Description: This presentation is a Teaching Session, number 38TS, from the 1999 Valuation Actuary Symposium, held September 23-24 in Los Angeles. The instructors provide an introduction into how complex adaptive systems [CAS] can be applied to solve actuarial problems. They provide a general overview of CAS, including genetic algorithms, neural network, nonlinear modeling, and agent-based modeling, provide recent examples of applications to business problems, and discuss how CAS approaches can be used to address actuarial pricing and valuation issues, such as modeling customer behavior. Discussion from participants is included.

Description: This presentation is a panel discussion, session number 34PD, from the 2002 Valuation Actuary Symposium, held September 19-20 in Lake Buena Vista, FL. A fundamental step in stochastic asset/liability modeling is to choose an appropriate underlying economic scenario generator. In this session, panelists discuss various scenario generation techniques and theories, assumptions underlying interest and equity scenario generators, and choosing suitable economic scenario generators for different applications, the two most common being pricing models [for assets and liabilities], and risk analysis models, used for asset/liability management, investment strategy, regulatory solvency, and capital adequacy. Discussion from participants is included.

Description: With the adoption of the C-3 Phase 2 Project of the American Academy of Actuaries, it is now required that stochastic scenarios be developed for establishing reserves for variable annuities. This project has led to the idea of Principle-Based Reserving PBR where many of the same techniques will be used to set life insurance reserves. This article describes the choice of a Random Number Generator RNG to develop the random scenarios.

Description: This presentation is a Teaching Session, number 3, from the 1987 Valuation Actuary Symposium. The first instructor discusses three methods of valuation [and methods of studying C-3 risk]: Gross Premium Valuation [the actuary’s traditional approach], Cash Flow Analysis, and Market Value Analysis, also known as ‘Modern Financial Analysis,’ with a focus on the third. Then the second instructor provides an update on C-2 risk [claims, expenses], and combinations of risk, noting that C-2 risk has not had much attention paid to it. He goes on to describe three types of C-2 risk: stochastic deviations, unpredictable single events, and ‘judgment errors, such as poor underwriting. The third instructor focuses on the C-1 [default] risk. Case studies are included.

Description: This session from the 1993 Valuation Actuary Symposium discusses the growth and deterioration of commercial mortgages and the modeling of commercial mortgages, real estate and CMO's.

Description: The author summaries several reports from the IBM Symposium on operations research techniques and their future outlook 1968.

Description: This abstract describes a paper that investigates model uncertainty arising from different ways of treating the operational loss data collection threshold.

Description: This abstract describes a paper that proposes a multi-population Global Vector Autoregression (GVAR) mortality model, in which the joint mortality dynamics of all populations are formulated in a large vector autoregression model with a set of contemporaneous global mortality factors.

Description: The Financial Reporting and Product Development Sections, along with the Committee on Life Insurance Research are pleased to make the following report available. Prepared by Bruce Rosner of Ernst & Young, the report summarizes the findings of a stochastic modeling efficiency study. With the help of twelve life insurance and annuity companies, the study tested six techniques for reducing the number of model points or scenarios required to achieve a given level of precision in stochastic actuarial modeling.

Description: This article discusses value at risk, VaR, and VaR modeling of the bank's portfolios that would recognize diversity and hedging. This article also covers the differences between classical VaR and the longer horizons required by the insurance industry and the feasibility of extending VaR to the insurance industry.