Why bother separating alpha and beta? Here’s why.

CAPM / Alpha Theory 08 Nov 2009

goingseparateways2The cast and crew of AllAboutAlpha.com were in Los Angeles this week meeting with some of our favorite alpha centric asset managers and investment management consultants.  One of those companies was quant manager Analytic Investors.  Regular readers will remember the names Roger Clarke, Harindra de Silva, Steven Thorley and Steve Sapra for their work on extending the “law of active management” and penning a seminal work on short extension (a.k.a. 130/30) strategies.

Clarke, de Silva and Thorley were at it again this year with the release of a very interesting “monograph” (translation: “100 page mini-book”) on alpha/beta separation.  This paper is required reading for anyone with an opinion (either positive or negative) on the somewhat controversial strategy.  With so-called “alpha” allocations producing decidedly beta-like returns the past couple of years, many have discounted the value of delineating alpha from beta in the first place.

As the trio writes:

“The separation of alpha and beta sources of return in institutional portfolios has arrived and is having a profound influence on the way investors view risk and return. Some observers believe that the impact of alpha–beta separation will be as transformative as modern portfolio theory was in the 1960s, while others consider it merely a passing fad. As usual, the truth is probably somewhere in the middle, but the need for a better understanding of alpha–beta principles and terminology among investment professionals is clear.”

Clarke, de Silva and Thorley present a methodical and cogent argument for why alpha and beta should be separated in the first place.  Much of the material on this topic is either too high level (marketing bumpf) or too technical (a lesson on how to execute a swap for beta exposure).  This monograph is a “Goldilocks” description of alpha/beta separation in our opinion.  Those of you who read “Portable Alpha: Theory and Practice” edited by PIMCO’s Sabrina Callin (who was also on our itinerary this week in LA), will find this to be a useful complement to that book.

One of the questions posed to proponents of alpha/beta separation is “Why?”  Why would you even want to separate the alpha and beta that is embedded in every active fund or investment mandate?  This monograph answers this question in a concise, yet sufficiently-detailed manner.

The problem with active management, say the trio is that the ratio of active and passive risk arises organically out of the myriad of separate investment decisions made by the manager.  The resulting ratio may not yield the highest possible Sharpe ratio, however.  In other words, it may be sub-optimal.

For example, assume you owned an actively managed fund with a 0.62 Sharpe ratio.  If you decide to allocate, say, 10% (or even 50%) to cash, you’d have a fund with a lower volatility, a lower return and, of course, the same Sharpe ratio.

But that (unlevered) 50% portfolio – like any actively managed portfolio – is made up of active and passive risk.  It’s essentially made up of a market beta portfolio and a “pure alpha” portfolio.  If you create an efficient-frontier-like line tying together all possible combinations of these two separate funds, you can see that the unlevered fund you began with is really just one arbitrary point on this line.

Unfortunately, it’s not necessarily the point with the highest Sharpe ratio.  Recall that our fund has a Sharpe of 0.62.  But by reallocating between the embedded “alpha fund” and the embedded “beta fund”, you can increase the Sharpe ratio to 0.73 (see chart below from the monograph).


So what?  Well, if you wanted to allocate 90% to the active fund, you could have a 0.62 Shape (“90/10 Mix” above) or a 0.73 Sharpe by allocating 90% to active fund and shorting the market.  The table below corresponds to the chart above:


Like the apparent free lunch served up in an article we covered recently by the rocket scientists at First Quadrant (another stop on our tour of LA this week), Clarke, de Silva and Thorley seem to have pulled a rabbit out of a hat here.  But as they point out, this is simply a logical explanation for the “active risk puzzle” identified by Goldman’s Robert Litterman earlier this decade.  The problem is essentially that active/passive combinations found in active mandates (whether they are of the mutual or hedge variety) leave part of lunch sitting on the table.

There’s plenty more in this paper worth highlighting.  We’ll get to some of it in an upcoming post.

Be Sociable, Share!

One Comment

  1. protege
    February 23, 2010 at 5:22 am

    can someone explain how they determine the -60% exposure to the index future?

Leave A Reply

← Equity long/short mutual funds "could easily grow twenty-fold over the next five years:" Report Can of worms? Supreme Court discusses "fair" compensation for fund advisory services →