Testing Conditional Factor Models
Columbia Business School - Finance and Economics; National Bureau of Economic Research (NBER)
University College London; University of Aarhus - CREATES; Cemmap (Centre for Microdata Methods and Practice)
January 21, 2011
Netspar Discussion Paper No. 01/2011-030
We develop a methodology for estimating time-varying alphas and factor loadings based on nonparametric techniques. We test whether conditional alphas and long-run alphas, which are averages of conditional alphas, are equal to zero and derive test statistics for the constancy of factor loadings. The tests can be performed for a single asset or jointly across portfolios. The traditional Gibbons, Ross, and Shanken (1989) test arises as a special case of no time variation in the alphas and factor loadings and homoskedasticity. As applications of the methodology, we estimate conditional CAPM and multifactor models on book-to-market and momentum decile portfolios. We reject the null that long-run alphas are equal to zero even though there is substantial variation in the conditional factor loadings of these portfolios.
Number of Pages in PDF File: 63
Date posted: April 19, 2011