Abstracts

Quality & Quantity -     

USING BAYESIAN BETAS TO ESTIMATE RISK RETURN PARAMETERS: AN EMPIRICAL INVESTIGATION

Since its original development by Sharpe (1964), the Capital Asset Pricing Model (CAPM) has been the focus of great interest, practical usage, modifications, testing, and controversy. The basic hypothesis of the CAPM is that the minimum expected return required by investors on any asset will equal the risk-free rate plus a premium for the asset's contribution to the variance risk of a diversified portfolio as measured by the asset's beta. The model is often utilized by investors to calculate the relevant risk and required return on an asset, while corporate officers widely employ the theory to compute the appropriate discount rate to use in estimating the net present value of capital budgeting projects when evaluating spending decisions (Gitman and Mercurio, 1982).     

USING THE CAPM AS A GENERAL FRAMEWORK FOR ASSET PRICING ANALYSIS

In this paper a model of asset pricing is derived that relaxes the restrictive assumptions of standard theories. The model leads to a tractable capital asset pricing model framework that can be used to test empirical hypotheses and draw important insights on capital market equilibrium.     

Empirical evidence of a positive inflation premium being incorporated into stock prices

This study demonstrates that in contrast to prior research findings on short-term stock returns, long-term stock returns are positively correlated with inflation. In addition, within the context of a more complete explanatory model, long-term stock returns are found to be negatively related to changes in long-term interest rates and negatively related to beginning price to earnings ratios. The significance of these variables in explaining almost all the time series variation in long-term stock returns demonstrates that changes in stock values are well explained by theory.     

Computations for constrained linear models

The article presents an algorithm for linear regression computations subject to linear parametric equality constraints, linear parametric inequality constraints, or a mixture of the two. No rank conditions are imposed on the regression specification or the constraint specification. The algorithm requires a full Moore-Penrose g-inverse which entails extra computational effort relative to other orthonormalization type algorithms. In exchange, auxiliary statistical information is generated: feasibility of a set of constraints may be checked, estimability of a linear parametric function may be checked, and bias and variance may be decomposed by source.