June 7, The Hartford Financial Services Group, Inc. Smith Barney Annuity & Life Risk Management Seminar

Transcription

1 The Hartford Financial Services Group, Inc. Smith Barney Annuity & Life Risk Management Seminar Craig R. Raymond Senior VP & Chief Risk Officer June 7, 2005

2 Safe Harbor Statement Certain statements made in this presentation should be considered forwardlooking statements as defined in the Private Securities Litigation Reform Act of These include statements about The Hartford s future results of operations. We caution investors that these forward-looking statements are not guarantees of future performance, and actual results may differ materially. Investors should consider the important risks and uncertainties that may cause actual results to differ, including those discussed in the Quarterly Report on Form 10-Q filed on May 2, 2005 and other filings we make with the Securities and Exchange Commission. We assume no obligation to update this presentation, which speaks as of today s date. 1

3 Industry Variable Annuity Sales The Hartford #1 Since VA Sales ($ Millions) Hartford Life $15,244 TIAA-CREF Equitable Metlife AIG/SunAmerica ING Lincoln National Pacific Life John Hancock Prudential Industry Variable Annuity Sales 2004 Source: VARDS $9,552 $9,355 $8,848 $7,674 $7,019 $6,401 $5,948 $5,767 $13,093 Rank Company 2004 Market Share 1 Hartford Life 11.87% 2 TIAA-CREF 10.20% 3 Equitable 7.44% 4 Metlife 7.29% 5 AIG/SunAmerica 6.89% 6 ING 5.98% 7 Lincoln National 5.47% 8 Pacific Life 4.99% 9 John Hancock 4.63% 10 Prudential 4.49% 2

4 How Do We Manage Risk? Product Design Reinsurance Hedging 3

5 The Hartford s GMWB Products Principal First The Hartford s Principal First Preferred Withdrawal Rate 7% 5% Reset Yes, every 5 years No Revocable No Yes, after 5 years Required Asset Allocation No No Transfer Restrictions No Yes, but not currently enforced Cost 50 bps 20 bps Introduction Date 8/ /2004 Advantages of Principal First Preferred Provides customers with basic principal protection at a lower cost Expands Hartford s withdrawal benefit product offerings Product design reduces capital requirements and risk 4

6 Principal First Assets 3/31/05 Individual Annuity Assets $109 Billion Total Principal First & PF Preferred $31.1 B 28.5% Hedged Principal First & PF Preferred $19.7 B 18.1% Reinsured Principal First $11.4 B 10.4% Fixed Annuities $10.9 B 10.0% Other VA (Non-Principal First) $67.0 B 61.5% Of the new variable annuity sales in the 1 st Q 2005 (based on initial deposits only), 63% elected the Principal First rider while 9% elected PF Preferred 5

7 How We Look at Principal First Risk & Hedging Risk is assessed over various time horizons Short-term Long-term Risk is considered across several financial frameworks GAAP Statutory Economic/cash flow Why do we hedge? Substantial short-term GAAP earnings volatility Statutory surplus usage and volatility Claim costs in extreme scenarios 6

8 How We Hedge The Hartford s Hedging Approach Approach: Use of derivative securities to protect against adverse market movements Goal: Change in value of hedge assets Change in value of liabilities The GAAP liability under FAS 133 uses option pricing techniques and capital markets assumptions Our liability hedge target starts with GAAP, but our strategy also contemplates the impacts of hedging on statutory results Desired Benefits Mitigates GAAP net income volatility Significantly reduces statutory capital usage -- especially under proposed NAIC C-3 Phase II risk-based capital requirements 7

9 Greek Hedging Put, Call & Exotic Options Index Futures & Swaps Interest Rate Futures & Swaps Delta Sensitivity to Market Movements Vega Sensitivity to Implied Volatility Changes Rho Sensitivity to Interest Rate Changes In addition to the three base Greeks, we also monitor/manage the Cross-Greeks (i.e., how do Delta, Vega and Rho behave as the capital markets move) 8

10 Long-Term Risk Range of Principal First Benefit Costs Basis Points on Account Value over Scenario Horizon Claims in basis points The long-term cost of Principal First is less than the 50 bps fee under a vast majority of scenarios % 4% 7% 10% 13% 16% 20% 23% 26% 29% 32% 36% 39% 42% 45% 48% 52% 55% 58% 61% 64% 68% 71% 74% 77% 80% 84% 87% 90% 93% 96% 100% % of 250 Scenarios 9

11 Impact of Hedging on GAAP Income Volatility 1995 Stress Test Scenario Quarterly GAAP Net Income With Hedging 150 #1 #2 #3 Hedge significantly reduces GAAP net income volatility GAAP Net Income ($ in millions) ) Mar-95 (50) (100) Mar-96 Mar-97 Mar-98 Mar-99 Mar-00 Mar-01 Mar-02 Mar-03 Mar-04 Mar-05 Mar-06 Mar-07 Mar-08 Mar-09 Mar-10 Mar-11 No hedging Hedging (150) (200) Based on $10B of Principal First written in the 1 st quarter of

12 Impact of Hedging on Statutory Capital Usage Principal First Annual Statutory Capital Usage $10 billion of Principal First issued on one day 300 Single Year Capital Usage ($ in millions) (50) (100) Hedging mitigates Statutory impacts Potential benefits of aggregation (150) 50.0% 25.0% 10.0% 5.0% 2.5% Probability of Capital Usage Event Stand-alone PF With Hedging 11

13 Principal First vs. Principal First Preferred without Hedging Principal First vs. Principal First Preferred Annual Statutory Capital Usage (both with no hedging) $10 billion of Principal First issued on one day Single Year Capital Usage ($ in millions) (50) (100) (150) Principal First Preferred has less risk due to its key design elements: 5% withdrawals (vs. 7%) No reset allowed (vs. every 5 years) 50.0% 25.0% 10.0% 5.0% 2.5% Probability of Capital Usage Event Stand-alone PF Stand-alone PFP 12

14 Actual GAAP Volatility 2004 Volatility of Realized Gain (Loss) Hedging stabilized monthly results 20 $ in millions Jan-04 Feb-04 Mar-04 Apr-04 May-04 Jun-04 Jul-04 Aug-04 Sep-04 Oct-04 Nov-04 Dec-04 Hedge fit was tight Diligent monitoring of risk & rebalancing Markets were not extremely volatile No major dislocations No qtr. had after tax/dac>$2 M Hedge cost after tax/dac was 3-4 BPS After tax/dac Net Gain (Loss) post-hedging Liability Gain (Loss) 13

15 Hedging Principle First Under a wide range of market conditions, the replication hedge will perform very well Cannot perfectly hedge the Principal First potential claim stream using the available universe of derivatives We understand the potential for hedging Gain/Loss and provide for this risk in the product pricing Proactive strategies to address challenges Strong consideration of the Cross-Greeks when managing the hedge portfolio Continually pursue more structured options to better match the liability Constant communication with Street to follow supply/demand dynamics 14

16 Daily Hedging Process Overnight: Liability models run over 100,000 scenarios to produce Greeks, Market Value (MV) and attribution of MV change Uses closing capital market data and prior day inforce file Start of Business Day: Liability files are sent from Life Co. to Hartford Investment Management Asset statistics are compared with Liability statistics Trades needed to rebalance the Greeks are identified and executed Trade to get Delta, Vega and Rho neutral P&L for prior day is produced Throughout the Business Day: As market conditions change, Cross-Greeks may create the need to rebalance The Asset and Liability Greeks are monitored on a real time basis Trades needed to rebalance the Greeks are identified and executed 15

17 Principal First Contractholder Experience Characteristics of current Principal First contractholders are similar to the rest of the variable annuity book (age, fund selection, etc) Contractholder experience studies Available data 1.4 million inforce variable annuity contracts 300,000 Principal First contacts We perform experience studies using 15+ years of experience Over 20 million contract years of experience Our experience data includes varying market environments the bull market of the late 1990 s through the market drops in Tracking of data Daily inforce and behavior snapshots Monthly experience trend analysis Sorting and tracking of behavior across many characteristics age, contract type, contract year, GMDB, fund mix, qualified status, use of auto-income and others 16

18 Risk Management and Hedging Recap We have a very good understanding of our Principal First business We diligently monitor our policyholder experience and use this information to improve our estimates We assess and measure risk across a variety of metrics and time horizons, which is factored into our hedging program Our hedging program has performed extremely well thus far. Given the growing size of our book and the potential for more volatile market conditions, we would expect more GAAP Gain/Loss than seen previously 17

19 C3 Phase II NAIC Initiative Reserves and Risk-Based Capital (RBC) Applies to Variable Annuity contracts and associated guarantees RBC is expected to be implemented for year-end 2005 Reserves expected to be implemented for year-end 2006 Model Based Methodology Model business over a broad range of stochastically generated scenarios Uses Prudent Best Estimate assumptions Include all revenue, benefits, and expenses Include currently held assets, including hedge assets Include impact of reinsurance and future hedge strategy 18

20 C3 Phase II CTE Measurement Conditional Tail Expectation measures Tail Risk CTE 90 for RBC means average of the results for the worst 10% of scenarios CTE 65 for Reserves means average of the results for the worst 35% of scenarios Advantages Aligns measurement of reserves and RBC to underlying risks Rewards risk management such as reinsurance, hedging, and product design The Hartford s View Actively involved in the development and analysis of the NAIC proposal Already pricing and managing capital based on this approach 19

21 Number of CPU's Risk Management Technology Risk management technology is all about enabling massive computation Hartford Life deployed Grid Computing in 2004 Pioneers in the Life Insurance Industry Hedging Program Hedging Starts Program Starts Analytical Capability Growth Grid Computing Grid Introduced Computing Introduced 0 3/2003 3/2004 3/2005 Grid Computing works like a virtual supercomputer Ties together servers and desktops into one big processing engine Allows us to make use of our latent computing power Date Enables tremendous scalability and resiliency 20 Hartford s Hartford s Analytical Analytical Computing Computing Facts Facts From 35 CPU s in our original Hedging From 35 CPU s in our original Hedging release, the platform has grown to > 350 release, the platform has grown to > 350 We have the capacity to process hundreds of We have the capacity to process hundreds of millions of calculations per second millions of calculations per second Newly optimized version of the hedging Newly optimized version of the hedging analytics runs times faster than before analytics runs times faster than before Hedging process alone consumes enough Hedging process alone consumes enough data monthly to fill the Library of Congress data monthly to fill the Library of Congress Analytics can run on any desktop, any Analytics can run on any desktop, any server, any operating system at the same server, any operating system at the same time ensuring massive expandability time ensuring massive expandability Grid computing environment runs Grid computing environment runs seamlessly across multiple data centers, seamlessly across multiple data centers, ensuring business continuity ensuring business continuity Exploring expansion to public grid centers Exploring expansion to public grid centers with major technology companies with major technology companies

22 The Hartford Financial Services Group, Inc. Smith Barney Annuity & Life Risk Management Seminar Craig R. Raymond Senior VP & Chief Risk Officer June 7, 2005