Have the New Factor Models Improved on the Theory Front? Evidence from Decomposing Asset Growth

Abstract

Contrary to the theoretical foundations of the asset-growth factor in both the Fama and French (2015) five-factor model and the q-factor model of Hou, Xue, and Zhang (2015), this study finds that investment level per se is not the reason asset growth is negatively associated with future stock returns. Instead, the source of the negative return predictability is the structure of the investments–specifically, the division of total investments between those that are carried on the balance sheet, versus those that are off the balance sheet, and is endogenously linked with uncertainty because accounting standards prohibit highly uncertain investments from appearing on the balance sheet. This study also sheds new light as to why these different types of investments are related to future stock returns in opposite directions: the fundamental difference is not a function of their tangible versus intangible nature; rather, it is their placement in the financial statements. As such, the opposite directionality represents two sides of the same coin.