset • A Board must be prepared to question and scrutinise management’s activities, present alternative views and have the courage to act in the face of obvious wrongdoing • The Board and management need to know how adverse a risk must be for it to impair the insurer’s financial position. This should include all risks arising from the insurer’s assets and liabilities • The members of the Board need to satisfy fit and proper requirements and have to minimize conflict of interests • The Board needs to define the risk appetite and see to it that it is in line with the actual risk capacity of the company

26. Reality vs Potentiality The Need for Probabilistic Thinking For a risk culture to develop, senior management, the board and supervisors must be able to understand the probabilistic nature of the world Example: A CRO hedges the risk of interest rates falling since the company is short in duration. Interest rates then increase and the hedge expires worthless. Senior management then criticizes the CRO for „destroying“ profit Example: A CRO models the exposure to hurricane risk according to industry best-practice but the actual loss is a multiple of the predicted loss. The supervisor then assume wrong-doing and an intentional optimistic assumption to minimize required regulatory capital In insurance, reality can – and often will be – different from prediction. While only one of the possible outcomes will be realized, there nevertheless are many a-priori potentialities, which the company and risk management have to consider

27. Contents • Global Regulatory Tendencies • The Impact of the Old Supervisory Framework • SST Principles • Risk • Valuation • Risk Management • Internal Models • Principle 10: Internal Models • Principle 10: Internal Model Review • Principle 9, 11 & 14: Modelling Deficiencies

28. Internal Models • Risk Quantification: • Using standard models for life, P&C and health companies, if the standard models capture the risk the companies are exposed to appropriately • Using internal models for reinsurers, insurance groups and conglomerates and all companies for which the standard model is not appropriate (e.g. if they write substantial business outside of Switzerland) • How to ensure that the results are comparable between different companies? • How to ensure, that a company is not punished if it models risks more conscientiously than its peers? • How to be able to distinguish between acceptable and not acceptable models? • How to be certain that a model is deeply embedded within a company? The use of an internal model is the default option, the standard models can only be used if they adequately quantify the company‘s risks When allowing internal models for regulatory capital calculation, the problems a supervisor faces are

29. Internal Model Review Even worse than having a bad model is having any kind of model – good or bad – and not understanding it If internal models are used for regulatory purposes, it will be unacceptable if the model is not understood within the company Senior management is responsible for internal models and the review process. The review of internal modes will be based on 4 pillars • Internal Review; • External Review; • Review by the Supervisor; • Public Transparency. • There needs to be • deep and detailed knowledge by the persons tasked with the upkeep and improvement of the model • knowledge on the underlying assumptions, methodology and limitations by the CRO, the responsble actuary, etc. • sufficient knowledge to be able to interpret the results and an awareness of the limitations by senior management and the board The regulator is responsible for ascertaining that the review process is appropriate Companies using internal models have to disclose publicly the methodology, valuation framework, embedding in the risk management processes etc.

30. Internal Model Review If reality turns out differently from the prediction: How can a supervisor distinguish between • an appropriate model which explains the outcome, but the event was rare and extreme by coincidence; • a model which has shown itself to be inappropriate but not due to an error of the modeller (e.g. reality might have changed due to global warming,…); • a model which has shown itself to be inappropriate and the modeller was responsible for the error (e.g. using Black Scholes for unhedgeable options,…); and • a model intentionally chosen in order to reduce regulatory capital requirement? • How can the supervisor give the modeller confidence that • in case a) and b), she will not suspect stupidity or fraud; • in case c) she will not assume a) and b) but also not foul play; and • in case d) she will not be fooled in believing a), b) or c)?

31. Internal Model Review The supervisor has to send the right signals such that there is trust that she will accept the random nature of reality but that she will sanction incompetence and fraud Is it better to err on the optimistic side (e.g. make some errors where incompetence or fraud models are not detected) or to err on the pessimistic side (e.g. identify some honest modelers with incompetent or fraudulent ones)? • To be too pessimistic is dangerous, since it destroys the trust of honest modelers • To be too optimistic emboldens fraudulent companies and will make incompetence acceptable • When incompetence or fraud have been clearly identified, the sanctions have to be swift

32. Modelling Deficiencies Observations based on three field tests of the SST: • A rule based mindset of some companies • Some companies do not deviate from standard model methodologies and parameters • Risk management is not always sufficiently embedded with the companies • Some models are not adequately embedded within risk management • Senior management sometimes is pushing for desired results • State dependent parameters are often calibrated to ‘normal’ experiences (e.g. correlations) • Data quality

33. Principles Based Supervision I believe we are on an irreversible trend toward more freedom and democracy - but that could change Dan Quayle The success of principles based supervision will depend crucially on: • Trust and an open and informed dialog between the industry and the supervisor • Development of a responsibility culture  the willingness to do the right thing rather than purely complying with a minimal set of rules • Adequate self-governance of the industry and relevant professional associations (actuaries, accountants,…) The ultimate responsibility for ascertaining adherence to principles lies with the supervisor but a principles based supervisory framework will depend on devolving responsibility for implementing the principles away from the supervisor to the board of directors, senior management and professional organizations