Skip to main content
Message Font: Serif | Sans-Serif
No. of Recommendations: 2
When looking up the aforementioned paper in the Journal of Financial Research by Darrat et al I also found this one


W. Scott Bauman
Northern Illinois University

C. Mitchell Conover
University of Richmond

Don R. Cox
Appalachian State University


Previous research finds that large companies
previously judged to be excellent growth companies
have subsequently been poor investments. Our study
examines small companies selected by Business
Week on the basis of multiple criteria used in annual
articles featuring highly rated growth companies. We
study the investment performance over the three
years before eleven annual Business Week
publications and the three years after publication. We
find positive excess returns in the pre-publication
period, but negative excess returns in the
post-publication period. This reversal in investment
performance appears to be due to a mean reversion
tendency in operating performance, in which the
earnings and the past rates of return on capital of
such companies subsequently decrease significantly.

i found this remark particularly telling

"We find positive excess returns in the pre-publication period,"

(hence that is why they are chosen by the publishers, see the mea culpa written by the TSCM writer that i mentioned earlier)

"but negative excess returns in the post-publication period."

they explain this effect as reversion to the mean in terms of operating performance, but i suspect that there is also a mean reversion with respect to valuation going on as well.


Print the post Back To Top
No. of Recommendations: 4
they explain this effect as reversion to the mean in terms of operating performance, but i suspect that there is also a mean reversion with respect to valuation going on as well.


I think you're hinting at something really big here. This basic idea is the core of my investment approach (although a strict Graham disciple would gag at my use of "investment" here).

In a nutshell: "Fade the Noise."

Really important fundamental information about a business probably doesn't come out more than 6 times a year (4 quarterlies plus a couple important developments or press releases). Add in a few more important events per year that are industry specific and you ought to get a share price that moves like one of your kid's connect the dot drawings.

Underlying these changes in business fundamentals you have economic factors that Buffett would say to ignore because you can't predict them anyway. But you can at least react to them in hindsight and alter your necessary discount rate based on what's happening in the bond market. So you get some additional "volatility" that's actually structural--not business related, but economy related. And this information is imperfect as well, so you get some justifiable "noise" as investors try to anticipate inflation rates and interest rates and what-not. And during major rate-changes like we've just gone through, events like this can alter valuations extensively.

But even so, superimposed on top of this company- and economy-driven process variation, there has to be one hell of a lot of absolute noise. Of course viewpoint/perception is everything, and to a day-trader looking at intraday charts, there is discernable structure to the noise and so they'll try to exploit it. But it's not fundamentally driven structure, so it's impossible to predict, as far as I'm concerned.

When I back up and look at a 1-year chart, it's mostly just noise to me. And so I'd rather just fade the intra-day intra-week variability--buy it when it sags and sell it (or sell it short) when it peaks. And a week or a month or 3 months later when it's bounced to the other side of it's trading distribution, close the position.

I try to capture as much of the noise as possible using limit-orders on both the buy and sell side. This is what I hate about stop-losses; you let the market-maker keep the profit from noise. You can make a very respectable annual return flipping stocks mid-term for 5-15% unit gains.

Of course if you're buying and selling noise, you have to make sure that you don't get absolutely trampled by massive fundamental changes that would trample you. So you can't just "Fade the Noise", you've got to have fundamentals behind you on your long positions (or lack of fundamentals on your short positions). Given that condition, you can ride out an unfounded trend that moves way against you (i.e., trades that move against you turn into "investments", a strict no-no according to traders, but I disagree.

Berkshire is of course a great company to invest in for the long term, but I've been more than content to capture 41% and 20% mid-term moves in it and flip it (and had those bouts of massive pessimism not been followed within 3-6 months by over-optimism, I'd still be holding it). Likewise, Nucor was a great company to buy in the low $30's, but at $65 I think it's riding too high on noise and I'm short.

I haven't really developed this as well as I'd like to, and I'm out of time. Pursue if you'd like to....

Print the post Back To Top