One possible reason why investable hedge fund indexes underperform

Over the years we’ve often marveled at the fact that investable hedge fund indexes tend to produce returns that are around 2% lower than their non-investable cousins over time.  Theories about as to why this is true.  In November, we even wondered if the demand for monthly (or greater) liquidity from the hedge fund underlying investable indexes was a cause.  The chart below from that post clearly shows that the HFRI (an example of a typical non-investable hedge fund index) beat the HFRX (a well-known investable index designed to roughly track the HFRI) during the financial crisis.

So is the performance of the investable indexes understated or is the performance of non-investable indexes overstated?  A new paper from Edhec-Risk Institute says the non-investable indexes are overstated – and proposes a “remedy” for this inaccuracy.

Edhec finds that the cumulative out-performance of non-investable indexes is mainly a result of the worst months for the hedge fund industry.  In other words, during periods of relative calm in Hedgistan, the difference between investable and non-investable indexes is relatively small.   The chart below was compiled with data from Table 1 in the paper.  It shows the out-performance of the (investable) HFRI over the (non-investable) HFRX.

As you can see, the delta between the HFRI and HFRX was around 4% per annum between December 2005 and December 2007.  But then it spiked to over 5% for Global Macro, over 15% for distressed securities and a whopping 25% for convertible arb.  Conversely, the delta between the HFRI and HFRX actually fell during the financial crisis for long/short equity, event driven, and equity market neutral.  As Edhec points out:

“…there was a striking contrast between liquid and illiquid strategies. For the latter, the significant increase in the excess returns of the non-investable indices during the second period perfectly coincided with the global credit crunch.”

When Edhec compared a portfolio of investable hedge fund indexes with its own “index of (non-investable) indexes”, it found that in extreme down months for the hedge fund industry, the delta between investable indexes and non-investable ones spiked dramatically (and non-linearly).  The picked the excess returns of its index of funds of funds indexes to illustrate the point:

They used the results of this model to adjust for the performance of the non-investable funds of funds index.  As you might guess, by applying this model to the data in the sample, they came up with a “readjusted” non-investable index that tracked the investable indexes pretty well (chart below to left of the vertical line).  But what’s really impressive is how the model worked outside of the time period from which it was born – the area to the right of the vertical black line showing February 2009 to April 2010. (Blue line = regular Edhec non-investable index, Orange line =investable indexes, Green line = “readjusted” Edhec index).

Note that the performance of the investable indexes (orange line) was about the same as the non-investable and non-adjusted index after February 2009.  In other words, investable indexes suffered most in Q4 2008 but held their own admirably before and after the crisis.

So it appears that one of the reasons behind the under-performance of investable indexes is indeed liquidity-related in that the investable versions of the most illiquid strategy indexes suffered the most while the investable indexes of the more liquid strategies actually performed a relatively little better during the crisis.  The authors caution  that the time period used in the analysis is “too short” to make the results “robust enough.”  But we think the results are intriguing nonetheless.

Be Sociable, Share!

One Comment

  1. Sol Waksman
    January 21, 2011 at 11:56 am

    Our firm, BarclayHedge, develops and maintains hedge fund and CTA indices. In particular, we maintain the Barclay CTA index and the Barclay BTOP50 Index. The BTOP50 is an investible subset of the Barclay CTA Index. In the 11 year period from Jan. 2000 through December 2010, the compound annual return of the investible BTOP50 exceeded the that of the non-investible Barclay CTA Index by 40 basis points. Perhaps one should take a closer look at index construction methodologies and the manager selection process in order to gain a better understanding of the reasons for the performance differentials discussed in the article above.

    Sol Waksman


Leave A Reply

← Researchers try to fix "unreliable" hedge fund measures Study finds private equity "four-peats" can be more difficult than previously thought →