A0490
Title: Time-varying ambiguity shocks and business cycles
Authors: Xiaojing Cai - Okayama University (Japan) [presenting]
Abstract: Relationships between ambiguity and macroeconomic variables are investigated. Following previous research, ambiguity is measured as a weighted average of the variances of probabilities. A longer dataset is employed, and whether ambiguity impacts key macroeconomic variables such as outputs and inflation in the U.S. over the post-World War II period is explored. In addition to ambiguity obtained from market risk premiums, it is estimated from size, value, and momentum risk premiums. To assess the effects of ambiguity over 70 years, a time-varying parameter vector autoregressive model with stochastic volatility (TVP-VAR-SV) that allows us to reflect structural changes in the economy is employed. The results provide evidence that the relationships between ambiguity and macroeconomic variables vary over time. Specifically, it is found that an increase in ambiguity led to an increase in output during the high inflation periods in the 1970s and the 1980s, which is consistent with the ambiguity lover behaviour previously. A negative relationship between ambiguity and inflation in the 1950s and a positive relationship in the 2000s are also observed, which indicates that unfavourable outcomes differ between the former and the latter period. Moreover, it is uncovered that ambiguity obtained from other risk premiums is weakly associated with macroeconomic variables, suggesting that size and value risk factors do not capture information about the entire market.
