When is alpha not really “alpha”? When it’s “beta-alpha.”

CAPM / Alpha Theory 12 May 2010

Put up your hand if you’ve heard some pundit tell you recently that “we’re now in a stock picker’s market”?  The thrust is, many of those pundits seems to say the same thing year after year, don’t they?

In fairness, there is a good reason to say that a sideways market is a “stock-picker’s” market since you can’t rely on beta to produce returns.  But the reality is that pundits forecast sideways or modestly trending markets more often then not.  If they argued otherwise – that the stocks were going to rise or fall en masse – then why would you need the help of their market punditry?  Besides, if markets are fairly valued at any given moment, it’s technically impossible to predict the market anyways – leaving no choice but to predict the safe bet: a sideways, stock-picker’s market.  It’s almost as if they have an interest in the retail investor being scared into hiring professional stock-picker to navigate the markets…

But when all is said and done, do “stock pickers” actually produce great returns, or do the best returns belong to those who time their sector or market allocations?  That’s the question tackled in an interesting paper by Eric Petroff, CFA, CAIA, and Curtis Yasutake of Wurts & Associates, an institutional investment consulting firm.

As the duo writes,

“There is more to finding excess returns than seeking out managers with purported stock picking abilities or by simply loosening tracking error constraints…Whether or not investors are aware of it, the bulk of excess returns are driven by differentiated sector (or Beta) exposures to a given benchmark, not security selection; or excess returns are mostly Beta-Alpha.”

Petroff and Ysutake define alpha as the return not attributable to sector weightings – in other words, returns directly resulting from security selection.  They define returns from sector bets separately as “beta-alpha.”

When fund returns are regressed against a broad market index like the S&P500, they may appear to produce alpha.  But the duo points out that much of this “alpha” might disappear if you ran a multiple regression of fund returns against various sector benchmarks.

They did just that with the US Large Cap Core managers listed in the eVestment Alliance database and here’s what they found…

As you can see, the alpha using a multiple regression pales in comparison to that calculated using a simple regression.  Petroff and Yasutake point out that a fund with a high alpha from a simple regression and a very low alpha from a multiple regression likely derives most of its returns from sector weighting.  A fund with a high simple-regression alpha and an equally high multiple regression alpha must be producing returns from something other than sector bets (namely, as the authors conclude, security-selection).

They call the ratio of the R-squared of the multiple-regression to the R-squared of the simple-regression, the “beta-alpha ratio.”

It turns out that funds with a high beta-alpha ratio have higher excess returns over both the past 10-year and the past 15-year periods.

Recall that a high beta-alpha ratio means that a higher proportion of returns are the result of sector bets, not security selection (a.k.a. “stock picking”).  So the non-stock-pickers seem to be teaching a clinic to their stock-picking comrades.

It makes intuitive sense that a fund with a high tracking error (low benchmark correlation) would have the opportunity to produce returns in excess of that benchmark.  But the duo finds that this is doubly true for funds with a low R-squared and a high beta-alpha ratio.  As you can see from the chart below from the paper, it’s just as true for index-huggers too…

So it would appear as though a “stock-picker’s market” may not be such a great thing after all.  In fact, as Petroff and Yasutake conclude:

“If you truly desire consistent excess returns over and above a benchmark, the single best way to do so is through Beta-Alpha, not Alpha.”

But all is not lost if you are a disciple of Graham & Dodd.  The authors continue…

“…This can be accomplished through deliberate top down sector allocations or bottom-up research that necessitates a deviation from benchmark sector targets.”

In conclusion, the duo has some words of advice for their fellow investment consultants.

“…the consulting industry has spent too much time building the systems to provide such analysis [sector exposures analysis] while entirely missing the implications thereof, as the true key to realizing excess returns is finding those who achieve them through Beta-Alpha.

“…through its fixation on benchmarking and attribution analysis, the consulting industry drove managers to seek excess returns through Alpha, as opposed to Beta-Alpha. And as we’ve shown in this paper, doing so is an unwise course of action.

“…Though we are not about to claim the Beta-Alpha Ratio is the definitive methodology to choose successful managers ex-ante, we do claim it will improve the chances of success over and above any other methodology prevalent in the consulting industry.”

Be Sociable, Share!

Leave A Reply

← Study finds private equity differs markedly across Asia A "new era" for funds of hedge funds: Sunrise or sunset? →