# How would you in practice use factor models (like Fama French) to make decisions?

So I understand that the Fama-French factor model relates a stock's excess return with its beta, market cap, and book to price. How does one use the model in practice? Do people assume that market cap and book to price are fixed (use today's values), and then just supply an estimate for next period broad market return and basically predict future stock return using expected broad market return with some adjustment for market cap and book to price?

Would one supply estimates for future market cap and book to price as well? This seems to be difficult as market cap and book to price are all dependent on price, which will be dependent on return, which is your response.

Not exactly.

People use those type of models (such as the fama-french model) to evaluate their portfolio.

Literally, you run a regression of a stock/portfolio agains the FF factor model to understand if your portfolio beats known risk factors (i.e. whether its $$\alpha$$ is positive).

If the $$\alpha$$ is negative you are better off taking the money out of your portfolio and investing in a passive index that replicates those risk factors.

Another way of looking at is: let's assume you are holding the factors and you want to know whether you should add a specific stock to your portfolio. Then you run that stock against the factors. If the $$\alpha$$ is positive you should add it. If not you shouldn't.

• Ok thanks! So the model is not for prediction/forecasting at all? In your second example of determining if a stock should be added...what's given? You can determine SMB, HML, just looking at financials, excess return by looking at historical returns. What about the coefficients and the response (expected return) - are they given as well? And then you solve for alpha? Jun 27, 2020 at 21:46
• Yes. Not for prediction at all (in terms of time series forecasting). Just run a regression of the return of the stock into the factor returns (you can find them on Kenneth French data library). The regression immediately gives you the loadings on the factors and the alpha. Jun 27, 2020 at 21:48
• An Active Investor doing research on proposed investment strategies can use FF to determine if a strategy has an $\alpha$. A Passive Investor can use the FF model to construct a portfolio which replicates a desired FF factor exposure with low costs. Jun 27, 2020 at 21:49
• Thanks for the responses, I'll have to look into it more. Jun 27, 2020 at 21:52