Knowledge that Transforms

The purpose of this study is to measure and evaluate the objectives, risk, and return of 123 American mutual funds using monthly returns in the period 1960–1969. The paper considers five questions: How were stated fund objectives related to risk and return, as measured over the subsequent decade? How did funds of various objectives perform in terms of return and return-to-risk measures? Did average excess return increase with risk? Was the return-to-risk performance of the average mutual fund better or worse than that of the stock market as a whole? How did the slope of the mutual fund line of returns versus beta compare to the capital market line; i.e., did funds at one end of the risk spectrum appear to “outperform” those at the other end?

In the linear market-index model of the return-generating process, return on security j is given bywhere αj and βj are constants characteristic of company j, is return on a market index, and is the company-specific component of return such that and . The coefficient βj is given by . It is known as market responsiveness, volatility, systematic risk, and, more commonly, simply as “beta.” It has been widely accepted as a measure of nondiversifiable risk and incorporated in popular performance measures. Many stock information services now provide estimates of beta.

Norman Toy, Arthur Stonehill, Lee Remmers, Richard Wright, Theo Beekhuisen, A Comparative International Study of Growth, Profitability, and Risk as Determinants of Corporate Debt Ratios in the Manufacturing Sector, The Journal of Financial and Quantitative Analysis, Vol. 9, No. 5, 1974 Proceedings (Nov., 1974), pp. 875-886

This study is concerned with the multiple-factor model of security returns, with its implications for a single-factor, market-index model applied to the same securities, and with statistical methods of estimating the parameters of a multiple-factor model and thereby operationalizing it. This part of the paper sets out the approach. The sequel will present the empirical results. The results show that there are highly significant extra-market components of covariance among security returns; moreover, these risk components are such that the loadings of individual security returns on the factors are determined by observable characteristics of the firm: income statement and balance sheet data, industry membership, and historical behavior of returns on the security. The results also show that the conventional security beta is a function of these same characteristics.

William W. Hogan, James M. Warren, Toward the Development of an Equilibrium Capital-Market Model Based on Semivariance, The Journal of Financial and Quantitative Analysis, Vol. 9, No. 1 (Jan., 1974), pp. 1-11

A link between the firm's operating decisions and the riskiness of its stocks was established. Differences in the production process affecting the relative shares of fixed and variable costs (i.e., the operating leverage) were found, both analytically and empirically, to be associated with risk differentials. Specifically, other things equal, the higher the operating leverage (i.e., the lower the unit variable costs) the larger the overall and systematic risk of the stocks.Various practical implications are suggested by these findings. On the firm level, it can be expected that large capital expenditures associated with an operating leverage increase will increase stock riskiness. In these cases, the cut-off rate used for the capital budgeting decision (i.e., the cost of capital) should allow for the increased risk. The use of the current cost of capital as the cut-off rate would probably result in a decrease in stock prices, adversely affecting stockholders' wealth. On the investor level, these findings might assist in the estimation of common stocks' risk given expected changes in the firm's operating leverage. Specifically, they suggest that, if a firm will experience a significant operating leverage change, the estimation of risk measures based exclusively on historical returns would be inappropriate.

In considering trade credit, we need to ask three questions:1. Why do nonfinancial firms commonly participate in the process of financial intermediation by extending credit to their customers?2. What explains differences in credit periods between firms and industries as well as over time for specific firms and industries?3. How do changing monetary conditions affect the credit that firms extend to their customers?To answer these questions, we first identify two reasons for credit sales: the first we might call a financing motive, and the second a transactions motive. The transactions motive can readily be understood—it costs something to match the time pattern of payment for goods with the time pattern of receipt of goods. Buyers benefit if bills are allowed to accumulate for periodic payment. Furthermore, trade credit gives buyers time to plan for the payment of unexpected purchases, enables them to forecast future cash outlays with greater certainty, and simplifies their cash management. To the extent that buyers benefit, sellers have an opportunity to sell credit. It is likely that, to a large extent, the aggregate stock of trade credit is explained by the transactions motive.

Although investors face multiperiod decision problems, there are conditions under which the results of the one-period two-parameter model apply period by period. In addition to the assumptions made in the development of the two-parameter model itself (a perfect capital market, investor risk aversion, and normal distributions of one-period portfolio returns), the critical assumption in a multiperiod context is that, for any t, returns on portfolio assets from t−1 to t are independent of stochastic elements of the state-of-the-world at time t that affect investor tastes for given levels of wealth to be obtained at t. One such element of the state-of-the-world is the nature of investment opportunities to be available at t. For example, if the level of expected returns on investment portfolios to be available at time t is uncertain at time t−1, and if the returns from t−1 to t on some investment assets are more strongly related to the level of expected returns at t than returns on other assets, then the former assets are better vehicles for hedging against the level of expected returns at t. This can affect the demands for assets and their prices in such a way that the simple results of the one-period two-parameter model do not hold. The empirical tests of this paper reveal no evidence of measurable relationships between the returns on portfolio assets from t−1 to t and the level of expected returns to be available at t. Indeed, in our opinion there is no reliable evidence that the level of expected returns changed during the 1953–1972 period.

Much controversy surrounds the use of the portfolio investment rules induced by maximizing the expected logarithm of terminal wealth (henceforth referred to as the MEL policy). It has been thought that the MEL policy is a good approximation to the optimal investment program when the utility of terminal wealth function is bounded and when the time horizon is long. However, I exhibit a class of bounded utility of terminal wealth functions for which the MEL policy is a very poor approximation to the optimal program. Hence, the wholesale use of the MEL policy as an approximation to the optimal program is unwarranted.

This paper examines stock market efficiency with respect to money supply data by testing (1) regression models of stock returns on monetary variables and (2) trading rules based on money supply data. The evidence indicates no meaningful lag in the effect of monetary policy on the stock market and that no profitable security trading rules using past values of the money supply exist. Therefore this evidence is consistent with the efficient market model. Current security returns incorporate all information contained in past money supply data and, in addition, appear to anticipate future changes in the money supply. A number of previous studies have concluded that lags exist and can be used in profitable trading rules. Analysis of these studies demonstrates that for a variety of reasons the evidence in these past studies does not sustain such conclusions.