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.

When measures such as the average or the midrange are used to report a typical value for the price series in each interval, the stochastic character of the underlying price process is subtly transformed. Fortunately, the spurious serial dependence introduced by averaging measures is sufficiently well understood to allow direct tests of many hypotheses to be made from averaged data. Moreover, a simple autoregressive transformation of the averaged data can be used to unscramble the effects of averaging on the lower-frequency components of the spectrum of the underlying process. These statistical devices are presented and are then illustrated by applications to the Cowles Commission Common- Stock Indexes, a massive collection of New York Stock Exchange price indexes tabulated in the form of monthly midranges.

Ex ante predictions of the riskiness of returns on common stocks — or, in more general terms, predictions of the probability distribution of returns — can be based on fundamental (accounting) data for the firm and also on the previous history of stock prices. In this article, we attempt to combine both sources of information to provide efficient predictions of the probability distribution of returns. We predict two parameters of the distribution of returns for each security in each year: the response to the overall market return (β), and the variance of the part of risk, specific to the security, that is uncorrelated with the market return. A cross section of time series data on returns and accounting variables, taken primarily from the Compustat tape, is used. Several recent developments in statistical methodology are applied.

A number of studies have compared the investment risk of various industries and of various individual corporations in developed countries (DCs). The purpose of this paper is to compare investment risk in DCs with less developed countries (LDCs). The variance of returns to investment in common stocks provides a natural measure of investment risk and will be used in this study. Studies in LDCs include work of Levy and Sarnat [12], [13]. Errunza [3], [4], and Lessard [11]. Levy and Sarnat and Errunza found low economy-wide investment risk on an average for LDCs (relative to DCs), with stock indices being used as surrogates for economy-wide risk. These results are not unambiguous, however, because there are probable difficulties due to infrequent trading and averaging in the broad market indices used in the above studies. Hence, we use a sample of the largest corporations that suffer little, if at all, from thin trading and/or averaging.

Barr Rosenberg, Andrew Rudd, Factor-Related and Specific Returns of Common Stocks: Serial Correlation and Market Inefficiency, The Journal of Finance, Vol. 37, No. 2, Papers and Proceedings of the Fortieth Annual Meeting of the American Finance Association, Washington, D.C., December 28-30, 1981 (May, 1982), pp. 543-554

Andrew Rudd, Barr Rosenberg, The "Market Model" In Investment Management, The Journal of Finance, Vol. 35, No. 2, Papers and Proceedings Thirty-Eighth Annual Meeting American Finance Association, Atlanta, Georgia, December 28-30, 1979 (May, 1980), pp. 597-607