New Estimates of the Market Parameters
使用随机系数回归模型（RCR）更有效地估计资本资产定价模型（CAPM）和Black双因子模型的市场参数（风险溢价和无风险利率），并与Fama-MacBeth方法的结果进行比较，同时采用价值加权指数作为市场组合的代理变量。
Over the past decade, the Capital Asset Pricing Model (CAPM) and the two-factor model of Black [1] have been the premier pricing models for the study of equity markets in finance. Fama and MacBeth (FM) [4] developed an empirical procedure for estimating the market parameters of these models (the premium and the riskless return). Many studies of capital market efficiency have used their methodology (for example, see [6, 7]). Given the importance of the FM methodology, an improved empirical technique for estimating these parameters over specified periods should contribute to the use of the model and permit more refined tests of the fundamental hypothesis that there is a positive relationship between risk and return. Here we employ a more efficient econometric procedure for estimating the means of the market parameters and then compare these estimates with the estimates obtained using the FM methodology. In addition, we also test the model with a more theoretically correct proxy for the market portfolio, the value-weighted index. The econometric technique employed in this paper to estimate the market parameters is the Random Cofficient Regression (RCR) model developed by Swamy [10]. Fabozzi and Francis [2] use a RCR to estimate individual stock betas. But there is little theoretical justification to assume that stock betas move randomly through time. In fact, a firm can deliberately and systematically affect its beta by changing its asset composition and/or by changing its financial structure. There is justification for employing RCR to estimate the market parameters, however, as we will demonstrate.