0
$\begingroup$

I have 2 questions which i can't seem to find no matter how I search. so:

1) If we have 2 portfolios. One based on risk-return tradeoff (with variables HML, SMB and beta ) (Fama French, 1993) and the other one based on the characteristics approach (Daniel & Titman, 1997) (still with market beta, size and book to market as variables) how to we compare which is better apart from the Sharpe ratio? With the t-stat of their coefficients?

2) Can someone please explain the difference between the factor and characteristics as variables for the returns of stocks?

Thanks

PS sorry if i am not clear enough but even I can't grasp exactly the idea

$\endgroup$
1
$\begingroup$

To Fama & French a "value stock" is a stock that fluctuates up and down together with a portfolio of (other) stocks that have high book to market. If you think about the 2 stage procedure Fama and French use to first identify the B/M factor and then calculate a stock's exposure to this factor, that is what it amounts to.

To K. Daniel & Titman a "value stock" is a stock that has high B/M. Period. There is no reference to any time series factor.

In principle, but it is very hard to do in practice, if we could identify a group of stocks that have high B/M themselves but do not go up and down with other high B/M stocks, but maybe go up when B/M stocks go down and vice versa, then we would have the solution at hand. We could just ask if these stocks have high returns over the long run (in which case K&D are right), or if they have low returns (then F&F are right).

But as I said, it is difficult to do this, and maybe impossible. (So 20 years later we still don't have an answer to this question).

$\endgroup$
  • 1
    $\begingroup$ Well firstly, thank you for your quick answer. I think I "kind of" grasp the idea of theory but not so much the practical aspect. I have been given 2 portfolios. One treats the the variables as risk (?) and the other one as characteristics(?). I have coefficients for all of them, t-stat (1 model only HML is significant, 2 model none), R squared and sharpe. I still though cant get if I just compare the stats or if there is something more in theory. Thanks and sorry for my messy line of thoughts. $\endgroup$ – John Est. Oct 10 '17 at 23:56
  • $\begingroup$ There was a lengthy article on AlphaArchitect about factors vs characteristics. My takeaway is that no one has a good answer so far as to which is best or how to compare them alphaarchitect.com/2017/10/31/… $\endgroup$ – Alex C Nov 11 '17 at 16:24

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.