Equity Valuation and Portfolio Management

QUESTIONS

1. What is the motivation of using the Alpha Repair investment process?
2. What's the objective function used to select factors in the Alpha Repair strategy?
3. Why should a relatively small number of factors such as three factors be selected?
4. Does the more frequent change of factors selected in recent years mean higher stock turnovers?
5. Are stock portfolios simply a basket of the top ranked stocks?

¹ The persistence of volatility and correlation are well documented, and are at the heart of generalized autoregressive conditional heteroscedasticity (GARCH) forecasting models. See, for example: Robert Engle and Joseph Mezrich, “Grappling with GARCH,” Risk 8, no. 8 (1995): 112–117; and Robert Engle and Joseph Mezrich “GARCH for Groups,” Risk 9, no. 8 (1996): 35–40. We do not use GARCH forecasting models here, but we do rely on the persistence of covariance.

² The surprisingly low number of factors is basically consistent with an observation from our principal components analysis that 80% of the variance in our set of factors (excluding beta and one-month return) is produced by three principal components.

³ See Joseph J. Mezrich and Junbo Feng, “Alpha Repair,” Nomura Securities International Inc., January 22, 2007.