Alpha effects have significant persistence. Some studies have failed to find alpha persistence, we suspect largely because the effects, although significant, are not long-lived. (Alpha performance advantage begins to dissipate after about a month; if a research alpha is constructed on 12-months or 36-months of data, it subsumes too much disparate information to have useful predictive value.)
To illustrate the Alpha Dynamics effect, we set up a monthly evaluation of four style funds; Large/Growth, Large/Value, Small/Growth, and Small Value. Their alphas are ranked each month. Percent return is measured for the top alpha and bottom alpha over the ensuing 21 days (beginning with the close of business the day following alpha evaluation).
For the ten years ending 12/31/03, this process accumulated an average annual +11.9% for top-alphas vs annual +5.1% for the bottom-alphas. The annualized S&P 500 rose by 9.1% over the same period.
During the bull market years 1993-2000 both top-alpha and bottom-alpha styles did well...roughly in line with (but a little less than) the S&P. In the bear market years, however, the top-alpha styles not only preserved value but actually grew.
Applied across 22 general equity funds, the same alpha-selection process accumulated +13.2% (top 3 alphas) vs -4.4% (bottom 3).
Click on Services link to request monthly Alpha Dynamics updates with past month's performance effects and current month's alpha rankings.