Recent empirical literature shows that the size effect, small capitalization stocks have higher returns than large capitalization stocks, has disappeared since the early 1980s. This paper shows that the disappearance of the size premium can be attributed to change in duration of the business cycle. We find that time variation in the size premiums is crucially dependent on the economic state. Size premiums are significantly positive mainly at the bottom of the business cycles. During other business cycle stages, the size premiums are indistinguishable from zero. We also find that this dependency of the size premium on the business cycle is preserved even after the early 1980s wherein there is absence of the size premium. This finding suggests that whereas unconditional size effect is dead, the size effect conditional on the state of the macroeconomy is alive and well. Given that the pattern of the size premium conditional on the phase of the business cycle remains strikingly stable before and after the early 1980s, we then examine whether any change in the lengths of phases of the business cycle is the primary cause of the disappearance of the (unconditional) size premium. Our formal statistical analyses reveal that the probability of economy being in recessionary periods has been significantly reduced since the early 1980s. Specifically, the probability of economy staying in recessionary periods after the early 1980s is less than half of that before the early 1980s. Finally, we illustrate the important role of duration of business cycle in explaining the disappearance of the size premium by using the stylized Markovian economy, and provide the economic intuition behind the results.
Keywords: Size effect; Business cycle duration and asymmetry; Markov switching model
