New York, N.Y., September 15, 2003 – Advances in stochastic modeling now offer insurers new ways to understand and manage volatile market conditions, according to consulting firm Tillinghast – Towers Perrin (Tillinghast). Tillinghast experts demonstrated stochastic modeling advances at the Stochastic Modeling Symposium on September 4 and 5, 2003. The symposium was sponsored by the Canadian Institute of Actuaries, the Society of Actuaries and the Actuarial Foundation, and cosponsored by the Casualty Actuarial Society and the Conference of Consulting Actuaries.
The turbulent stock market, threats of terrorism and new Sarbanes-Oxley pressures have collectively heightened the demand for better information and financial transparency. Insurance executives now have an even greater need to know and are applying increased internal pressure for improved management information. Best practice companies are turning to the next generation of stochastic modeling techniques to help respond to these concerns.
“The broader use of stochastic models by senior management is essential in today’s uncertain environment,” says Jack Gibson, Tillinghast’s North American Life Insurance and Financial Services Practice Leader. “Management can make better strategic decisions with a more rigorous understanding of the link between risk and value. Stochastic models can also help insurers better anticipate and address the expected volatility of future earnings, as well as the implications of key assumptions being wrong.”
Tillinghast’s Alastair Longley-Cook and Jason Kehrberg were presented with an award at the conference in recognition of their paper “Efficient Stochastic Modeling Utilizing Representative Scenarios.” According to Longley-Cook and Kehrberg, a growing number of financial services organizations are using stochastic modeling of large numbers of cash flow projections to determine capital requirements or assess complex risks. However, there is a need for efficient methodologies that reduce the number of projections to manageable levels. The paper focuses on efficient use of a small number of representative cash flow scenarios that provide representative results, even in the tail of the distribution.
“We were delighted to accept the Outstanding Paper Award at this year’s Stochastic Modeling Symposium. The run-time issue highlighted in our paper is proving to be a significant hurdle for companies; representative scenario methodologies can help reduce run time and enable companies to get more out of their stochastic models,” says Kehrberg. “Many insurance companies are faced with the need to run thousands of stochastic scenarios for internal purposes, and now capital requirements in Canada and the U.S. are moving toward more stochastic modeling.”
The other papers presented at the Stochastic Modeling Symposium addressed other recent advances that extend the theory and application of stochastic modeling techniques:
- “Modeling of Mortality” by David Czernicki, Noel Harewood and Michael Taht: While actuaries have long recognized the stochastic nature of investment risks inherent in insurance, there has been much less focus on the stochastic modeling of insurance risk. This paper explores how stochastic mortality techniques can be used to gain further insights into the economics of insurance products.
- “The Impact of Dynamic Policyholder Behavior on Capital Requirements” by Marc Altschull and Douglas Robbins: It is widely recognized that the choices or standards used to create a scenario set can dramatically affect the results of stochastic modeling. However, it is not always recognized that policyholder behavior, due to variations in economic scenarios, can affect results just as, or more, dramatically. This paper describes the potential effects on required capital levels under C3II testing for variable annuities.
- “Long-Term Equity Returns” by David Bayliffe and Bill Pauling: Volatility of equity returns is an important but often overlooked aspect of actuarial modeling. The volatility of compound returns has a significant impact on product pricing, reserving, economic capital and asset allocation. This paper investigates the distribution of long-term equity returns in order to define a range of estimates for long-term volatility. The authors apply a financial econometric approach to a variety of models to infer the volatility range. Some good short-term models are found to have a poor long-term fit. Models that incorporate fat-tails and mean reversion generally provide superior fits to both short- and long-term data.
“To be useful, stochastic models must be practical,” adds Gibson. “The models need to be simple enough to run quickly, but sufficiently detailed to provide meaningful conclusions. The techniques discussed in these papers represent real advances in making these tools more practical and accessible for the average insurance company.”
About Tillinghast – Towers Perrin
Tillinghast provides actuarial and management consulting to financial services companies and advises other organizations on their self-insurance programs. Tillinghast is a premier independent advisor to the insurance industry; its major clients include most of the world’s top insurers. It operates as one global business, through a network of 42 offices in 20 countries. Tillinghast is a division of Towers Perrin, one of world’s largest management and human resource consulting firms. The Towers Perrin family of businesses also includes Towers Perrin Reinsurance, a leading global reinsurance intermediary. Together, these businesses have over 9,000 employees in 23 countries. More information about Tillinghast is available at www.tillinghast.com.
Tags: predictive modeling, Towers Perrin
