NEW YORK (TheStreet) -- Warren Buffett's Berkshire Hathaway (BRK.A) underperformed the market for the past five years. While this garners headlines, what does it mean from a statistical perspective? Using BRK as a proxy for Buffett, how do we process such an underperformance? Perhaps Buffett should follow Chicago Bulls' Michael Jordan to the sidelines.
Probabilists can easily adjust their interpretation of Warren's performance, given what's happened recently. In this note, we looked at the statistical test for outperformance, and arbitrarily threw in Buffett's recent underperformance, as part of the study. BRK has a long audit trail that makes it easy to contrast with the market, using CFA performance standards.
As confirmed by the BRK annual review, the company has seen its fifth year of underperformance. This is also the worst continuous, 5-year performance ever. But it's not so easy to blame the S&P, as Warren does, since it has never been an issue to outperform this large benchmark previously. Given the underperformance event, however, we see a large number of curious fans trying to make sense of what this might mean.
To be clear, from a statistical performance perspective, this recent record for BRK is bad. There is no argument about it, or some way to excuse it. If it weren't known in advance that it was Warren's performance, we would outright laugh at it.
It also argues from a Bayesian perspective that his skill in the market has now dissipated. It seems Buffett himself is not convinced that his methods could outperform the market going forward and so his will is now explicit that his wife needs to be invested in the S&P and not the BRK successors. Nevertheless, his overall cumulative performance, including these past five years, is still a tremendous success on any measure, as stated in the note on my blog.