Quarterly Review
Introduction to Investment Strategy

Review of Active Portfolio Management by Richard C. Grinold and Ronald N. Kahn (Probus Publishing Company, 1995, $65).

by Robert J. Finger

It is likely that investments will become a much more important field for the casualty actuary in the future. Dynamic Financial Analysis may become almost as important as reserving and ratemaking to practicing actuaries. Life and pension actuaries receive more training in investments than casualty actuaries. In addition, there has been consolidation within the financial services sector, which is likely to continue. A better understanding of investment practices and principles will be useful to the casualty actuary in at least two ways: in understanding the asset side of the balance sheet and in understanding how property/casualty operations are viewed by the owners.

I believe Active Portfolio Management provides an excellent overview of current investment theory and practice. It has an appropriately mathematical perspective (including extensive technical appendices) and it describes a broad range of theories and practices in clear terms. It focuses on stock investments (which may not be as relevant to property/casualty companies), but it provides a process that is applicable to all types of investments, including bonds, derivatives and currencies. After reading the book, one should have a better idea of how to manage investment advisors and what one can consider a realistic level of investment performance.

The book is written from the perspective of a manager of institutional assets. The manager has been given a certain amount of assets and is charged with investing it in a certain asset class (for example, domestic large capitalization stocks). The manager's performance will be judged against a benchmark such as the Standard & Poor's 500 Index. The basic theme of the book is that the manager can only do better than the benchmark if he/she has better information regarding the expected returns of individual investments within the asset class. If the manager's estimates are the same as the market's consensus estimates of expected returns, the manager should invest in the benchmark. If the manager's estimates are at least a little bit better than the consensus estimates, the manager can add value to the investment decision by overweighting the investments where he/she expects higher than consensus returns. Of course, this added value can be dissipated by inefficient portfolio construction and transaction costs.

A great deal of research has questioned whether the capital markets are efficient and if any given investment advisor actually can beat the market. This book takes the position that it is possible, but difficult, to beat the market. It develops concepts and tools to deal with this issue. It does not provide the answer because yesterday's answer is unlikely to be today's or tomorrow's answers.

Perhaps the most important concept is the "Fundamental Law of Active Management." This law states that the value added by a strategy will be proportional to the "information ratio" squared. The information ratio is the product of the "information coefficient" and the square root of the strategy's "breadth." The information coefficient is a measure of the manager's skill (at making better asset value forecasts than the market's consensus estimates). The breadth is the number of independent forecasts (per year). If the information coefficient is zero, the manager adds no value and should invest in the benchmark. Breadth will be higher if the manager makes more estimates (for example, follows 1,000 stocks rather than 100) or makes estimates more often (quarterly rather than annually). Of course, it should be more difficult to make equally good estimates on more stocks or independent estimates more often.

Part I of the book includes five chapters on theoretical foundations. Chapter 2 discusses the Capital Asset Pricing Model. The authors discuss various limitations on CAPM, but conclude that it provides a good structure for active portfolio management. Indeed, the authors use a mean-variance approach for virtually all of their analyses. Other approaches, such as shortfall risk, do not fit easily into their framework.

Chapter 3 discusses risk, which the authors define as the standard deviation of return. Of interest is their "multiple-factor" risk model, which structures the variability in returns of various assets into an "exposure" element, for each asset, and an excess "return" element, for each factor. For example, returns may be divided into returns due to such factors as: external economic influences, industry, market capitalization, dividend yields or earnings momentum. The key is for the active manager to devise and test various models that will efficiently explain historical returns. Better models (and better forecasts of important factors) may yield better forecasts of individual asset returns.

Chapter 4 discusses exceptional returns, benchmarks, and added value. Exceptional returns are returns in excess of the consensus forecasts of return. This chapter also quantifies the risk in taking a strategy that deviates from the benchmark. The book does not discuss strategic asset allocation, which is the decision by the asset owners to allocate assets to various investment classes.

Chapter 5 deals with the "information ratio," which is defined as the ratio of the residual return (versus the benchmark) to the residual risk. It also discusses the "residual risk aversion," which is the aggressiveness with which the manager chooses to exploit superior information. Chapter 6 ties all of the foundations together in the "Fundamental Law of Active Management," which was discussed above.

Part II includes three chapters on expected returns and valuation. Chapter 7 explains the Arbitrage Pricing Theory. This is presented as a multiple-factor model of expected returns. Chapter 8, "Valuation in Theory," presents discounted cash flow models, including risk-adjusted interest rates. Chapter 9, "Valuation in Practice," discusses dividend discount models.

Part III discusses implementation. Chapter 10, "Forecasting," is concerned primarily with the variability of forecasts. Chapter 11, "Information Analysis," shows how exceptional returns from forecasts can be measured. Chapter 12 describes "Portfolio Construction." Chapter 13 explains "Transactions Costs, Turnover, and Trading." An important insight is that 75 percent of the value added usually can be gained with about 50 percent of turnover (and transactions costs). Chapter 14, "Performance Analysis," has the goal of distinguishing the skilled from the unskilled manager. Chapter 15 discusses "Benchmark Timing" and concludes that it is very difficult to add value due to timing the market (such as, cash versus fully invested positions).

In summary, the book provides a good discussion of many important investment concepts and principles.