外汇干预反应函数非线性FTR模型的提出及其实证

中央银行外汇干预反应函数是描述央行干预行为与外汇市场条件变量间关系的重要工具,对于揭示央行外汇干预规律和特征具有重要意义。现有的干预反应函数模型在充分体现外汇干预的不频繁性特征、以及提高描述精度两方面上,还不能同时满足实际的需要。在参考相关研究基础上,从兼顾模型实际意义与描述精度的角度出发,提出了一种新的非线性FTR模型,并基于日本央行的外汇干预数据对其进行了实证研究。将其与线性模型、2-机制TR模型进行实证比较后发现,所提出的FTR模型在样本内拟合和样本外预测两方面均具有比其它两种模型较小的误差,从而证明了FTR模型在描述央行外汇干预行为上具有更好的精度,更加适合作为央行外汇干预反应函数。

Advancement and Empirical Research of Nonlinear FTR Model as Central Bank’s Foreign Exchange Intervention Reaction Function

The foreign exchange intervention reaction function of central bank is an important tool to describe the relationship of intervention behavior and some market conditions,and it is significant for revealing the rules and characters of intervention behavior At present,the models of reaction function provided are not good enough to meet the need of explaining the infrequency of intervention,and acquiring high precision simultaneo usly Based on the analysis of related research,and from the view of practical significant and high precision,this paper provides a new nonlinear FTR model,and executes experiments based on the intervention dat a of Japanese central bank.Experiment al results demonstrate that,comparing with linear model and two regime TR model,the FTR model provided by this paper has smaller errors both in in sample fitting and out of sample forecasting processes,that is,it is more precise in describing the intervention behavior,and is more suitable to be the central bank's intervention reaction function.