A Synthesis of Two Factor Estimation Methods

Abstract:     
Two-pass cross sectional regression (TPCSR) is frequently used in estimating factor risk premiums. Recent papers have argued that the common practice of grouping assets into portfolios to reducing the errors in variables (EIV) problem leads to loss of efficiency and masks potential deviations from asset pricing models. One solution that allows the use of individual assets while overcoming the EIV problem is iterated TPCSR. We show that ITSCSR converges to a fixed point regardless of the initial factors chosen. ITPCSR are intimately linked to the asymptotic principal components method of estimating factors since the ITPCSR estimates are the APC estimates.

Number of Pages in PDF File: 27

Keywords: Cross-sectional regression, Factor model, Principal Components, Errors in variables

JEL Classification: G1, G12, C3

working papers series