In an earlier post we addressed the question as to whether there is any bias in the delay between a fund's month end and the performance of that month. Our cynical theory is that managers with good news to report report it early and those with bad news to report report it late.

Since September, 2009, I've been sampling the reported monthly return of the BarclayHedge Hedge Fund Index. This is a simple average of the monthly returns of all funds that have reported to the group at that time. From time-to-time during each month I've sampled the main index's reported average monthly rate of return and the number of funds that have reported. You can find this data on my blog at the page Return Index Accumulation Report. The chart below is an analysis of the error, meaning the difference between the average monthly rate of return for the entire universe reporting on the sample date and that value finally reported at the end of the month. To gauge the average scale of the bias we fit a simple model by least squares:

Here B is the bias, or the average error between the sampled monthly rate of return and the final monthly rate of return; S is the scale we seek to estimate; and, p is the proportion of funds reported (i.e. the number in the sample divided by the final number of funds reporting in that month). Our estimate of the average scale of the error is (51 ± 8) bp/month.

It appears that Microsoft's file replication service, NTFRS, decided to delete the copies of my charts and regression results from both my web servers. This is frustrating, but I am in the process of reproducing them. Some of them may end up “more up-to-date” than the post text refers to — but apart from being contextually jarring that's no big deal. (This is presumably why they replaced NTFRS with a completely new product in R2 version of the operating system.)

As of this afternoon, a total of 1149 of approximately 2,500 hedge funds have reported data for October, 2009; so it's now time to update our data and forecasts for the dynamic trading risk factor. This gives us an early estimate of a number that is not much likely to change during the rest of the month.

Out-of-sample, our final forecast of the return for September, 2009, was 1.10% (our early forecast was 1.11%), and the realization was a substantial 3.25%. Our final forecast for October, 2009, was 1.61%, yet our current estimate of the performance for the most recent month is a loss of −0.20%. As a result we are forcasting a 0.30% return for November, 2009. These forecasts represent an a priori expected monthly return for any fund or firm that makes it's living by trading.

The Compact Model Portolio, which has been quiescent for a long time, has traded again. This time substituting FAS, which is the triple leveraged financials ETF for SDS, which is the double short S&P 500 tracking ETF. (The history of the composition of this index, and it's daily return relative to the benchmark, have are available on the blog side panel.) This trade actually occured on 20/10/2009, but I've been very busy on other projects and have been unable to update the blog for a while.

Astute observers of the Compact Model Portolio, which had been quiescent for a long time, will have noted that it has once more traded again. This time substituting EEM, which is the ETF following the MSCI Emerging Markets Index, for J.P. Morgan. (The history of the composition of this index, and it's daily return relative to the benchmark, have are available on the blog side panel.)