The Value Proposition of DFA Basic Applications for Real World Decision Making Bob Mueller ACAS Executive Vice President Middlesex Mutual / Country Companies Group
Perspectives... CFO, Chief Actuary, Corporate Strategist Small to Mid - sized Company
Illustrations... 1) Asset Allocation Decision 2) Strategic Partnering Decision
Anatomy of an Asset Allocation Decision Year: 1988
Anatomy of an Asset Allocation Decision Historic Financial Profile: Low leverage, high equity investment
Anatomy of an Asset Allocation Decision Emerging Issues –Premium growth % per year –Significant new investment in Real Estate –Pressure from A.M.Best
Anatomy of an Asset Allocation Decision Board Finance Committee: –University president with investment background –Regional bank president –Two president/owners of boutique investment management firms –CFO of a national investment management firm Key Decision Makers
Approach Direct, intuitive Focus on concept and principal Quick
“Back of the Napkin” Model Total Return = Investment + Underwriting Return T / S = I/S + U/S = I/A x A/S + U/P x P/S = I/A x (1 + L/PxP/S) + U/P x P/S S= Surplus, A= Assets, L=Liabilities, P=premium
“Back of the Napkin” Model Total Return = Investment + Underwriting Return T / S = I/S + U/S = I/A x A/S + U/P x P/S = I/A x (1 + L/PxP/S) + U/P x P/S “constant” leverage factors
“Back of the Napkin” Model Total Return = Investment + Underwriting Return T / S = I/S + U/S = I/A x A/S + U/P x P/S = I/A x (1 + L/PxP/S) + U/P x P/S random variables
Model Results No stocks 50% stocks 100% stocks
Model Results
Business Conclusions Risks too high, returns too low at any level of equities Need to focus on core business issues rather than asset allocation
In Summary... Advantages –not a “black box” approach –clear linkage of investment and underwriting –gets management acquainted with DFA Disadvantages –one year projection –not good for “fine tuning” asset allocation or underwriting strategy
Strategic Partnering Decision Year: 1996
Strategic Partnering Decision Business Situation: –$80MM PL in Connecticut and Maine –Good auto results, volatile and unprofitable property results –100 year return time wind loss = 1.5 x Surplus –Catastrophe insurance ROL approaching 4% –Depressed returns from large real estate investment –Ratings pressure
Alternatives Open Market Reinsurance Product – Contingent Surplus notes etc. Demutualization and Sale Merger Reinsurance Pooling Arrangement
Pooling Arrangement All premiums, losses and expenses pooled with member companies Retrospective reinsurance treatment of prior loss reserves Pool results redistributed based on “economic value” surplus Companies maintain separate surplus and investments
DFA Modeling Approach Focus on the underwriting side of the model –underwriting volatility changes –investment projections left deterministic State conclusions in terms of relative rather than absolute returns –eliminate need for economic scenarios etc. Custom built model to accommodate pooling transaction accounting
Underwriting model Accident year loss ratios by line –decomposed historical to cycle + randomness No reserve uncertainties Separate catastrophe loss modeling –Distributions from IRAS model before and after pooling –Catastrophe reinsurance modeling
Model Results Parameter Assumption sets –Forecast (expected )underwriting results –Better than past average results –$100MM catastrophe loss Model outputs –Probability that surplus in 5 years would be X% better under pooling than stand-alone
DFA Benefits Demonstrated significant risk reduction benefit to transaction Helped achieve regulatory approval Kept all “stakeholders” focused on the real issues
Summary Avoid the “model in search of a problem” Avoid “black box” modeling Understand the people you’re influencing Focus on relative performance