A study by Vikas Agarwal, Naveen D. Daniel and Narayan Y. Naik, Why is Santa so kind to hedge funds? The December return puzzle! has prompted many observers to conclude that hedge funds mis-report their December performance. The report finds that hedge fund returns spike in December with an average return of 2.51% in that month alone, versus an average monthly return of 0.96% for the January-November period (see adjacent chart from the study). This spike can not be explained by higher market returns in December and therefore the authors conclude that it must be caused by return management. To back up their claim of large scale wrongdoing, they also show that funds with higher performance fees, more illiquid strategies and higher volatility exhibit higher December outperformance than others. In addition, they claim that funds borrow January performance to boost December returns.
Not so fast. For a start, we have doubts whenever someone tries to model returns of non-linear investment strategies with linear market factor analysis. (As an aside, many of the hedge fund replication studies suffer from that fallacy, too.) More importantly, we can think of a few fundamental reasons for the December spike that are not captured by the professors’ study, or by any other Fama-French factors.
The first one pops up in our own merger arbitrage portfolio every December. That is the month when more mergers close than in any other month. To back up our empirical observation with hard data, we looked at the closing month of all cash mergers of public companies in the MergerStat database from 1990 through 2006. We restricted the analysis to the 2,659 cash mergers. It can be seen from the adjacent chart that the month of December experiences the most closings of any month over that 17-year period.
We will readily accept that the closing of a merger in a given month does not necessarily translate into a higher return in that month. However, our own experience tells us that very often, a rapid spread compression occurs near the closing when any remaining obstacles have been removed. Not being full time quants, we leave the careful analysis of this distinction to someone with more free time on their hands.
Moreover, the closing of a merger leads to a cash inflow that must be invested. December is not exactly known as a month with lots of merger announcements, so managers have to wait until January to put the cash to work. In the meantime, it acts as a drag on returns, and makes the statistics look as if a fund has borrowed performance from January to boost December returns.
Another factor that has the potential to explain some of the December performance spike is the seasonality of implied option volatility. Many hedge funds use option writing in one form or another in their investment strategy. As Bill Luby points out in his VIX And More blog, the VIX index of implied volatility peaks in July and October and subsequently drops through November and December. A seller of volatility will make above average returns in the runup to the option expiration on the third Friday of December. [Credits for the chart and all the underlying work go to Bill Luby]
A final reason that contributes to higher December returns is the ability of hedge funds to take advantage of various year-end effects, such as tax-related selling or the squaring of banks’ books ahead of 12/31, that are not reflected in indices and that other investors may not take advantage of. We have seen such effects ourselves in our past life in fixed income, although we understand that the particular effect that we have in mind has since disappeared. Nevertheless, while tax codes as well as bank and insurance regulations become more and more complex, we have no doubt that savvy arbs will continue to find new opportunities at year-end.
It would be an oversimplification to assume that only funds listing option or merger arbitrage as their principal strategy can justify their December spike. Style drift is a significant problem in hedge fund returns, so that a clear attribution of the above style factors to specific strategies is futile. Recall that after the demise of Long Term Capital Management, it became known that this fixed income fund had taken large positions in merger arbitrage. Similarly, convertible bond fund Amaranth went belly up over natural gas trades.
It is clear that alternative strategies depend on many different nonlinear factors that can not be explained or replicated easily by market indices, and that are difficult to capture in a traditional mean-variance analysis. We do not want to deny the possibility of ill faith in some of the reported return figures, but we do believe that it takes a leap of faith to use traditional factor analysis and conclude that hedge fund returns have been tampered with. Surprisingly enough, two of the authors of the December spike study have previously published a paper that emphasized the option-like return characteristics of hedge fund strategies, so we would have expected them to take these characteristics into account, if not quantitatively, then at least qualitatively in the interpretation of the results. We hope that some PhD candidate will give us a more complete and credible explanation of the December performance spike by modeling factors more relevant to non-linear strategies.