Analyzing the Fama French Five Factor and AQR six factor models

In this assignment you will be testing factor models on different sets of test assets. You will use the Gibbons, Ross and Shanken test statistic to examine if alphas are jointly equal to zero.

1.1. Initial Reading:

  • Read Fama and French’s “A Five-Factor Asset Pricing Model” paper (Link)
  • Read AQR’s (Cliff Asness’) post on the five-factor model “Our Model Goes to Six and Saves Value From Redundancy Along the Way.” (Link)
  • Read Karl Diether’s slides on the Gibbons, Ross and Shanken test (Link). You can also read the Gibbons, Ross and Shanken (1989) paper if you are interested (Link)

1.2. Gather Factor Data:

  • Download the Fama/French 5 Factors (monthly), the Momentum Factor, the 25 Portfolios Formed on Size and Book-to-Market and the 10 Industry Portfolios from Ken French’s website (Link)
  • Download the HML-DEV (monthly) from AQR’s website (Link)

1.3. Analysis:

  • Replicate Tables 1-4 in AQR’s blog post
  • Using the 25 portfolios formed on size and book-to-market as the test assets:
    • Examine the ability of the the Fama and French Five Factors to explain returnson the test assets.
      • Calculate the five-factor alpha for each of the test assets and report thealphas and the t-statistics
      • Test if the alphas are jointly equal to zero using the GRS test.
    • Examine the ability of the AQR six-factor model (Add the momentum factorand replace the Fama-French HML factor with the AQR HML-DEV factor) to explain returns on the test assets.
      • Calculate the six-factor alpha for each of the test assets and report thealphas and the t-statistics
      • Test if the alphas are jointly equal to zero using the GRS test.
    • Repeat all of 1.3.2 using the 10 industry portfolios as the test assets