The Cross-Section of Expected Returns: A Case of over Aggregation

Beta, as a measure of risk, is used by many investors and academicians for forecasting the expected earnings. Fama and French (1992) first used data of New York Stock Exchange (NYSE), American Stock Exchange (AMEX), and National Automated Security Dealers Automated Quotes (NASDAQ) from July 1963 to December 1990 in the univariate regression model. They formed ten size-based portfolios and then divided each of these ten portfolios into ten sub-portfolios based on beta. They found that the size-beta portfolio returns show a negative relationship with portfolio betas. Hence, it was concluded that beta does not describe the average stock market returns. In the appendix of their paper they extended the time period to fifty years covering from 1941 to 1990 and arrived on the same results. Hence, it was concluded that beta cannot be salvaged. Many other studies arrived on similar conclusions using data from different countries.
This paper is using the data of NYSE, AMEX, and NASDAQ following the methodology of Fama and French (1992), for a period of 28½ years from July 1964 to December 1992, arrived on the same results as those of Fama and French (1992) that beta does not describe the average stock market returns. After this, many other studied arrived on the similar conclusions in different markets from Latin America to Indonesia. In this paper we made slight adjustment to the data set by dividing the total time period into two parts based on market condition (bear and bull). Using the same model, we will show that beta is still alive, and it is more powerful than previously thought. This result shows that the conclusion of Fama and French (1992) are not the problem of beta, instead it is a problem of over aggregations.

Total Reviews: 0

Add to Wishlist

Product added! Browse Wishlist

The product is already in the wishlist! Browse Wishlist