ARE EMPIRICAL MEASURES OF MACROECONOMIC UNCERTAINTY ALIKE?
研究了时间序列模型产生的通胀和增长不确定性度量与专业预测者调查中的不确定性度量是否可比，发现使用实时数据和固定样本递归估计的非对称双变量GARCH模型能产生与调查度量相似的通胀不确定性估计，但增长不确定性度量普遍偏高。
Abstract There is a plethora of time series measures of uncertainty for inflation and real output growth in empirical studies but little is known whether they are comparable to the uncertainty measure reported by individual forecasters in the survey of professional forecasters. Are these two measures of uncertainty inherently distinct? This paper shows that, compared with many uncertainty proxies produced by time series models, the use of real-time data with fixed-sample recursive estimation of an asymmetric bivariate generalized autoregressive conditional heteroskedasticity model yields inflation uncertainty estimates which resemble the survey measure. There is, however, overwhelming evidence that many of the time series measures of growth uncertainty exceed the level of uncertainty obtained from survey measure. Our results highlight the relative merits of using different methods in modelling macroeconomic uncertainty which are useful for empirical researchers.