Testing the Lucas Hypothesis on Output-Inflation Trade-offs
推导了一个考虑异方差和两阶段自相关的贝叶斯最高后验区间检验，用于检验卢卡斯方差假设，并利用日本和美国数据展示了异方差和自相关对检验结果的影响。
A Bayesian highest-posterior interval test that accounts for heteroscedasticity and for two regimes of autocorrelation is derived to test the Lucas variance hypothesis on output-inflation trade-offs. Japanese and U.S. data show that heteroscedasticity and autocorrelation tend to influence the test results. A Bayesian procedure and Akaike's information criterion are used to determine the order or autoregressive schemes, and a join point to partition the sample is estimated by a Bayesian procedure.