Comments on: How P/E juices P/B returns

By: MJ

MJ — Tue, 20 Apr 2010 20:51:15 +0000

The E/P CXO cites is actually LTE (long-term earnings) defined as “1 – (B/P + STE)” where STE is defined as “short-term residual earnings, with the required return set equal to the risk-free rate and that forward residual earnings then converted to a no-growth residual earnings forecast by capitalizing one-year ahead residual earnings at the risk-free rate, as indicated by component 2 of equation (8) [page 19] in the text. It is then divided by price.”

In other words, to match the returns shown in the research paper, there are some required adjustments. It is not simply a matter of running two screens. To produce STE, the authors use consensus earnings estimates as well as their own models.

By: MJ

MJ — Tue, 20 Apr 2010 19:48:36 +0000

“The incremental return for B/P is most striking in low E/P quintile.”

Shouldn’t this read “high” E/P quintile? I was under the impression that the 5th quintile represents the lowest P/E, lowest P/B stocks.

By: Nate

Nate — Mon, 01 Mar 2010 17:22:28 +0000

Seems like a lot of the recent things you’ve been discussing are basically a rehash of the Fama/French work. My guess is the conclusion will be similar which is buy a value index weighted by p/e or p/b.

Have you considered comparing any of these results with the Russell value index or something similar?