Sunday, March 29, 2009

Fortune Tellers

I often don’t agree with columnists in the Sunday NYT Business section, but today three unrelated articles came together for me.   I thought that they contained some good investment lessons.  I’ll address the first two here, and leave the last for a subsequent article. 

I may tend to run on here, so let me give you the conclusions first and you can decide whether to read more or save time and go about your business. 

1)      Everyone tries to use correlations and statistical analysis to predict the future, but it’s a very difficult if not useless endeavor. 

2)      Analysts focus on the things that they can analyze, but many other factors are more important but unanalyzable. 

3)      Times of extreme stress bring big changes.  That’s the natural order of things, and while it can be painful, it’s usually necessary and good. (See next post). 


Mark Hulbert writes about a recent study  by two finance professors who challenge the conventional notion that long-term equity investments always produce superior returns Strategies - Now the Long Run Looks Riskier, Too, for Investors.  Essentially, they endorse the old saw that past performance is no guarantee of future performance.  Although we all nod our heads in agreement, most of what passes for equity investment research is based on the premise that past performance does indeed  predict future performance. 

The professors note that uncertainty increases with the holding period, and that the reason for the good historical equity record in the U.S. over the past century could be attributable to some non-investment related effects like having won two World Wars.  They conclude that “other things being equal… you (should) probably lower your portfolio allocation to stocks”.   I conclude:  other things are never equal.

Robert Schiller, another finance professor (wouldn’t all of our problems be solved if the world was run by finance professors?) addresses the field of social psychology in the investment world It Pays to Understand The Mind-Set.  Professor Schiller explains social psychology as “a theory of mind—defined by cognitive scientists as humans’ innate ability… to judge others’ changing thinking… It is a judgment faculty, quite different from our quantitative faculties.” 

Of course, forecasts based on a theory of mind are subject to egregious error. They cannot accurately predict the future. But the uncomfortable truth has to be that such forecasts need to be respected alongside econometric forecasts, which cannot reliably predict the future, either.” 

This is a very important point.  Stock investments are like icebergs:  only a small portion of the essential information is visible.  There’s a vast mass of unknown and unpredictable information, including the changing motivations of customers, competitors, and other investors, beneath the surface.  Yet rather than to acknowledge the huge inherent unknowns, analysts engage in silly exercises to micro analyze the small part that is visible and extend their conclusions to the whole.  For the most part, they all have access to the same data, often provided to them by the company or the government.  They produce surveys, and channel checks, and sophisticated models.  They issue buy and sell opinions in an effort to distinguish their particular outlook.  But in the end, they all have essentially the same understanding of the investment, and it’s a limited understanding. 

I’m not trying to pick on analysts (buy side or sell side).  My comments apply equally to most economists.  It’s not that they aren’t smart, or hard-working, or experienced.  But the nature of the job requires that they make predictions, and they almost never acknowledge the extremely high degree of randomness. 

So what’s the bottom line?  Be skeptical of anyone who offers investment predictions.  But be particularly skeptical of investment predictions that are based on extensive and rigorous statistical analysis, because they’re probably based on only an obvious and small part of the whole.

No comments:

Post a Comment