Authors ’ Coordinates

It is now generally agreed that multiple factors drive asset returns. Their identities, however, remain subject to wide debate, despite a plethora of factor candidates nominated in a voluminous literature. We propose a protocol for sorting them out. Our protocol involves seven empirical stages that successively identify non-risk factors that are not priced, non-risk factors that are (i.e., arbitrage opportunities), risk factors that are not priced, and those that are. Certain principles apply. Factor movements should not be easily predictable. A characteristic such as firm size, or anything else known in advance, cannot be a factor. However, characteristics can be related to mean returns either because they happen to align with factor loadings or because they represent arbitrage opportunities. Truly pervasive factors with accompanying risk premium must be related to the covariances among returns on real assets only, not to the covariances of any assets in zero net supply, such as bonds and derivatives. Time variation in covariances is a challenge, but soluble with recently developed statistical procedures