Prime brokers help funds of hedge funds identify hedge funds, which creates what the authors of a new paper call a PB bias. This means that portfolios are overweighted to the hedge funds serviced by the connected prime brokers. The scholars tested (and confirmed) the intuitive hypothesis that the bias gets greater as the cost of due diligence rises relative to capital.
The paper’s authors are George O. Aragon, Arizona State University; Ji-Woong Chung, of Korea University Department of Finance; and Byoung Uk Kang, of Hong Kong Polytechnic University, School of Accounting and Finance.
Quantifying the Bias
More than a decade ago (in 2007) Andrew K. Golden, president of the Princeton University Investment Company, testifying before the House of Representatives’ Committee on Financial Services, said: “We spend at least 400 person-hours in our due diligence process before investing in a hedge fund.” Even in monitoring existing relationships the cost of due diligence for a platform such as Golden’s endowment—likewise presumably for funds of funds—remains considerable. Golden said monitoring takes up 70 person hours per year per fund.
So let us quantify the resulting bias: Aragon et al find that a fund of funds’ weighting on its prime broker’s hedge funds is on average 39.14%. The market weight on the same hedge funds is 25.84%. This large difference is robust even when one accounts for alternative explanations of the overlap: geography or style focus for example. Indeed, the data indicates that when a hedge fund is headquartered far away, or when its style is outside of the fund of funds’ comfort zone, the “information frictions are more severe,” and that this drives the PB bias up.
“Not That That’s a Bad Thing.”
What about auditors? The authors of the new paper use auditors as a placebo. Is there any common auditor bias in the portfolios of funds of funds? It turns out there is not, which is what one would expect on the presumption that PB bias arises from the information advantages that a PB contact can provide. Auditors “are less likely to gain and share special insights about the investment ability of hedge funds.” This supports what the authors call the “information story”
A related but distinct point: the PB bias exhibited by a fund of funds varies positively with the number of hedge fund clients that the prime broker in question possesses but varies negatively with the number of other funds of funds that the prime broker possesses. The positive variation fits with the information story, because as the PB gets more clients it becomes better informed about the marketplace and its advice to its fund of funds client becomes more valuable. The negative variation (with the number of other funds of funds clients) also fits because as that number shrinks, the exclusivity makes its advice to the fewer funds of funds it does deal with that much more valuable.
“Bias” is so often in colloquial usage, a “bad” word. If Jones says, “Smith has a bias,” Smith is bound to say, “No I don’t!” as the statement certainly sounds like an accusation. In this case, though, the “bias” in “PB bias” isn’t an accusation at all, and this isn’t something that fund of funds investors ought to urge their managers to avoid or resist. Rather, PB bias positively predicts fund of funds performance. When these authors sorted funds of funds into quartile portfolios on the basis of their PB bias, they found that the highest portfolio outperforms the lowest by 2.89%. The highest portfolio performed the total of the other three portfolios by 2.77% per year.
A Final Thought: Information Frictions
This paper is part of a growing literature about information frictions in the search for informed asset managers. Eleven years ago, Brown, Fraser, and Liang found considerable economies of scale in the performance of funds of funds. Size matters, and in a good way. The authors suggested the reason might be that the larger funds are better able to shoulder the costs of proper due diligence, overcoming information friction. Aragon et al., agree with that, but add that the larger funds of funds are more valuable to their PBs, and so presumably are in a better position to draw valuable information, notwithstanding the friction, from those PBs.