A Forecast-Based Approach to Economic Capital Models in the Insurance Industry

• Activity Number: 442
• Type: Topic Contributed
• Date/Time: Wednesday, August 6, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Business and Economic Statistics Section
• Author(s): Alan Kessler*+ and Scott Farris
• Companies: State Farm Insurance and State Farm Insurance
• Keywords: economic capital ; forecast ; insurance

Abstract

In the insurance industry, it is important that underwriting economic capital models relate to pricing, reserving, and other company planning mechanisms. By incorporating the methods used in these activities, the model best reflects the underwriting risk. We will discuss a model with both the flexibility and sophistication to generate one million unique scenarios of operating returns compatible with current pricing methods. The output is made up of several components including losses, exposure, and expenses. Each of the components is forecast using ARIMA models. The usual normality assumption around forecasts is relaxed so that univariate distributions with more representative tail values are used for simulation. Traditional loss development methods are applied to forecasts to obtain reserve risk simulations. Hurricane and earthquake catastrophes are added using vendor models. Other catastrophe outcomes use forecast methods in conjunction with compound distribution models for simulations of counts and event amounts. The different components are combined to calculate operating returns and combined with investment simulations to create economic capital results.