Theory and Applications of Statical Inference for Economic Time Series

点击翻译按钮获取中文摘要

We investigated the following topics and obtained significant results which were or will be published in journals. Among others we focused on non-standard cases associated with non-stationarity, non-linearity and non-normality.1. Non-stationary time series.(1) Unit root and co-integration test.Several new tests for testing a unit root and co-integration were proposed and studied their performance.(2) Non-invertible time series.Condition are examined in non-standard time series including fractionally differenced ARIMA process(3) Dynamic regression with an integrated regressor.Statistical properties of the OLS are studied and a two step estimator is proposed.2. Economic time series analysis of Japanese monetary and fiscal policy.(1) Analysis of demand and supply function of money by error correction model.(2) Time series analysis of Japanese deposit interest rate.(3) Quantitative evaluation of fiscal policy by multiplier.(4) Time series analysis of inter-governmental fiscal policy3. Non-linear regression and asymptotic expansion(1) The Durbin-Watson statistics in a non-linear regression.(2) SUR estimators and asymptotic expansions of their sampling distribution.(3) Error bounds of asymptotic expansions.