Complex Models and Biases - Dreman and the Contrarian
Reynolds,
You've got to read the Dreman book on contrarian investing. You could have written it. That's how similar his thinking is to yours. You will love it.
I have a contribution to make on the complex information overload and bias issue that we have discussed on numerous occasions. That's probably the main thing that separates our current view on the market and has lead to numerous heated discussions. We are often looking at different information and models and therefore coming to different conclusions. Mr. Dreman discusses this subject at length.
One example Mr. Dreman gives on this subject involves horse racing. Basically he postulates that studies indicate that the more information a supposed racing expert has, the more confidence he has in his ability, yet there is very little improvement in the final results. This may be partly due to bias introduction and partly due to the inability to process the additional complex information properly. Similar studies in other areas indicate the same thing. He therefore advocates simplicity. This would seem to argue for the point of view you have expressed to me repeatedly. It is one that I have questioned. Perhaps, surprisingly to you, I happen to agree with this observation as it pertains to horse racing to some degree. I also believe that I have bridged our gap on this issue as a result of the comparison to horse racing. First a little background.
Before I got into investing 12 years ago, I was a serious horse racing handicapper. I would have to say that I know more about valuing odds on horses than I do about valuing businesses. I have been at it for 24 years. During that time I became acquainted with some of the best racing handicappers in the country. Several of them make their living exclusively by betting on horses! This is no small feat since the mathematical expectation is around negative 15%-18%. (that is the track take) In order to win big you must find inefficiencies in the MARKET that are much greater than that. Some other players show consistent profits (though not enough to live on.) I am in the latter category as a result of significant personal effort and the help of some of those acquaintances.
Here is the crux of my observation. About 90% of the information needed to make money at the racetrack is contained in the Daily Racing Form and is very simple. The 10% that is missing requires some extra calculations, visual skills, film watching, statistics keeping, and research. It is the 10% that is missing that puts you over the top from a profitability point of view. I don't know a single winning player that uses just the 90%. All do some if not most of the extra research and information gathering. You can cut the house edge without it (outperform the market), but not enough to win. The missing 10% does not improve your overall win percentage very much though. In fact, the improvement in WIN percentage is hardly noticeable. Most of the ability to select winners is in the simple and obvious information. One factor alone will produce results very similar (from a win percentage point of view) to the best expert. What the extra 10% does, is enable you to separate the good value from the poor value among the contenders of similar ability and similar probability of winning. That is because most of the players do not have the information or do not understand how to use it. Interestingly, I have also known numerous players who DID have the extra 10% and did not improve their results at all. It is even possible that they may have gotten poorer results. They simply could not value and weigh the additional information well enough for it to be useful. I was in that category in the early period of my skill development. It takes time and practical experience to use the information properly.
This leads to my updated conclusion on the subject of complex additional information and bias. It's one that I think you will be comfortable with also. "Additional information of complexity is valuable in some hands and not in others." Applying this to investing, one might say that supplementing the Value Line (or other basic data) with adjustments for stock option compensation, post retirement benefits, one time charges, point of business cycle considerations, and other adjustments etc... will add to the returns of some investors and hurt others. The same can be said for some of the more complex valuation models. The net, as Mr. Dreman might argue, may be of little "aggregate benefit" to investors.
From my own perspective, this type of additional information (even if a bit complex - without going overboard) as it pertains to both horse racing and investing has been a big plus. That's why I will continue argue for it even if it is not appropriate for some people. I think it's a matter of reaching a competency level at using it so that it becomes value enhancing instead of confusing and distorting. There are higher returns to be achieved from getting to that stage in both horse racing and investing. Just maybe not for everybody.
Here is a specific example. My girlfriend and her boss are short sellers. They have been very successful at identifying companies in certain industries that are about to have serious problems. These problems would be virtually impossible to find using basic data. They do it through a careful analysis of the balance sheet receivables, inventories, and other significant numbers. In my hands, this information is practically useless. It is only useful if you have a high level of understanding of those types of businesses. I do not have it at this time. In their hands, it is value enhancing and provides better results. The intrinsic value of this complex information and analysis is almost without doubt based on my observations of their performance. The net of everyone trying to use it may be zero. The NET may not be what counts and determines whether you should use it though! It is your own personal level of competence in the area and with the data. So while the Dreman advice may be good when talking to the broadest possible audience, it is not really correct in my view. It is essentially advocating that investors choose simplicity and somewhat above average results when many could do much better by applying more complex and additional data to their study. That is certainly the case at the track. Have a nice weekend. I'm off to Saratoga Springs for the weekend!
|
