4A: The Money Pit - Reflecting the Risks We Are Taking In Pricing Products

Similar documents
Market Consistent Embedded Value and Implications for Pricing

Session 55 PD, Pricing in a MCEV Environment. Moderator: Kendrick D. Lombardo, FSA, MAAA

Rethinking Fixed Deferred Annuities: Applying a Risk-Based Economic Value Approach

Disclosure of European Embedded Value as of September 30, 2010

Market-Consistent Embedded Values

Allianz. European Embedded Value Report

Economic Capital: Recent Market Trends and Best Practices for Implementation

Disclosure of Market Consistent Embedded Value as at March 31, 2018

Embedded Value for Insurance Company

Disclosure of European Embedded Value as of 30 September 2015

Life 2008 Spring Meeting June 16-18, Session 94, Impact of IFRS Insurance Accounting. Moderator Simon R. Curtis, FSA, FCIA, MAAA

AvivaSA Emeklilik ve Hayat A.Ş. Market Consistent Embedded Value Report. Half-year 2017

AvivaSA Emeklilik ve Hayat A.Ş. Market Consistent Embedded Value Report. Half-year 2018

Groupama European Embedded Value Report

MCEV : Practical approaches in

Version VI. White paper. April White paper Danica version VI. Consolidation policy and business activities. at Danica Pension.

2009 Market Consistent Embedded Value. Supplementary information 3 March 2010

Market Consistent Embedded Value Basis for Conclusions

AvivaSA Emeklilik ve Hayat A.Ş. Market Consistent Embedded Value Report. Full-year 2017

Disclosure of European Embedded Value as of March 31, 2016

Disclosure of European Embedded Value as of March 31, 2016, using an Ultimate Forward Rate

Valuation methods for life insurers: Conversion or chaos?

Disclosure of European Embedded Value as of March 31, 2018

Disclosure of European Embedded Value as of March 31, 2017

Disclosure of Market Consistent Embedded Value as of March 31, 2016

Disclosure of European Embedded Value as of March 31, 2018

Disclosure of Market Consistent Embedded Value as of March 31, 2018

Disclosure of European Embedded Value as of September 30, 2015

Embedded Value in Non Life Insurance a suggested approach

Deep dive into IEV and views from the market

Market Consistent Embedded Value 2016

Disclosure of European Embedded Value as of September 30, 2016

KBC Embedded Value Report 2007 Contents

2006 Tillinghast Pricing Methodology Survey Results

UNIQA Insurance Group AG. Group Embedded Value 2014

SWEDBANK FÖRSÄKRING AB European Embedded Value

2004 European Embedded Value for Life & Savings activities. December 12, 2005

2010 EEV/MCEV. Focus on value and cash flow. Table 01. Summary of approach to allowing for risk companies reporting EEV/MCEV 1

Embedded Value 2009 Report

Market Consistent Embedded Value (MCEV)

2014 Embedded Value Results - Europe Generating Value

Stochastic Analysis Of Long Term Multiple-Decrement Contracts

The directors of Talanx acknowledge their responsibility for the preparation of this disclosure document.

1. INTRODUCTION COVERED BUSINESS DEFINITIONS... 4

KBC 2006 Embedded Value Results Content

Session 83 PD, Modeling Managing and Pricing Living Benefits Risk. Moderator: Sean Michael Hayward, FSA, MAAA

Industrial Alliance Post-Demutualization and Financial Disclosure Tools. Denis Ricard Senior Vice-President Chief Actuary A PARTNER YOU CAN TRUST.

Stochastic Modeling Concerns and RBC C3 Phase 2 Issues

SWEDBANK FÖRSÄKRING AB European Embedded Value

Munich Re Market Consistent Embedded Value Report 2012

TWOTHOUCEENDAND FIFTEEN

UNIQA Versicherungen AG. Group Embedded Value 2008

UNIQA Insurance Group AG. Group Embedded Value 2017

Supplementary Information on the Group Embedded Value Results 2016 CAN YOU COUNT US ON 17PG001/HE16 (17.03 J )

UNIQA Group Group Embedded Value May 2012 Kurt Svoboda, CRO

Disclosure of European Embedded Value as of March 31, 2017

Enterprise Risk Management and Stochastic Embedded Value Modeling

Session 61, Overview of Embedded Value. Moderator: Zeeshan Ramzan Ali Rehmani, FSA, MAAA. Presenter: David Lawrence White, Jr.

The directors of Talanx acknowledge their responsibility for the preparation of this disclosure document.

European Embedded Value 2010

Disclosure of European Embedded Value as of September 30, 2014

TABLE OF CONTENTS. Lombardi, Chapter 1, Overview of Valuation Requirements. A- 22 to A- 26

Article from: Risk Management. March 2008 Issue 12

UNIQA Versicherungen AG. Group Embedded Value 2010

SOA Risk Management Task Force

Practical issues in ALM and Stochastic modelling for actuaries. Shaun Gibbs FIA Eric McNamara FFA FIAA

2015 Embedded Value Results Overview Focus on Switzerland

Insights Life supplementary reporting. Entering a new era? EV reporting remained prevalent across Asia-Pacific, although reduced in Europe

EUROPEAN EMBEDDED VALUE 2005

IASB s Insurance Contracts Exposure Draft: Risk in the Next Decade

LIFE INSURANCE & WEALTH MANAGEMENT PRACTICE COMMITTEE

Insights 2011 EEV/MCEV

EUROPEAN EMBEDDED VALUE 2006

European Embedded Value as of September 30, EEV as of September 30, 2016: 32,008 million yen

DYNAMIC ASSET LIABILITY MANAGEMENT

US GAAP Update What s New

Market Consistent Embedded Value (MCEV)

Variable Annuity Market Trends. Presented by : Ken Mungan, FSA, MAAA Financial Risk Management, Practice Leader

UNIQA Group Austria Group Embedded Value Hannes Bogner CFO May 25, 2011

Practical application of Liquidity Premium to the valuation of insurance liabilities and determination of capital requirements

Inforce Management 2014 ACHS Fall Meeting

The Actuarial Society of Hong Kong DISCOUNT RATES Insurance IFRS Seminar. Ben Lovelock. Session 8

Appendix 1: Strategy, Targets and Remittances per segment Appendix to ING Group and NN Group Press Release of 5 June 2014

The Financial Reporter

Disclosure of European Embedded Value (summary) as of September 30, 2011

MATH FOR CREDIT. Purdue University, Feb 6 th, SHIKHAR RANJAN Credit Products Group, Morgan Stanley

Embedded Value. & AFR report. Cash and Value Report- AXA / FY2016 1

Session 76 PD, Modeling Indexed Products. Moderator: Leonid Shteyman, FSA. Presenters: Trevor D. Huseman, FSA, MAAA Leonid Shteyman, FSA

Economic Scenario Generators

Supplementary Information on the Life Health Embedded Value Results 2017 WE EMBRACE DIVERSITY. Protecting what matters. (18.

GN47: Stochastic Modelling of Economic Risks in Life Insurance

Explaining Your Financial Results Attribution Analysis and Forecasting Using Replicated Stratified Sampling

Wiener Städtische Versicherung AG Vienna Insurance Group

The Evolution of ALM in Asia. Some Food For Thought. Mark Saunders, FIA, FASI, FSA

Article from Financial Reporter. December 2017 Issue 110

Embedded Value 2013 Report

Additional Unaudited Financial Information (New Business and Value of in-force) 35

Session 61 L, Economic Scenario Generators: Risk-Neutral and Real-World Considerations from an Investment Perspective

CEIOPS-DOC-35/09. (former CP 41) October 2009

An Introduction to Solvency II

Transcription:

9 th Annual Product Development Actuary Symposium June 2009 4A: The Money Pit - Reflecting the Risks We Are Taking In Pricing Products Dominique Lebel

Market Consistent Pricing Risk Management at the Point of Sale Product Development Symposium Dominique Lebel, FSA, FCIA, MAAA June 2009 2009 Towers Perrin What is your primary pricing metric? 1.Real world IRR 2.Real world profit margin 73% 3.Risk neutral profit margin (MCEV) 4.Real world VNB (EV) 15% 3% 9% 2 1 2 3 4 2009 Towers Perrin

Shortcomings of traditional pricing methods Most common pricing measure is statutory internal rate of return ( IRR ) Pricing objective is based on achieving a rate of return in excess of the company s hurdle rate Hurdle rate often based on company s overall cost of capital Hurdle rate typically does not vary by product But different products have different levels of risk A product with a higher pricing IRR does not necessarily create more shareholder value than a product with a lower pricing IRR Depends on risks inherent in each product Products are priced under the implicit assumption that arbitrage opportunities exist Asset risk premiums are capitalized during pricing These risk premiums are passed to policyholders before insurers are released from risk If insurers believe that these arbitrage opportunities exist, why not just borrow at the insurer s credit rating and invest in riskier assets rather than manufacture and distribute insurance products? Consideration should be given to pricing products such that all risks undertaken are measured in an objective and consistent way 2009 Towers Perrin 3 Traditional vs. market consistent pricing approaches Traditional Typically the discount rate used is the same for all products and is based on the cost of capital of the company Credit spreads and equity risk premiums are capitalized at issue The costs of options and guarantees are generally not valued in a manner that is consistent with how they are valued in the market Market Consistent The discount rate is set to reflect the risks inherent in each product Credit spreads and equity risk premiums are earned as insurer is released from risk The costs of options and guarantees are valued in a manner that is consistent with how they are valued in the market 2009 Towers Perrin 4

Typically, for each product, a value of new business ( VNB ) is determined which reflects the value to shareholders created through the activity of writing new business VNB = Present value of future profits after tax Time value of financial options and guarantees Frictional costs of required capital Cost of non hedgeable risks Common metric is the profit margin: VNB / present value of premiums Implied discount rate is the discount rate such that a traditional VNB will equal the market consistent VNB This is sometimes used to compare the relative level of risk between products Market consistent pricing provides a robust, transparent and objective economic perspective on new business profitability that is consistent across products. If the VNB is greater than zero, the return is greater than the market price of the risks undertaken. A VNB less than zero reduces shareholder value. 2009 Towers Perrin 5 PRESENT VALUE OF FUTURE PROFITS With market consistent valuations, cash flows are valued using a discount rate which reflects the risk inherent in each cash flow Assume that the investor has capital of 20, borrows 80 (at 5% pa) and then invests 100 in equities (with an expected return of 8% pa) Day 1 One year on Assets 100 108 Liabilities (80) (84) Capital 20 24 Note: The difference between the traditional and market consistent results may be partly compensated for by a higher cost of capital under the traditional methodology (see later) The actions of borrowing and investing do not create value on Day 1 Using traditional pricing techniques the 24 might be discounted at, say, 10%. This would give a value of 21.8 on Day 1 Under MCV the asset cash flows are discounted at 8% and the liability cash flows at 5%. This would give a value of 20 on Day 1. The effective discount rate for profits is 20% 2009 Towers Perrin 6

PRESENT VALUE OF FUTURE PROFITS We can also risk adjust the cash flows Assume that the investor has capital of 20, borrows 80 (at 5% pa) and then invests 100 in equities (with an expected return of 8% pa) Day 1 One year on Assets 100 104.0 Liabilities (80) (83.2) Capital 20 20.8 The actions of borrowing and investing do not create value on Day 1 An alternative approach is to project and discount everything at the risk free rate (assumed to be 4%) As the example shows, we get the same Day 1 result as the methods shown on the prior page This is called the certainty equivalent approach and is commonly used in practice 2009 Towers Perrin 7 TIME VALUE OF FINANCIAL OPTIONS AND GUARANTEES Time Value vs. Intrinsic Value Total value of options and guarantees = Intrinsic value + Time value Intrinsic value already captured in the certainty equivalent PVFP Time value = PVFP (certainty equivalent) PVFP (stochastic) Intrinsic Value Time Value PVFP (deterministic - no guarantee) PVFP (deterministic - guarantee) PVFP (stochastic - guarantee) 2009 Towers Perrin 8

TIME VALUE OF FINANCIAL OPTIONS AND GUARANTEES Stochastic Modeling Generally, risk neutral valuations Economic scenarios are derived from swap rates such that, on average, all assets follow the forward rates Model should be calibrated to market values Equity option implied volatilities Swaption implied volatilities Initial swap rate curve Market consistency is tested by checking that initial asset market value = mean present value of asset cash flows discounted at risk free rate (1 = 1 test) Purpose is to ensure that the asset model is simulating risk free returns for all assets Assumptions used for stochastic models should be consistent with deterministic models Allowance should be made for dynamic management actions and policyholder behavior Relative to real world projections, the time value of financial options and guarantees will generally increase 2009 Towers Perrin 9 FRICTIONAL COSTS OF REQUIRED CAPITAL Frictional costs should reflect the tax and investment expenses on the assets backing required capital Generally based on the difference between: Earned rate (after tax and investment expenses) on assets backing required capital Discount rate In a market consistent world: Earned rate (before tax and investment expenses) is equal to the discount rate Only costs left are the frictional costs of taxes and investment expenses on investment income on assets backing required capital Relative to real world projections, the cost of capital will generally decrease Additional frictional costs include agency costs and the cost of financial distress, but the CFO Forum decided to exclude these General corporate risks that individual investors should assess 2009 Towers Perrin 10

COST OF RESIDUAL NON HEDGEABLE RISKS Cost of residual non hedgeable risks (NHR) Finance theory states that investors do not need to be compensated for risks that can be diversified, provided the best estimate assumptions used in valuation reflect the mean An allowance needs to be made for the cost of non hedgeable risks not already allowed for in the time value of options and guarantees or the present value of future profits Should include the impact of non hedgeable non-financial and non hedgeable financial risks NHR allowance is intended to allow for the following items: Incomplete best estimates in the base model (e.g., operational risk) Asymmetric risk distributions or asymmetric impact on results Non-linearities/correlations with market risk (e.g., dynamic policyholder behavior) Non hedgeable financial risk (e.g., swap market not deep and liquid at long durations) The allowance is distinct from the frictional costs of required capital 2009 Towers Perrin 11 COST OF RESIDUAL NON HEDGEABLE RISKS Reconciliation between the CRO and CFO Forum approaches CRO Forum CFO Forum Regulatory capital in excess of EC Regulatory capital in excess of EC Total EC EC for hedgeable risks EC for hedgeable risks Frictional cost on total capital EC for non hedgeable risks EC for non hedgeable risks Fixed charge for NHR EC for non hedgeable risks Residual Cost of NHR Total Required Capital Capital Annual Charge Capital Annual Charge 2009 Towers Perrin 12

Implications for North American products and companies are numerous Market consistent pricing can have significantly different impacts by company depending on several factors including: Product mix Split of fixed vs. variable/segregated fund business Level of guarantees Amount of asset risk (e.g., credit quality of assets) Treatment of future flexible policyholder credited rates Everything else being equal (e.g., assuming the same product charges), a product (Product A) with more guarantees, more asset risk and without management levers to mitigate adverse experience ought to be considered more risky than a similar product (Product B) with opposite characteristics The pricing metric used should show a less favorable result for Product A relative to Product B This is the case under a market consistent pricing framework 2009 Towers Perrin 13 Published information shows that Market Consistent Pricing can have significantly different impacts on VNB margins by company Comparison of Market Consistent vs. Traditional European Embedded Value Profit Margins Company Year Value of New Business Profit Margin Traditional Market Consistent Difference Allianz US 2005 1.9% 2.5% 0.6% Aviva US 2007 / 2008* 1.9% / 2.0% 0.9% / -0.2% -1.0% / -2.2% AXA US 2004 1.8% 1.3% -0.5% Old Mutual US 2007 2.3% 1.0% -1.3% Zurich/Farmers 2005 4.4% 6.6% 2.2% *June 30 2009 Towers Perrin 14

Comparison of market consistent vs. traditional profit margins (positive figure indicates market consistent greater than traditional) Product Company A Company B Company C Term Insurance 2.3% 1.4% N/A Universal Life/Variable UL 2.7% -9.1% to +1.4% N/A Fixed Annuities -4.1% -3.6% to -0.5% -1.0% Payout Annuities -4.9% -7.6% -6.1% VAs/Segregated Funds 0.0% N/A -1.8% Fixed Indexed Annuities N/A -5.3% to -3.9% N/A Critical Illness 1.0% N/A N/A Group Life & Health/Employee Benefits 1.1% N/A 1.9% 2009 Towers Perrin 15 Typical Implied Discount Rates Relative to Typical Internal Rates of Return Used to Price Common Products Product Term Insurance Universal Life/Variable UL Fixed Annuities Payout Annuities VAs/Segregated Funds Fixed Indexed Annuities Critical Illness Group Life & Health/Employee Benefits Typical Implied Discount Rate Lower than typical IRR used Varies Varies, but generally higher Higher than typical IRR used Varies, but generally higher Higher than typical IRR used Lower than typical IRR used Lower than typical IRR used 2009 Towers Perrin 16

The approach used to set the cost of guarantees on variable annuity business has evolved over time Method Used to Determine Cost of Guarantees on VAs (e.g., Guaranteed Minimum Death Benefits, Guaranteed Minimum Income Benefits) (Percent of Responses) 61% 26% 10% 19% 16% 6% 6% 24% 26% 8% Mean cost based on real world stochastic scenarios Fiftieth or other percentile based on real world stochastic scenarios Set equal to reinsurance cost 2006 2008 Set equal to expected hedge costs based on risk neutral stochastic scenarios Other Note: Companies selected multiple responses if they used different methods for different guarantees. Source: 2006 and 2008 Tillinghast Pricing Methodology Surveys (i.e., methodology used to price products in 2005 and 2007) 2009 Towers Perrin 17 Market consistent pricing methodologies or assumptions are most often used in the US for product features that are hedged in the capital markets Use of Market Consistent ( MC ) Pricing Methodologies or Assumptions (Percent of Responses) Use MC VNB as alternative pricing measure 10% Use MC VNB as primary pricing measure 8% None 44% Use MC assumptions for product features hedged in capital markets 38% Source: 2008 Tillinghast Pricing Methodology Survey (i.e., methodology used to price products in 2007) 2009 Towers Perrin 18

While market consistent pricing was not broadly used in 2007 for a wide range of products, this is gradually changing Market consistent techniques are making their way into a wide range of applications US GAAP (FAS 157 and 159) The European Insurance CFO Forum MCEV Principles Published in June 2008, requires member companies to publish year-end 2011 embedded values and VNBs using market consistent techniques Many companies, domestic and international, are using market consistent methodologies to determine economic capital (a la Solvency II) More and more merger and acquisition and securitization transactions are being valued using both traditional and market consistent techniques Some companies are embracing market consistent techniques because they believe these methods provide useful insights into asset-liability management Others use market consistent pricing for incentive compensation to align compensation with the risks undertaken The above developments have motivated many companies to look at the profitability of their products using market consistent techniques Some of these companies have made or are in the process of making changes to their products and/or pricing Other companies have embraced market consistent pricing for its own sake IFRS Phase II, which is based on a fair value approach, could become required in 2014 in the US and in 2013 in Canada Consequently, the use of market consistent pricing should continue to increase in North America 2009 Towers Perrin 19 Those that act early can gain a competitive advantage Market consistent pricing could be used to develop strategic options Companies could target products where current product charges are greater than prices required by the market Companies moving first would gain leverage by targeting profitable products Eventually inefficiencies will be corrected as competitors catch up Companies could also use market consistent pricing analyses to better understand the relative risks of their products Depending on a company s risk appetite, measures could then be taken to derisk certain products by increasing product charges or making changes to the product design Product design changes could include decreasing interest rate guarantees making variable annuity/segregated fund guarantees less rich introducing market value adjustments upon surrender changing premiums from a guaranteed basis to an adjustable basis Companies could use market consistent pricing techniques to protect themselves against similar tactics used by competitors 2009 Towers Perrin 20

Markets are displaying highly unusual characteristics, compared to historic norms Nil/negative LIBOR swap spreads (over government) at medium to long durations Corporate bonds, CDS protection and LIBOR swap yields suggest potential arbitrage opportunities exist, at least in credit markets High implied volatilities Different sources of data at times show very different results Does the concept of willing buyer, willing seller still hold? Are market prices reliable? 2009 Towers Perrin 21 Highly unusual market characteristics Time series plot showing comparison of swap spot rate and notional spot rate over time 11.00% 10.00% 9.00% 8.00% Spot rate 7.00% 6.00% 5.00% 4.00% 3.00% 2.00% 1.00% Corporate Bond Yield Corporate Bond Yield less adjusted CDS spread Swap spot rate 0.00% Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Time Source: Tillinghast analysis of Bloomberg and Markit data 2009 Towers Perrin 22

Current market conditions support the use of a liquidity premium in pricing predictable liability cash flows Corporate bond spreads have risen sharply Part of this is an increased perception of default risk However, this component can be measured to some degree by CDS spreads The remainder includes a liquidity premium and any unexplained components Replicating portfolios are a common technique used to calculate market consistent values Liability cash flows are replicated using a theoretical, easy-to-value portfolio of assets In a perfect market, all replicating portfolios for a liability will have the same value. In practice they do not due to market imperfections and other characteristics (e.g., liquidity) Where there are alternative replicating portfolios, the cheapest portfolio could be used to value liabilities In MCEV, the LIBOR swap rate is already used instead of the government yield 2009 Towers Perrin 23 Cheapest replicating portfolio in mark-to-market valuations For example, payout annuities have largely predictable cash flows over long terms Provided an annuity writer is protected against defaults, it will have no need to trade the investments and is unaffected by market value fluctuations Pricing should report sales of payout annuities as profitable if they are priced with an interest rate up to Risk free rates, plus The residual corporate bond spread over the CDS spread However, if a replicating portfolio cannot be constructed it should not be used 2009 Towers Perrin 24

Theory also supports the inclusion of an own credit risk adjustment in mark-to-market valuations MCEV Principles treat guaranteed liabilities as being certain to be paid. In reality, there is a chance of default and this should theoretically reduce the value of the liabilities held Some banks have (controversially) introduced this element to value their debt Until recently, the impact was small In current conditions the economic impact of allowing for own credit risk is very significant The impact is of the same order of magnitude as the current liquidity premium 2009 Towers Perrin 25 End 2008 CFO Forum MCEV & market-consistent EEV reporting: comparison of reference rates COMPANY Allianz, Hannover Re, Munich Re, Zurich Aviva AXA, Fortis CNP Generali Old Mutual Prudential Scottish Widows Standard Life REFERENCE RATES END 2008 EV Swaps (government bonds for Allianz Korea) Swaps +150bps: UK and NL immediate annuities Swaps +300bps for US immediate annuities / +250bps for other US contracts Swaps: other businesses Swaps +50bps for European businesses, +100bps for most other businesses Swaps +70bps Government bonds: Italy, Czech Republic Swaps +50bps: most other European businesses Swaps for other business Swaps +300bps for US onshore business Swaps for other businesses MCEV approach used for UK annuities only: gilts + c.250-290bp (exact adjustment not disclosed) Gilts +154bp for UK annuities; otherwise gilts Reference rate set to 3.42% (gilts); for UK annuities investment return 6.44% 2009 Towers Perrin 26

Key statements for market consistent pricing 1 2 3 4 Market consistent pricing addresses some of the shortcomings of traditional pricing methods by providing a framework for understanding the tradeoffs between shareholder risks and rewards using a robust, transparent and objective economic methodology that is consistent across products The use of market consistent pricing has recently extended beyond variable annuity guarantees to a wide range of life, health and annuity products More and more companies are looking at the profitability of their business using market consistent approaches motivated by FAS 157 and 159, MCEV Principles, economic capital calculations, mergers and acquisitions and securitization transactions, asset-liability management, incentive compensation and IFRS Phase II When markets are dislocated, adjustments can be made for liquidity premiums and own credit risk 5 Companies that are among the first to take action may benefit 2009 Towers Perrin 27 Thank You Dominique Lebel, FSA, FCIA, MAAA Towers Perrin 525 Market Street, Suite 2900 San Francisco, CA 94105-2708 Direct: 415.836.1081 E-mail: dominique.lebel@towersperrin.com 2009 Towers Perrin 28

Market Consistent Pricing Reflecting the Risks in our Products Brian Campbell, FSA Product Development Symposium June 30, 2009 Do you consider embedded value in your pricing process? 1.Yes 2.no 57% 43% 30 1 2 2009 Towers Perrin

Outline History of Embedded Value at Allianz Implementation Challenges and Considerations Challenges for Our Products and Actions Taken Concluding Remarks Copyright Allianz July 14, 2009 31 History of EV Reporting at Allianz Life First calculated in 1999 2000 EV guidelines developed at Group level Movement analysis 2001 Reviewed by external consultants Group published results 2003 Cost of options and guarantees estimated but not published 2004 Move towards EEV principles Values of options and guarantees published 2005 Published and disclosed EEV results to public 2006 Published and disclosed Market Consistent Basis Copyright Allianz July 14, 2009 32

Why We Added MCEV as a Component of AZ Life Pricing Our Parent Reports on MCEV Always the first Question What is the VNB? Especially Now! Complements a Real World Approach Add on to Stochastic Real World Pricing Highlights Financial Payoffs required to Achieve Expected Profits Less Biased Approach to Valuing Guarantees Cannot cheat the market As Unbiased as the Market is Unbiased Copyright Allianz July 14, 2009 33 Challenges to MCEV Pricing Getting Buy-In from Senior Management as to Value of Measure Viewed initially with Skepticism Difficult to link IRR target with MCEV result RW PM helps provide a bridge Lack of Calibration Techniques No formal secondary market Limited disclosure by competitors Balancing Real World and MCEV Targets Pure MCEV Approach may be uncompetitive Cannot satisfy both without Compromise Possible Approaches Negative MCEV Budget Assumption or Methodology Adjustments Monitoring Issues Selling positive MCEV business not locked-in MCEV much more sensitive post-sale Copyright Allianz July 14, 2009 34

Implementation Challenges Requires Stochastic Processing Most Lines already there Requires Risk Neutral Scenario Set Modified Internal Log-normal Generator Hedging Group Generators Scenarios Provided by Parent Outside Vendor If Already using Stochastic Pricing, to be added Fairly simple add on to existing modeling Use Statutory earnings No need to model hedging frictional cost main consideration Capital cost is also frictional tax and investment expense Result = average of PV of DE s Copyright Allianz July 14, 2009 35 Assumption and Methodology Considerations Market Environment to Utilize Current, Near-term or Long-term Swap rates Market volatility Fund parameters Credit spreads Dynamic Policyholder behavior Same as RW or Adjusted Parameters to use Areas of Discussion Liquidity Premiums Mean Reversion Sub-optimal behaviors Copyright Allianz July 14, 2009 36

Challenges with our Product Lines Variable Annuities Variable Annuity Market driven by demand for Guarantees Guarantees are Sensitive to Swaps, Volatility and Market Movements Regulatory and Competitive constraints to constant Re-pricing Limited ability to add management levers to control MCEV Cannot enter and exit market without impact on market share Fixed Annuities MYGA s and SPIA s typically have negative MCEV at issue Index and Annual Rate Annuities have higher MCEV at issue Spread compression at low swap levels leads to negative MCEV s Copyright Allianz July 14, 2009 37 Actions Being Taken within our Product Lines Focus on EV at Issue and Over Life of Product Both MCEV and IRR are Important Criteria Close Monitoring of Guarantees Relative to Current Market Continued Balance of Competitiveness vs. Prudent Pricing Take Advantage of Product Flexibility Closer Tie between Pricing and Reporting Build Expectations and Faster Feedback Loop Copyright Allianz July 14, 2009 38

MCEV Emergence An Example from the CFO Forum Illustrative Immediate Annuity MCEV Profile 4.0% % premium 2.0% 0.0% -2.0% -4.0% 0 MCEV Earnings 2 4 6 8 10 12 14 16-6.0% New Business Margin Time MCEV MCEV earnings Copyright Allianz July 14, 2009 39 Conclusion Introducing MCEV into Pricing has its Challenges but The Benefits make it worth it - - MCEV is a Valuable Complement to Real Word Pricing Especially for Product with Complex Guarantees - Allows for Greater consistency in Valuing Risk across Products - Takes Asset Arbitrage out of the Equation - Does Require Perspective as to how Given Products will React under MCEV Copyright Allianz July 14, 2009 40

Will you consider embedded value in your pricing process? 1.Yes 2.no 82% 18% 41 1 2 Copyright Allianz July 14, 2009 42