There is a false, but widely-held belief about orthogonal (“zero-beta”) portfolios: for a given market index, all zero-beta portfolios have the same expected return and the minimal-variance, zero-beta portfolio is unique. This is true only when the index is mean/variance efficient. Every nonefficient index possesses zero-beta portfolios at all levels of expected return. For a given index, minimal-variance zero-beta portfolios corresponding to different expected returns lie along an “orthogonal frontier” in the mean/variance space. The frontier has some unusual properties which turn out to be relevant for empirical work on asset pricing. It is functionally related to deviations about the “securities market line.”

In this paper, we present a new version of the capital asset pricing model CAPM) that provides a linear pricing equation substantially different from that implied by the traditional CAPM of Sharpe [18], Lintner [12], and Mossin [14, 15] (hereafter SLM model). It is assumed that each of the investors has an initial endowment of real resources (say, corn) which can be either consumed invested in investment opportunities available to the investor. A set of simultaneous equations is derived from the model. The set of equations determines the equilibrium values of these interdependent endogenous variables: the amount to be consumed by the investor; the proportion of each investment project be owned by the investor; the amount to be invested in each of the available investment projects; the market value of each project; the market price of risk; and the return imputed by the capital market for a risky project which has a zero-beta risk. If a riskless project exists, the zero-beta rate is just a risk-free rate.

The Capital Asset Pricing Model (CAPM) developed and popularized by Treynor [27], Sharpe [26], Lintner [16], Mossin [19], and Fama [6] is of the form

where

E is the expected return at time t for firm i (conditional on information available at time t); the subscript m denotes the analogous market variable; rf is the risk-free rate; and . Black [2] has developed a similar form with expected returns from a zero beta portfolio, , assuming the role of the risk-free rate.

Recently the standard market model has been used to examine holding period returns of corporate bonds. These studies have involved issues such as: the impact of accounting earnings data on bond price behavior [5]; the relationship of bond betas and ratings [19, 21]; the effect of ratings changes on bond prices [27]; the relationship of bond betas to duration and yield [3, 13, 15]; bond performance of bankrupt and nonbankrupt firms [26]; and tests of the Capital Asset Pricing Model based on bond returns [7]. While the empirical appropriateness of applying the market model to common stock returns has been demonstrated, similar tests have not been conducted with regard to long-term corporate bonds. Section II of this paper will examine the assumptions of the normal error regression model when used in the form of the market model and applied to a sample of long-term corporate bonds during the early years of their lives. The issue of systematic changes in the regression parameters will be addressed in Section III. Lastly, conclusions will be presented in Section IV.

Most of the recent empirical tests of the Black-Scholes option-pricing formula have been joint tests of three types of hypotheses:

1) mathematical structure of the formula,

2) measurement of formula inputs and outputs, and

3) the efficiency of the listed option market.

Based on the theory of the pricing of capital assets developed by Sharpe [12], Lintner [9] and Mossin [11], Professor Jensen formulated a return-generating model to measure portfolio performance [5]. In a subsequent paper, Professor Jensen [6] investigated the impact of the investment horizon on the functional form of the model. Lee [8] has proposed a generalized specification of the model to resolve this problem. Alternative estimation methods for testing the linearity of the model in terms of time-series data have also been suggested by Lee. Moreover, the stability of the beta coefficient over time and the impact of the market's condition on both the alpha (or, Jensen's measure of performance [5]) and beta of the model have come under scrutiny in financial research.

The purpose of this paper is to demonstrate mathematically that the skewness of securities' returns--the ratio of the third moment to the standard deviation cubed--is sensitive to the length of the differencing interval over which returns are measured. Empirical observations of this so-called intervaling effect on skewness have been reported in at least three articles in this Journal. There have been no attempts, however, to examine this effect analytically. The empirical evidence presented in the literature is often contradictory and remains unexplained because of a lack of an analytical insight into the causes of the intervaling effect.

Unequal costs of obtaining and processing information may lead to trading of securities and wealth redistributions among investors. Those investors with easy access to information about a firm may be able to profit from prior knowledge of the information before public release. Public policy making bodies such as the SEC have attempted to alleviate this phenomenon by promoting public disclosure of information through litigation and regulation of the trading activities of insiders. Whether these procedures have been successful in curtailing trading due to privileged information is still open to debate. Academicians have also been concerned with resolving the existence of asymmetrically distributed information and “efficient markets.” In spite of the social and academic importance of this phenomenon, there has been little empirical work in this area. This is primarily due to a lack of a testable theoretical framework explaining investor behavior in securities markets with asymmetric information distribution. The ensuing paper provides a tentative testable theory on trading in markets with asymmetrically distributed information as well as an empirical investigation of this theory.

In capital budgeting when the net present value method (NPV) is used and future cash flows are uncertain, the measurement of the standard deviation of the project's net present value (σNPV) is unusually difficult. Typically the investment life of the project is more than one year, and the cash flows of each period are correlated with each other. The difficulty in determining the exact inter-temporal correlation (ρt, t+l) of cash flows creates special problems in estimating the overall variance of the investment project. For example, see [5, pp. 125–131]. This paper attempts to isolate the possible impact of the inter-temporal correlation of cash flows upon the total risk of the project under the assumption that the standard deviations of cash flows for each time period of the project are known. It also provides upper and lower bounds of σNPV for a multi-period investment project where the degree of the inter-temporal correlation is not exactly known.

During the past 20 years several theoretical valuation models with empirically testable implications regarding the impact of financial decisions, particularly debt policy and dividend policy, have appeared in the financial literature. Some of these models have been subjected to extensive testing. However, results of these investigations are inconclusive and/or conflicting as to the impact of financial policies on share value. For instance, on the basis of a study of electric utility firms, Miller and Modigliani (M&M) find that neither dividend nor debt policies of the firm affect the value of the firm. On the other hand, Weston, as well as Brigham and Gordon, disagree with M&M's findings based upon their own investigations of electric utility firms.