统计研究

本文运用Bai-Perron内生多重结构突变检验方法和结构突变协整方法,针对2000年1月-2008年12月的月度数据,对人民币均衡实际有效汇率及其失调程度进行了重新估计。发现样本期间人民币实际有效汇率确实发生了两次结构突变,突变时点分别为2002年12月和2007年4月。进一步分析发现,实际有效汇率数据生成过程为均值突变单位根过程,这意味着2005年7月汇改对汇率数据生成过程确实产生了结构性变化的冲击,其结果是使得汇率均值无法回复到突变前的水平,此外,样本期内汇率仍各发生了两次明显的低估和高估,但其失调程度却并未像一些未考虑结构突变的研究那么严重。本文最重要的发现是,升值性汇改政策的实施,不仅基本扭转了汇率长期处于低估的局面,还导致其出现了一定程度的高估(截至2008年底,约高估10%)。这与胡春田(2009)、刘玉贵(2009)、彭国富(2010)等最近从完全不同的角度对此问题进行研究所得出的结论一致,故我们可以认为,一个时期以来的(汇改后近年来的绝大部分时期内)人民币汇率并不存在低估,更不存在欧、美等国最近一直所指责的被严重低估。

In this paper,Bai-Perron endogenous structural breaks tests and cointegration with structure method,for the 2000. 01-2008. 12 monthly data,are used to re-estimate the equilibrium real effective exchange rate of RMB and its misalignment. We found that the real effective exchange rate of the RMB during the whole sample,have two structural breaks,respectively in December 2002 and April 2007. Further analysis found that the real effective exchange rate data generation process for the mean mutation unit root process,which means that exchange reform in July 2005 on the exchange rate data generation process does have the impact of structural changes,the result is to make the exchange rate can not return to the mean mutation the level before,in addition,two each in the exchange rate is still significantly undervalued and overvalued in the whole sample,but the misalignment is not,as some studies did not take into account structural breaks get less serious. The most important finding in this article is that the appreciation of the nature of exchange rate reform policy implementation,not only essential to reverse a long period of undervalued exchange rate situation,but also led to the emergence of a certain degree of overestimation (at the end of 2008,about over 10% ). This is Hu Chuntian et al (2009) recently from a totally different perspective on this issue the conclusions from research conducted fairly,and therefore we can assume that the RMB exchange rate does not exist in Europe,America and other countries have recently been accused of being grossly underestimated.