I have to predict (monthly) returns on Stock Indices (S&P 500) with the FF-Model (3 and 5 Factors). Therefore I shall use a predictive regression and an in-sample analysis. I started off with a (normal, not predictive) regression on the whole dataset (last 28 years). The adjusted R2 was about 50% for the 3 factor Model (maybe someone can tell me if this is still OK, it seems too small in my opinion). After that I tried the predictive regression

$y_t = a + b\ R_{t-1} + s\ \text{SML}_{t-1} + h\ \text{HML}_{t-1}$

[adj.R2: 30%]

but it seemed not quite right.

There are three options I can think of:

  1. I have done something wrong
  2. The model's predictive ability is that poor (but the explanation ability is not great either)
  3. My dataset is too big, but I'm not sure if it is OK to divide it into smaller pieces and analyze them individually.


  1. Can anyone validate my findings or tell me what I have done wrong?

  2. What would be the appropriate way to test the prediction ability of the model(s)?

  • $\begingroup$ is this school work or you're managing a pension fund? $\endgroup$ – Aksakal May 14 '18 at 15:27
  • $\begingroup$ I am an undergraduate business student, so it is school work $\endgroup$ – Husten May 14 '18 at 17:53
  • $\begingroup$ then mark it with self-study flag and read the instructions $\endgroup$ – Aksakal May 14 '18 at 17:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.