The problem of solvency is, in fact, as old as insurance. The history of the industry knows many ways to meet the risks involved with underwriting, such as spreading the risk portfolio (Cato, Senior already applied it), risk selection, reserve funds, reinsurance, etc. Whilst these measures too often proved ineffective, the establish ment of legislative control and public supervision ensued. However, not until the last few decades has the solvency issue become an ob ject of intensive studies, very much thanks to the progress of related empirical and theoretical knowledge, and in the under standing of the concerned complicated processes. The research activities have grown extensively in many countries in recent years. The more the studies advance the more new relevant aspects are detected and a great variety of alternative proposals have come up for discussion. Therefore, it has become necessary to attempt a survey of the whole problem area in order to be able to place the quite numerous pieces of knowledge in their proper context, and also, among other things, to avoid the pitfalls of handling isolated problems omitting vital tie-ins to the environment. Many of the rele vant problems and subproblems are still lacking adequate and well tested solutions. Therefore, a survey of the whole problem area can also hopefully serve as guidance for future research efforts."

Two different applications have been considered, automobile claims from Massachusetts and health expenses from the Netherlands. We have fit 11 different distributions to these data. The distributions are conveniently nested within a single four parameter distribution, the generalized beta of the second type. This relationship facilitates analysis and comparisons. In both cases the GB2 provided the best fit and the Burr 3 is the best three parameter model. In the case of automobile claims, the flexibility of the GB2 provides a statistically siE;nificant improvement in fit over all other models. In the case of Dutch health expenses the improvement of the GB2 relative to several alternatives was not statistically significant. * The author appreciates the research assistance of Mark Bean, Young Yong Kim and Steve White. The data used were provided by Richard Derrig of The Massachusetts Automobile Rating and Accident Prevention Bureau and by Bob Van der Laan and The Silver Cross Foundation for the medical insurance claim data. 2~ REFERENCES Arnold, B. C. 1983. Pareto Distributions. Bartonsville: International Cooperative Publishing House. Cummins, J. D. and L. R. Freifelder. 1978. A comparative analysis of alternative maximum probable yearly aggregate loss estimators. Journal of Risk and Insurance 45:27-52. *Cummins, J. D., G. Dionne, and L. Maistre. 1987. Application of the GB2 family of distributions in collective risk theory. University of Pennsylvania: Mimeographed manuscript. Hogg, R. V. and S. A. Klugman. 1983. On the estimation of long tailed skewed distributions with actuarial applications.

The First International Conference on Insurance Solvency was held at the Wharton School, University of Pennsylvania from June 18th through June 20th, 1986. The conference was the inaugural event for Wharton's Center for Research on Risk and Insurance. In atten dance were thirty-nine representatives from Australia, Canada, France, Germany, Israel, the United Kingdom, and the United States. The papers presented at the Conference are published in two volumes, this book and a companion volume, Classical Insurance Solvency Theory, J. D. Cummins and R. A. Derrig, eds. (Norwell, MA: Kluwer Academic Publishers, 1988). The first volume presented two papers reflecting important advances in actuarial solvency theory. The current volume goes beyond the actuarial approach to encom pass papers applying the insights and techniques of financial economics. The papers fall into two groups. The first group con sists of papers that adopt an essentially actuarial or statistical ap proach to solvency modelling. These papers represent methodology advances over prior efforts at operational modelling of insurance companies. The emphasis is on cash flow analysis and many of the models incorporate investment income, inflation, taxation, and other economic variables. The papers in second group bring financial economics to bear on various aspects of solvency analysis. These papers discuss insurance applications of asset pricing models, capital structure theory, and the economic theory of agency."

Predictive modeling involves the use of data to forecast future events. It relies on capturing relationships between explanatory variables and the predicted variables from past occurrences and exploiting this to predict future outcomes. Forecasting future financial events is a core actuarial skill actuaries routinely apply predictive-modeling techniques in insurance and other risk-management applications. This book is for actuaries and other financial analysts who are developing their expertise in statistics and wish to become familiar with concrete examples of predictive modeling. The book also addresses the needs of more seasoned practicing analysts who would like an overview of advanced statistical topics that are particularly relevant in actuarial practice. Predictive Modeling Applications in Actuarial Science emphasizes life-long learning by developing tools in an insurance context, providing the relevant actuarial applications, and introducing advanced statistical techniques that can be used by analysts to gain a competitive advantage in situations with complex data."

Two different applications have been considered, automobile claims from Massachusetts and health expenses from the Netherlands. We have fit 11 different distributions to these data. The distributions are conveniently nested within a single four parameter distribution, the generalized beta of the second type. This relationship facilitates analysis and comparisons. In both cases the GB2 provided the best fit and the Burr 3 is the best three parameter model. In the case of automobile claims, the flexibility of the GB2 provides a statistically siE;nificant improvement in fit over all other models. In the case of Dutch health expenses the improvement of the GB2 relative to several alternatives was not statistically significant. * The author appreciates the research assistance of Mark Bean, Young Yong Kim and Steve White. The data used were provided by Richard Derrig of The Massachusetts Automobile Rating and Accident Prevention Bureau and by Bob Van der Laan and The Silver Cross Foundation for the medical insurance claim data. 2~ REFERENCES Arnold, B. C. 1983. Pareto Distributions. Bartonsville: International Cooperative Publishing House. Cummins, J. D. and L. R. Freifelder. 1978. A comparative analysis of alternative maximum probable yearly aggregate loss estimators. Journal of Risk and Insurance 45:27-52. *Cummins, J. D., G. Dionne, and L. Maistre. 1987. Application of the GB2 family of distributions in collective risk theory. University of Pennsylvania: Mimeographed manuscript. Hogg, R. V. and S. A. Klugman. 1983. On the estimation of long tailed skewed distributions with actuarial applications.

The problem of solvency is, in fact, as old as insurance. The history of the industry knows many ways to meet the risks involved with underwriting, such as spreading the risk portfolio (Cato, Senior already applied it), risk selection, reserve funds, reinsurance, etc. Whilst these measures too often proved ineffective, the establish ment of legislative control and public supervision ensued. However, not until the last few decades has the solvency issue become an ob ject of intensive studies, very much thanks to the progress of related empirical and theoretical knowledge, and in the under standing of the concerned complicated processes. The research activities have grown extensively in many countries in recent years. The more the studies advance the more new relevant aspects are detected and a great variety of alternative proposals have come up for discussion. Therefore, it has become necessary to attempt a survey of the whole problem area in order to be able to place the quite numerous pieces of knowledge in their proper context, and also, among other things, to avoid the pitfalls of handling isolated problems omitting vital tie-ins to the environment. Many of the rele vant problems and subproblems are still lacking adequate and well tested solutions. Therefore, a survey of the whole problem area can also hopefully serve as guidance for future research efforts."

The First International Conference on Insurance Solvency was held at the Wharton School, University of Pennsylvania from June 18th through June 20th, 1986. The conference was the inaugural event for Wharton's Center for Research on Risk and Insurance. In atten dance were thirty-nine representatives from Australia, Canada, France, Germany, Israel, the United Kingdom, and the United States. The papers presented at the Conference are published in two volumes, this book and a companion volume, Classical Insurance Solvency Theory, J. D. Cummins and R. A. Derrig, eds. (Norwell, MA: Kluwer Academic Publishers, 1988). The first volume presented two papers reflecting important advances in actuarial solvency theory. The current volume goes beyond the actuarial approach to encom pass papers applying the insights and techniques of financial economics. The papers fall into two groups. The first group con sists of papers that adopt an essentially actuarial or statistical ap proach to solvency modelling. These papers represent methodology advances over prior efforts at operational modelling of insurance companies. The emphasis is on cash flow analysis and many of the models incorporate investment income, inflation, taxation, and other economic variables. The papers in second group bring financial economics to bear on various aspects of solvency analysis. These papers discuss insurance applications of asset pricing models, capital structure theory, and the economic theory of agency."

Predictive modeling uses data to forecast future events. It exploits relationships between explanatory variables and the predicted variables from past occurrences to predict future outcomes. Forecasting financial events is a core skill that actuaries routinely apply in insurance and other risk-management applications. Predictive Modeling Applications in Actuarial Science emphasizes life-long learning by developing tools in an insurance context, providing the relevant actuarial applications, and introducing advanced statistical techniques that can be used to gain a competitive advantage in situations with complex data. Volume 2 examines applications of predictive modeling. Where Volume 1 developed the foundations of predictive modeling, Volume 2 explores practical uses for techniques, focusing on property and casualty insurance. Readers are exposed to a variety of techniques in concrete, real-life contexts that demonstrate their value and the overall value of predictive modeling, for seasoned practicing analysts as well as those just starting out.