Why The Search for Alpha is More Like Baseball Than Mining

10 August 2006

By: Alpha Male 

The hedge fund lexicon seems to borrow heavily from the world of mining. Pundits talk about “mining” alpha, “prospecting” for alpha, and “extracting” alpha. They say that, like oil, there is a “limited supply” of alpha in world. In fact, like “peak oil” theorists, some are frightened that we are quickly exhausting the finite supply of alpha in the world – call them “peak alpha” theorists.

These “Peak alpha” theorists argue that structural inefficiencies that allow repeatable market-beating returns will soon be ironed out of financial markets. But those managers who are able to exploit such inefficiencies are taking them not from the ground, like oil or gold, but from other investors. We use the mining analogy because these “other investors” are such a diverse, faceless, and indefinable group. You never know exact who you took the alpha from – so you treat it as if it was simply lying there in the ground.

But this is where the resource extraction analogy falls down. Unlike oil, alpha was not left to us en masse by Mother Nature. There is no finite amount of alpha in the world ready for us to exploit. Alpha is situational, transitory, and relative. It has no physical dimensions and cannot be measured directly. As William Sharpe famously taught us, Alpha is a zero-sum game. For every investor who receives alpha – another investor looses it. Tris Lett, the “elder statesman of the Canadian hedge fund industry”, recently pointed out to me that such alpha “losers” might have less information, inferior skill, or simply have a different utility curve (witness the non-economic players in the commodity futures markets).

However, mining is not a zero-sum game.  Every day, we all have more oil and gold.  That’s why alpha is more like another well known zero-sum game: Major League Baseball.

You don’t need to be a big baseball fan to figure out that, across major league baseball, the number of overall victories in a season is exactly the same as the number of losses. In other words, the sum of all baseball victories (represented by +1) and losses (represented by -1) is zero. Still, in the face of overwhelming evidence that all teams will end up at .500 in the long-run (assuming a level playing field across the league), we continue to root for our home team.

But while wins are a zero sum game, some teams seem to be able to repeatedly beat this .500 “benchmark”. They display return “persistence” by being good year after year. This can only be possible by exploiting some structural, extra-competitive inefficiency such as the increased TV revenue resulting from local monopolies in larger markets or a fan base that supports a team even if it looses year after year. (Obviously, these fans have different utility curve than others. Like the “non-economic players” in the commodities market, they have “non-competitive” utility functions.)

So, across all of major-league baseball, wins equal losses. However, thanks to inter-league play, the total number of wins in the National League might be larger than the total number of losses in that league. The corollary must be that the American league has the opposite record. For hedge funds, every game is an inter-league game. They compete not only against themselves, but also with traditional long only funds and individual investors. So in theory, it is possible for the National League to have a better record than the American League. But, as above, structural inefficiencies must exist for this situation to persist. Perhaps the National League and the hedge fund industry are able to pay for more skilled players. Perhaps the National League spends more, in aggregate, on researching & analyzing their opponents. Or perhaps the American league teams (and fans) simply don’t have as strong an incentive to win as do the National League teams.

In any case, baseball victories cannot be “mined” or “extracted”. They are offered up by the losers. One could argue that techniques for persistently winning are finite and disappearing – although if they were, all teams would be at .500 (or at least normally distributed around .500).

Treating alpha as a finite and measurable resource leads to many colourful metaphors. But they ultimately lead end investors to conclude, erroneously, that alpha and hedge funds, like oil or gold, are a asset classe, not investment strategies.

 Email this post to a friend   Print This Post 

2 Responses to “Why The Search for Alpha is More Like Baseball Than Mining”

1. All About Alpha » Blog Archives » A Mass of Alpha Says:
   September 28th, 2006 at 11:57 pm

   […] Alpha Male agrees that, like oil, there will always be new sources of alpha to exploit.  But the oil metaphor does have its weaknesses.  In this post on why the search for alpha is more like baseball than drilling for oil, Alpha Male argues: “Unlike oil, alpha was not left to us en masse by Mother Nature. There is no finite amount of alpha in the world ready for us to exploit. Alpha is situational, transitory, and relative. It has no physical dimensions and cannot be measured directly.” […]

2. All About Alpha » Blog Archives » Alpha as a Net Zero Sum Game: How Serious a Constraint? Says:
   December 14th, 2006 at 12:12 pm

   […] Hill makes an argument that we have likened to interleague play in Major League Baseball.  She says that certain segments of the market might generate net positive aggregate alpha on the backs of other segments of the market.   […]