The housing cycle leads the business cycle, or as pointed by Leamer (2007): \Housing is the Business Cycle". Across all components of U.S. GDP, residential investment contributes most to the weakness of GDP prior to recessions and almost all recoveries can first be observed in residential investment as well. In this study, we focus on five GDP components, residential investment, durables, nondurable consumption, investment in business structures, and investment in equipment and software.
The growth rates of all examined GDP components are also well-known risk factors in the asset pricing literature. Consumption growth is motivated by the model of Lucas (1978) and Breeden (1979). The growth rate of the stock of durables is used by Yogo (2006), and the use of the change in durables investment as a factor is introduced by Belo (2010). Asset pricing implications of residential investment growth and nonresidential investment growth are examined fi rst by Cochrane (1991, 1996).
However, none of these studies or their successors take into account the strong lead-lag pattern in the GDP components, although this stylized fact is well documented in the macroeconomics literature. Novel, and so far overlooked, we focus on exactly this strong lead-lag behavior when examining and comparing di fferent GDP components as factors for asset pricing and ll this gap. The housing cycle leads the business cycle, thus the hypothesis we test is that, among all GDP components, residential investment can best explain the cross-section of stock returns.
If asset returns (the size and value premium in particular) are related to the business cycle, they should covary with the leading component and much less with the lagging components. Durables closely follow residential investment. We hypothesize that durables perform second best. The model with consumption growth should perform fair, by placing itself in the middle of the field of the real economic variables. And finally, we expect that equipment and software and business structures as factors perform worst, because business investment lags the cycle.