Recently I had a student post a message to me complaining about his grade on an assignment. His tone was insulting and clearly inappropriate. In addition to the fact that he turned in the assignment late, his derogatory tone did nothing to help his case. This was the first time in more than three years of teaching for the University of Phoenix that I have had a student display unacceptable behavior toward me personally. While this was certainly disappointing, the experience reminded me of the normal, very human way in which most of us make short-term, emotional decisions that often have detrimental consequences.
I have always been fascinated with the reasons people make decisions, especially when it relates to investments. The human factor inherent in the action of securities markets is a very real, very significant aspect that all investors must consider when formulating their investment strategies, and when making specific investment decisions.
I recently watched Moneyball again. For those who have not seen the movie, it tells the true story of Billy Beane, the Oakland A’s general manager who used statistics to pick his baseball players, and found surprising success against teams with much larger budgets After his success, the Boston Red Sox, using the same basic strategy, won the World Series two years later.
What is interesting about this, and the way it relates to investing, is that baseball was completely transformed by this experience, with many teams now incorporating some level of statistical analysis into their operations. This is not dissimilar to the way in which many investment funds now use sophisticated computer models to choose investments and time purchases and sales.
The problem with these models is that they can’t model human factors — why and how people make decisions. In fact, the more sophisticated and complicated the models get, the more likely they are to fail, at least in my experience. The way I describe these models is that they are two-dimensional in a three-dimensional world.
In a more general sense, even if you exclude human factors, computer models must incorporate assumptions to reduce the number of variables included in the model. It just isn’t possible to model every potential variable. The problem with making simplifying assumptions is that to simplify you have to exclude certain possibilities. This is exactly what happened when companies like Bear Stearns, AIG, Lehman Brothers and so many others made massive bets on credit default swaps. These firms basically made a bunch of assumptions about the likelihood of catastrophic events — they assumed (incorrectly) that they wouldn’t happen, but they did.
To me, the best approach whether it’s with investment models, baseball models or pretty much any other structured methodology for decision-making, is to incorporate some combination of statistical analysis and traditional analysis that includes the impact of human factors. For baseball, you have to consider things like age, injuries, attitude, confidence, family, behavior on and off the field, etc. For investing, all decisions must incorporate the impact of investor behavior — their attitudes and impressions of the current and future economic situation, their expectations for the relative attractiveness of the various asset classes, the impact on their decisions from tax policies and other political issues, and the like. Only by thinking through the collective impact of all of these factors on investor decision-making and then incorporating the most likely impact of those decisions on the financial markets, can we hope to formulate an effective investment strategy.
It is easy to understand investors’ attraction to fully automated investment strategies. The vast majority of investors make decisions, not based on their true, long-term investment objectives in an unemotional, logical, reasoned manner, but rather on the short-term, emotional forces of fear and greed. Taking people out of the process and relying 100 percent on a computer model to make decisions certainly sounds appealing! There have been some of these “quant” funds that have enjoyed some measure of success, at least over the shorter term. But, just as those using strictly statistical models for choosing baseball players will surely realize, ignoring the impact of human factors will eventually lead to poor performance. There is no arena where this is more valid than investing.
The answer may lie in combining some type of quantitative model with good old-fashioned research techniques that include an analysis of the impact of human factors. The study of behavioral finance specifically deals with analyzing, evaluating and quantifying the impact of human decision-making on financial markets and the economy. Investors considering purely quantitative strategies should do their homework and understand the simplifying assumptions that have been used in the model that drives the investment decisions before risking their capital. These models simply cannot function without making at least some simplifying assumptions, and it is these assumptions that limit the effectiveness of these strategies.
Caveat emptor (buyer beware).