人民币非抛补利率平价为什么不成立:对4个假说的检验

Uncovered Interest Rate Parity Puzzle: Four Hypothesis Tests

本文采用市场调查的汇率预期数据,运用非线性时变平滑转换模型验证了"风险溢价"、"交易成本"、"外汇市场干预"和"套利受限"4种非抛补利率平价不成立假说。结果表明,交易成本会改变汇率对利差的反应方向,在交易成本较低的区制,人民币非抛补利率平价倾向于成立。时变的风险溢价对人民币非抛补利率平价系数的偏离并没有显著影响。央行的外汇市场干预和跨境套利收益变化会改变人民币对利差的响应时间和程度,导致非抛补利率平价系数的偏离更加严重。总体上,人民币非抛补利率平价并不成立,而且近期偏离程度越来越高。这意味中国的资本账户开放政策,与以中间价为基础的人民币汇率形成机制,在短期正面临越来越突出的矛盾。

This paper uses survey data onexpected market exchange ratesto carry out an empirical investigation on the RMB against the U.S. dollar uncovered interest parity. It incorporates a nonlinear time-varying coefficient depends on 'risk premium', 'transaction costs', 'foreign exchange market intervention' and 'limited arbitrage', which are four uncovered interest parity hypotheses The results show that transaction costs will change the direction of exchange rate reaction to interest spreads: in the regime with high transaction cost it has persistent deviations from UIP, but the UIP holdsin the regime with lowtransaction cost. Time-varying risk premium has no significant effect on UIP deviation. The central bank's intervention and the variance of cross-border arbitrage return will change the RMB response speed to interest rate spreads, which will cause the deviation of UIP coefficient become more serious. To sum up, the RMB UIP wasprovedto be incorrect and RMB UIP deviations are increasingly penetratingrecently, which means that the contradiction between the gradual opening of China's capital account and the RMB exchange rate formation based on middle rate is increasingly severe.