Getting asset allocation right is far more important than selecting stocks or funds, as several studies have clearly demonstrated. But is Chris going too far in wanting us to throw out 50 years of insights provided by modern portfolio theory (MPT)?
The problem with MPT is that it generally relies excessively on past performance and recent past performance at that.
What’s more, its more uncritical followers tend to interpret statistical interpretations of the past as determinsitic predictors of the future. MPT is about measuring probabilities. And it is true that normal distribution curves are probably deficient but that does not mean that the whole approach of MPT is wrong – just that normal distribution curves underestimate the frequency of big catastrophes and we all forget their impact.
We have seen this sort of market turbulence before – not in the same way – in some ways worse and in others less threatening. Chris Gilchrist and myself were both sallow new-comers to the bear market phenomenon in the early 1970s. We have been here before.