Tuesday, March 29, 2011

Paper: "Perils of the Periphery: Social Network Position and Hedge Fund Performance"

I just read an interesting paper called Perils of the Periphery: Social Network Position and Hedge Fund Performance. The author, Joon Nak Choi, looks at how the position of hedge funds in a social network (with network ties consisting of fund managers and employees who have worked together previously) affects results.

Ordinarily, sociologists talk about social networks in terms of a core-periphery structure, having a core that consists of prestigious, highly-connected nodes, and a periphery with nodes that have fewer connections.

Researchers already know that highly-connected funds in the core of the industry perform better than peripheral funds with fewer connections. A common hypothesis is that peripheral funds are likely to be late adopters of ideas, because innovations must be first be legitimized by being adopted by funds in the core. The result is that,

"As early adopters, the core is likely to carefully assess the technical merits of the innovations they adopt. As late adopters, the periphery may not focus on innovation’s technical benefits, as much as their legitimizing effects. Given this tendency, late adopters [i.e. the peripheral funds] may be particularly vulnerable to adopting practices and innovations that reward early adopters, but punish late adopters."
The reason that legitimizing effects are so important, and the pattern of johnny-come-lately investing continues, is that the younger managers (who are running peripheral funds with fewer connections) "seek safety and legitimacy by copying prominent managers." One of the results of this is herding behavior, or as we call it on Credit Bubble Stocks: groupthink. And the investors who are late to the party get left holding the bag. 

This paper extends the core-periphery analysis of hedge fund performance to include isolates: nodes that are "disconnected from the main component (i.e. main group of connected actors) in the core-periphery structure".

The paper asks, "Who performs better, connected outliers (i.e. periphery) or disconnected outliers (i.e. isolates)?" The conclusion of this paper is that,
"While first-movers in the core should fully benefit from inflating asset bubbles, their imitators in the periphery should be left 'holding the bag' when the bubbles collapse. In contrast, isolates have no way of knowing about core managers' investment positions, and consequently no means of imitating these positions."
There are a number of fascinating implications of this. The copycat, me-too, funds in the periphery apparently serve as a dumping ground for the trades that the funds in the core have put on and are looking to unwind. Meanwhile, there are funds which are isolated, either by choice or not, from Wall Street groupthink.

I've written about some of the dynamics of Wall Street groupthink before. For example, Andrew Ross Sorkin's account of a lunch with Oliver Stone in Manhattan. Also in my review of Conquer the Crash, I noted that "people who 'got it' for whatever reason have gotten constant, positive, psychic feedback for three decades. They have been carefully programmed to ignore the types of issues that we discuss on Credit Bubble Stocks." And in my review of Diary of a Very Bad Year, about how the advantage of being in Manhattan has turned into a disadvantage for investment operations: "parasites have gathered to tax you, and there's no space for independent thought".


Taylor Conant said...

Judging by the quality of comments typically found on this blog, it will be a long time before any copy-cat blames their poor performance on their attempt to mimic you and your strategy!

(Which is probably in and of itself a virtue for your strategy and another reason you have an edge. Compared to Warren Buffett, for example, you have a lot more room to maneuver because you have a lot fewer people trying to predict your next move and bid up those assets before you get there.)

CP said...