Working Paper: CEPR ID: DP13049
Authors: Martin Lettau; Markus Pelger
Abstract: We propose a new method for estimating latent asset pricing factors that fit the timeseries and cross-section of expected returns. Our estimator generalizes Principal Component Analysis (PCA) by including a penalty on the pricing error in expected returns. We show that our estimator strongly dominates PCA and finds weak factors with high Sharpe-ratios that PCA cannot detect. Studying a large number of characteristic sorted portfolios we find that five latent factors with economic meaning explain well the cross-section and time-series of returns. We show that out-of-sample the maximum Sharpe-ratio of our five factors is more than twice as large as with PCA with significantly smaller pricing errors. Our factors are based on only a subset of the stock characteristics implying that a significant amount of characteristic information is redundant.
Keywords: cross section of returns; anomalies; expected returns; high-dimensional data; latent factors; weak factors; PCA
JEL Codes: C14; C52; C58; G12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| RPPCA (R50) | identification of latent asset pricing factors (G19) |
| RPPCA (R50) | detection of weak factors with high Sharpe ratios (C58) |
| RPPCA (R50) | quality of factor identification (C38) |
| five latent factors from RPPCA (C38) | explain covariance and expected return structure of anomaly portfolios (C10) |
| characteristics used (C52) | factors identified (D91) |