This paper proposes a simple method to measure and compare the average relative return contribution of proposed risk factors. The method is applied to six common risk factors, including market, size, value, momentum, profitability, and investment, using 49 U.S. industry portfolios in the period 1969 to 2014. We find that the average relative return contributions of the market factor and mispricing alpha are highest in all models and sample periods. If multifactors are included, their main effect is to reduce the contribution of the average market factor return with some reduction in the contribution of mispricing alpha also.
Objectives:
The growing list of multifactors naturally raises a number of questions. What are the relationships between different factors? How can the relative importance of different factors be evaluated? Regarding the latter, it is possible that their joint characteristics are more important than individual separate effects.
This paper proposes a new measure of the contribution of individual risk factors to the required rate of return for a set of portfolios. To assess the average relative impact of each individual risk factor on the portfolio’s required rate of return, we develop a novel metric dubbed the average relative absolute return contribution ARARC. This goodness-of-fit measure summarizes in percentage terms the average economic contribution of each risk factor in a sample of priced test assets. Based on returns of 49 U.S. industry portfolios in the period July 1969 to December 2014, our empirical results show that major return contributions come from the market risk factor as well as the mispricing component in the traditional market model. By comparison, the contributions of multifactors, such as size, value, momentum, profitability, and investment, are relatively smaller. Moreover, the average absolute return contribution of multifactors is partly overlapping with the market risk factor.
Data:
We utilize exogenous industry portfolios to examine the relative return contribution of risk factors. Monthly returns for 49 U.S. industry portfolios are sampled over the time period July 1969 to December 2014. The analysis is also carried out over three subperiods: July 1969 to December 1984, January 1985 to December 1999, and January 2000 to December 2014. We use the traditional market model with the market excess return as the single risk factor. The following multifactor models are investigated: Fama-French 3-factor model, Carhart 4-factor model, and Fama-French 5-factor model.