Financial Analysts Journal, November/December (1997): 54-66
Campbell Harvey
Duke University, Durham, NC 27708
John Graham
Duke University, Durham, NC 27708
ABSTRACT
We grade the performance of 326 newsletter asset allocation strategies over the 1983-1995 period. These newsletters offer market timing advice. That is, they are supposed to recommend increased stock market weights before market appreciations and decreased weights before market declines. However, our evidence suggests that newsletters as a group do not possess any special information about the future direction of the market. Nevertheless, we present evidence that investment newsletters that are on a hot streak (have correctly anticipated the direction of the market in previous recommendations) appear to have valuable information about future returns.