In an earlier post we asked Is Berkshire Hathway a Hedge Fund?. Since that work was done, in mid. February, Berkshire has rebounded sharply off its low. Hedge funds have also been making money again, so I decided to revisit the data and see how Berkshires regression significance has changed. Has the recent months reinforced or weakened our prior conclusion that there was a significant element of the monthly returns that can be simply attributed to the risk premium arising from dynamic trading?

The prior regression showed that, over almost the entire previous decade, the monthly returns of Berkshire Hathaway common stock have a β of 0.70 ± 0.24 onto the dynamic trading risk factor, with a significance level (p-Value) of 0.005. The α is positive, but not significant, at 0.10 ± 0.46. The R² was 8%, which corresponds to a correlation coefficient of 28%.

In this regression we find a correlation of the monthly returns (to shareholders) with the dynamic trading risk factor, which should correspond to the returns of a typical hedge fund, has strengthened slightly. The β is now 0.83 ± 0.26, which is statistically indistinct from unity (t-Statistic is 0.63 referenced to the β = 1 hypothesis). The t-Statistic for the β is 3.15 for the null hypothsis of β = 0, which is equivalent to a p-Value of 0.002. The α is now estimated at (−0.01 ±0.50) % per month, which is statistically indistict from zero. The R² for the OLS regression (in the left hand panel of the chart) is 9%, which corresponds to a correlation coefficient of 31%.

In the prior article we chose not to draw a bold conclusion from the data, but I think the conclusion has to be stated more firmly. The data suggests that Berkshire Hathaway, Inc., as far as the returns available to the ordinary shareholders are concerned, is making money solely by following hedge fund trading strategies. One caveat to this, as compared to our regressions for Goldman Sachs and Morgan Stanley, is that the R² is much smaller. However, this could be attributed to skill. i.e. We could conclude that Morgan and Goldman are just better at executing hedge fund style trading strategies, with less zero expectation noise trading getting in the way, and this is why their returns are better explained by the factor (remember that the factor has a non-negative mean).