Saturday, June 12, 2010

Contrarian Investment Strategies - The Next Generation: Chapter 4

The founder and Chairman of Dreman Value Management (est. 1977) shares his views on how investors can beat the market with this book (written in 1998). In reference to the efficient market hypothesis (EMH), Dreman writes "Nobody beats the market, they say. Except for those of us who do." More on this book is available here. One of his earlier books (from 1982) has already been summarized here.

Dreman quotes many "experts" from military to political to corporate leaders that have made famous predictions that went terribly wrong. Experts in the stock market are no different. While people gather en masse to hear the opinion of experts, Dreman argues that this actually prevents investors from achieving superior returns. Like a bad golf swing that must be unlearned, investors must unlearn the way they currently use "experts" if they have any hope of earning superior returns.

The psychological research explaining the apparent inability of experts to forecast the future is discussed. The way humans best handle problems is in a linear fashion (e.g. Step 1 do this, Step 2 do that etc.). But more complex problems require interactive reasoning, whereby the interpretation of one input can change depending on how the other inputs are evaluated. Research suggests that humans tend to apply linear approaches to solving problems that are optimally solved interactively.

The more complex the problem (i.e. as more inputs need to be evaluated), the more this "line" of thinking can lead to poor results. But exacerbating the problem is that the more inputs that are provided, the more confident the decision-maker becomes. Studies testing respondent predictions for events that are highly uncertain show that when a decision-maker is provided with more information, his ability to predict stays flat (or rises moderately at best) but his confidence increases with every piece of new information, which can be a dangerous combination.

Dreman argues that stock research requires similar interactive thinking, and that the problems that must be solved (predicting a future stock price etc.) are complex, requiring the simultaneous evaluation of scores of data points. Despite the availability of unimaginable amounts of information, the outcome is nevertheless very difficult to predict, leading to low predictability, but overconfidence on the part of experts.

No comments: