Paul Meehl, who died in 2003, was a brilliant psychologist who wrote what is probably the most significant work on decision-making published in the last century. His short treatise, Clinical Versus Statistical Prediction, is still read and referenced today more than fifty years since it was written in 1954.

In what he would later call "my disturbing little book," Meehl advanced the theory that experts would make more reliable predictions if they used their expertise to construct models to make the decisions for them. Meehl felt that experts, because they were human, were inconsistent in the application of their expertise when it came to making decisions based on judgment. By contrast, a mechanical model would make the same decision for the same set of data each time and thus be more reliable.

During his lifetime, almost two hundred studies validated his theory that a defined set of rules - what we call naked strategies - would make more accurate predictions than the judgment of experts.

So why has Wall Street continued to ignore this important work? Money is one reason. After all, if a portfolio manager who is earning $5,000,000 per year can be replaced with a model, what justification is there for high management fees? Ego may be the second reason. What portfolio manager wants to believe he or she can be replaced by a relatively simple rules-based model?

If Meehl was right in his theory, then investors who use judgment-based portfolio managers may be paying a premium for mediocrity.

Don't expect Wall Street to change course any time soon though. As investing sage John Bogle says, if you pay a man a tremendous amount of money not to understand something, it is amazing how difficult it will be for him to learn it.

Labels: Bogle, Meehl