Sharpe takes a look at the factor models in his discussions including the 3 factor model of Fama & French. Two interesting points were that the magnitude of the effect of small and value was entirely within the realm of randomness and that if any such effect existed it seems to have disappeared about the time that DFA started to offer a fund that really targeted that factor (arbitraged away?).
.
Hyper,
I took the time to wade through two of Sharpe's papers looking for this reference, (It is on pages 46-51 of the second lecture) and would not want the average reader to walk away thinking that Fama French has been made irrelevant.
What my (highly opinionated) read of Sharpe on this point leaves me with is that:
a) He wishes it were easier to dismiss the 3-factor model because it doesn't fit into his neat world of hypothetical fish famers
b) he eyeballs "Ocular Analysis" some graphs and decides that something in recent years isn't going up as fast as it used to, thus wishing to dismiss it.
c) He does find significant results in the small and small value categories, but dismisses these as 'only a few percent of the total market cap". That may be a problem if you only global-market-cap-weight as he recommends, but it is no problem at all if you slice with small and value tilts -- overweighting these small outperforming sectors has done nice things to my portfolio actual and expected returns.
d) he definitely doesn't say there are no longer value or small effects. After admitting that there are definite small and value effects, he simply asks, in that peevish academic way, whether they can possibly persist, since they are irrational, and that now that we all know about them, they are all but sure to disappear.
It reminds me of the joke about the two economists who find $100 lying on the ground in the park, and can't decide whether to split it, pick it up or whatnot, until one comes to the conculsion that, since $100 bills are valuable, if it really were one it wouldn't still be lying on the ground in a public place , so they happily leave it and walk off having assured themselves that what they saw couldn't possibly have been a $100 bill. (I was always lousy at telling jokes, sorry.)
While giving a nod to psychological "Market Behavior' issues, Sharpe doesn't touch what I view as the real reason a value premium persists (Fama and French admit the small premium may be declining, although if you rope in microcap and private equity, small cap premia may in fact be alive and well). The reason a value premium persists, imho is purely psychological: value companies generally have problems. They are in the news for negative surprises. (or they are boring un-noticed companies). When bad news hits, active investors, except those curmudgeonly value investors, are afraid their reputations will suffer if they own these things. Who wants to risk looking like a fool? So investors dump them and make them into bargains. Meanwhile management of these companies works hard to fix the problems (we are talking big companies here -- NYSE listed securities for most US value funds) and when they do, everyone acts surprised and gives them back their keys to the executive washroom (invites them back into the club of market-PE stocks). The value investor then has a fully priced security, which she sells and moves on. This works across the asset class, not on anecdotal stock-by-stock basis: I wouldn't want to hang all my chips on Delta or Enron, but a fund will balance these out with the Tycos and can come out ahead.
One point: value index funds may not capture this cycle that efficiently as their indexes only get re-set annually (usually) and as should be clear from the above, companies move in and out of the value camp possibly more rapidly than that. For this reason, I am comfortable paying for an active manager on the value side -- the only place this old BogleHead can really say that with any conviction.
I've posted on this topic before, so sorry if it is boring to some, but I feel value investing deserves an advocate and a fair hearing, especially when Nobel Prize Winners like Sharpe are lined up to diss it.
ESRBob
Additional note: I sent a polite version of this to Professor Sharpe, basically saying if he has real data showing that the value premium is gone, that this Board would very much like to see it. Will post any reply.