This article was first posted November 1, 2012.
Daniel Kahneman is the 2002 Nobel Laureate in Economics for his work on prospect theory -- the study of decision-making between alternatives that involve risk, where the probabilities of the outcomes are known. In stark contrast to the 'rational man' of classical economic theory, prospect theory tells us the people often make sub-optimal economic decisions, displaying risk aversion when presented with the opportunity for a gain, and risk seeking behavior when threatened with a loss. Thinking, Fast and Slow is a layman's introduction to Kahneman's work.
Kahneman describes several insights that should prove useful in retirement planning and investing. Perhaps the most germane is the superiority of algorithms over expert opinion. For example, psychologist Paul Meehl's groundbreaking study on the validity of a clinical psychologist's expert judgment vs. a statistical prediction according to a rule rocked the profession. Meehl's self-described disturbing little book spawned over 200 follow-on studies of the topic. To this day, 60% of the studies show significantly better accuracy for an algorithm versus expert opinion.
I particularly liked Professor Kahneman's comments on the puzzle that "a major industry (i.e., Wall Street and mutual funds) appears to be built largely on the illusion of skill." He then recounts a humorous story of a seminar that he and fellow Nobel Laureate Richard Thaler held at a major investment firm. Kahneman asked for some data to prepare his presentation and was given a spreadsheet with the investment performance of 25 anonymous "wealth advisers" at the firm for 8 consecutive years. Calculating correlation coefficients for the matrix he found that average was 0.01 -- in other words zero. The performance of this sample of highly compensated financial advisers matched what you'd expect from a dice-rolling contest.
A Duke University study of Chief Financial Officers (CFOs) of large corporations yielded similar results. Asked to estimate the return of the S&P 500 over the next year, the correlation between the prediction and actual value for the group of CFOs was less than zero. (In other words, when they predicted a decline, the market actually went up slightly.)
Giving this hapless group a larger target to hit, the CFOs were asked to provide two values for the S&P 500; the value that they were 90% sure the market wouldn't exceed, and the value they were 90% sure the S&P 500 wouldn't fall below. The results: 67% of the time the actual value for the S&P 500 fell outside of the range given by our CFO experts.
Throughout the book, Kahneman provides similar anecdotes where experts' perceptions are at odds with statistical reality. It will certainly make you think twice before you pay for an expert opinion.
Thinking, Fast and Slow is a very entertaining book. It's worthwhile reading.
Retire Early rating: