A0211
Title: Strong trending time series models with endogeneity: A bias-correction method
Authors: Li Chen - Xiamen University (China) [presenting]
Abstract: Endogeneity problems are studied in strong trending time series regression models. We explore the properties of the simple OLS estimator under difference cases of trending magnitudes for the nonstationary time series regressors. A bias corrected estimator is proposed to adjust for the bias in the simple OLS estimator. For the hypothesis tests, numerical simulations show that compared to the OLS estimator, the probability of making the type I error is significantly reduced based on the bias corrected estimator. We also apply our method to estimate the linear regression model of aggregate disposable personal income on the aggregate personal consumption and the real interest rate as an empirical example.
