Uncovering Preference Parameters with Higher-Order Consumption Moments
2006/06/13
研討會日期 : 2006-06-13
時間 : 15:00
主講人 : Prof. Biing-Shen Kuo
地點 : B棟110會議室
演講者簡介 : Prof. Biing-Shen Kuo為Ph.D. in Economics，University of Rochester。
現為政治大學國貿系教授。
其主要研究領域為Time Series Theory及Empirical International Finance。
演講摘要 : Empirical attempts using panel data to unveil preference parameters by a log-linearized version of the Euler equation have produced many buzzling results. Recent studies have shown that the approximation bias may be responsible for these anomalies. Motivated by its potential success in reducing the bias, we investigate the economic significance and empirical relevance of higher-order approximations to the Euler equation with simulation method. The approach mounted has the advantages of easy implementation and being less prone to misspecification. Our simulation results clearly reveal that the approximation bias can be significantly reduced when the higher-order moments are introduced, but at the cost of efficiency loss, the conventional tradeoff between bias reduction and efficiency loss. Our study contributes as well to the selection of the approximation order. With the selection criteria proposed in the paper, we find that the mean-squared errors may be substantially reduced simply by adding a small number of the higherorder moments to the second-order approximated consumption regression. Remarkably, when applying our approach to PSID, ‘reasonable’ preference parameter estimates, as suggested in the literature, can now be produced.
