摘要：Prior research on the relationship between output volatility and
growth has produced mixed results. This paper investigates the
volatility-growth relationship from a different empirical angle, focusing
on the signs of the volatility-growth relation across business
cycle phases. We fit two strands of models (i.e., the ARCH-type
model and Markov-Switching model) to the U.S. real GDP data,
and find results supporting that the volatility-growth relation is
positive when the economy is in the expansion phase, while higher
volatility lowers growth rate in the contraction phase. Therefore,
empirical investigations of data without taking this feature into
account will fail to recover non-trivial relation between volatility
and growth.
