In portfolio analysis, uncertainty about parameter values leads to suboptimal portfolio choices. The resulting loss in the investor's utility is a function of the particular estimator chosen for expected returns. So, this is a problem of simultaneous estimation of normal means under a well-specified loss function. In this situation, as Stein has shown, the classical sample mean is inadmissible. This paper presents a simple empirical Bayes estimator that should outperform the sample mean in the context of a portfolio. Simulation analysis shows that these Bayes-Stein estimators provide significant gains in portfolio selection problems.

This article explores the price behavior of a sample of corporate securities in which trading was temporarily suspended by the SEC. Suspensions are found to coincide with substantial devaluations of the suspended securities. Further, significant and prolonged negative abnormal returns are observed in the postsuspension period, an apparent violation of semistrong form market efficiency.

Numerous studies have analyzed the stock market reaction to information released in proxy statements. Two possible biases in proxy statement research are suggested frequently: misspecification of the return benchmark and a sample selection bias from analyzing only “clean” events. This paper presents evidence that most of the conclusions of existing studies are not affected by these potential biases. A significantly positive abnormal return was found around a random sample of shareholder meeting dates. The result indicates that interpreting event study results for announcements occurring around annual shareholder meetings must be conducted carefully. The results are consistent with the findings of Kalay and Loewenstein [14], who argue that risk and expected return can increase around predictable, information-producing events. Alternatively, the study can be viewed as being consistent with several recent studies that find anomalous results, using daily return data and large samples.

This paper examines whether investors with power utility functions choose mean-variance-(MV) efficient portfolios when returns are approximately normally distributed and there is borrowing or lending at a riskless interest rate. The results show that the unlevered portfolios of power utility investors plot very closely to the MV-efficient frontier. However, there are marked differences in the mix of risky assets, regardless of whether the portfolios are highly concentrated or widely diversified. Such differences allow power investors to remain solvent even when they lever their optimal portfolios to a greater extent than “less risk-averse” MV investors who risk bankruptcy. It is concluded that the investment policies of power utility and MV investors with similar risk aversion measures are not as similar as is commonly believed. This is particularly true for high power investors, unless explicit solvency constraints are imposed on the MV problem, and for low power investors when quadratic utility approximations are made to the power utility functions. These differences in the investment policies of power utility and MV investors lead us to question the widely-accepted assertion that the assumptions of homogeneous beliefs, normality, a riskless asset, and risk-averse investors imply the simple MV CAPM where all investors, including power utility investors, hold combinations of the market portfolio and the riskless asset.

Recent empirical studies have found ex post common stock returns to be consistently positively skewed. The frequency of positive skewness in this study is found to be relatively stable over varying time periods from 1961 to 1980. However, the skewness of individual stocks and portfolios of stocks does not persist across different time periods. Positively-skewed equity portfolios in one period are not likely to be positively skewed in the next time period. Past positively-skewed returns do not predict future positively-skewed returns.

It is shown that the existing tax law with its incomplete tax-loss offset will often lead to leasing contracts being advantageous to firms. This result obtains even if firms are in the same tax bracket and have the same probability of having positive taxable income.

The advance of the theory of contingent claim pricing has made it possible to model and analyze very complex financial claims. When the value of the firm can be represented as a contingent claim, then the firm's optimal financial policy can be determined with only a slight modification in standard solution techniques for contingent claims. In this paper, stochastic control theory is used to determine a dynamic investment policy for the valuemaximizing firm. The value of future, stochastic economic rents (i.e., the net present value of the firm), and the firm's optimal investment policy must reflect a rational reaction on behalf of its competitors. A computationally efficient methodology is presented for solving the simultaneous investment-valuation problem for an n-firm game.

Recent studies indicate that the widespread assumption of parameter stationarity in empirical applications of asset pricing models may be inappropriate. This paper investigates the feasibility of modeling parameter instability as a sequence of persistent stable regimes. Recursive residual and log likelihood techniques are combined to detect and locate shift points. The results indicate that regime shifts are widespread, frequent, and often large enough to significantly effect empirical findings. The nature of the shifts appears to be a rotation of the regression line, indicating that correction of both alpha and beta parameters is required.

When portfolio optimization is implemented using the historical characteristics of security returns, estimation error can degrade the desirable properties of the investment portfolio that is selected. Given the problem of estimation risk, it is natural to formulate rules of portfolio selection within a Bayesian framework. In this framework, portfolio selection is based on maximization of expected utility conditioned on the predictive distribution of security returns. Most researchers have addressed the problem of estimation risk by asserting a noninformative diffuse prior that reduces the detrimental effect of estimation risk, but does not directly reduce estimation error. Portfolio performance can be improved by specifying an informative prior that reduces estimation error. An informative prior that all securities have identical expected returns, variances, and pairwise correlation coefficients is asserted. This informative prior reduces estimation error by drawing the posterior estimates of each security's expected return, variance, and pairwise correlation coefficients toward the average return, average variance, and average correlation coefficient, respectively, of all the securities in the population. The amount that each of these parameters is drawn toward its grand mean depends upon the degree to which the sample is consistent with the informative prior. This empirical Bayes method is shown to select portfolios whose performance is superior to that achieved, given the assumption of a noninformative prior or by using classical sample estimates.