New York – If alpha is finite, what does that imply for research? Research has constraints on the value it provides, which is one of the reasons that the industry is becoming increasingly fragmented. The concept of informational alpha suggests that the research is more valuable the earlier it can provide insights and the less widely it is disseminated.
Hedge fund alpha
An academic study released in February examined the relationship between alpha and hedge fund assets under management, concluding that there are constraints on alpha both at the individual fund level and at the level of investment style. In “Crowded Chickens Farm Fewer Eggs” Oliver Weidenmuller and Marno Verbeek of the Rotterdam School of Management argue that the slow decline in hedge fund alpha earlier this decade is not just individual funds being overwhelmed by new assets (commonly associated with lower returns) but a finite amount of alpha being divided among an ever-growing field of competitors.
At the fund level, new entrants tend to generate more alpha than existing funds. Weidenmuller and Verbeek find that capital inflows have a negative effect on smaller funds. But when funds become large, the drag of excessive inflows is swamped by the fact that the fund is simply too large to produce great returns.
The paper also examined alpha among all funds within a given investment approach such as long/short equity, event driven, macro, emerging markets, etc. At this ‘strategy’ level, the paper finds that alpha is a zero-sum game. “It appears as any additional inflows in strategy segment decrease obtainable alpha returns, and that there is a maximum level of capital allocated to a strategy segment beyond which the performance of all funds suffers.”
Our colleague Tom Hutchinson introduced the concept of ‘informational alpha’ as a way of describing the research analogue of finite alpha. There are two primary dimensions to informational alpha: timeliness and scarcity. The earlier the information and the less widely known, the more valuable the information becomes. This sounds straightforward, but the reality is more complicated.
Forensic research adds value by identifying accounting legerdemain and interpreting whether this is an early warning signal. We recently organized a lunch hosted by UBS which showcased three forensic research providers which demonstrated some of the challenges of informational alpha. Forensic research is looking for obscure clues in the accounting footnotes which are easily overlooked (scarcity). But timing can be a problem because sometimes signals can be early. There are accounting anomalies, but they are not yet material. Forensic research then focuses on whether and when the anomalies might become material.
Distribution ≠ Alpha
Another dimension of scarcity is how widely the research is distributed. The more widely research is used, the less informational alpha it contains. This is the research equivalent to the academic finding that funds with more assets under management have a harder time delivering alpha. In the forensic research segment, there are firms such as Assay or Badger, which try to limit their client base in order to maintain a level of informational alpha. The challenge for research providers is whether they can be adequately paid for the scarcity value, to compensate for the self-limited client base.
The increasing popularity of primary research is another illustration of informational alpha. One of the appeals of expert networks is that it is bespoke — consultations are customized to each request. This enhances the scarcity value, unless everyone is asking the same questions of the same experts. Paradoxically, some clients have told us that for some of the largest expert networks, this has been the case. On the larger expert networks, clients tend to concentrate their attention on a few experts, diluting the informational alpha. Some clients go so far as to say that they use the large expert networks for ‘consensus’ information, and then hire boutique expert networks for their informational alpha requirements. If so, this helps explain why there has been a proliferation of expert networks — by our count there are over forty-five expert networks out there…
Which also explains why the alternative research marketplace continues to expand. As research firms become widely used, their informational alpha declines, creating the impetus for new, ‘nichier’ firms. In this way, informational alpha is one of the forces that keeps the long tail of research getting longer.