Uncertainty About the Persistence of Economic Shocks
通过似然函数分析，发现经济产出时间序列中冲击持续性的不确定性远大于通常置信区间所暗示的程度，并指出向量自回归估计中也需要谨慎评估不确定性。
The behavior of time series of economic output has been characterized in terms of the persistence of shocks. For a series generated by an autoregressive integrated moving average model, persistence is a function of the model parameters. We demonstrate that a careful analysis, based on the likelihood, can lead to considerably more uncertainty about persistence than is implied by the confidence intervals that are usually calculated. We examine also the estimation of persistence from a vector autoregression, in this case we also find nonstandard concentrated log-likelihood functions, concluding again that considerable care is necessary to assess uncertainty.