以美元/日元远期汇率为例,利用马尔可夫机制切换算法研究远期汇率对数收益率的异常波动,并将异常波动与宏观经济形势和宏观事件相对照,较好地解释了宏观因素对汇率波动的影响.同时,利用GARCH(1,1)模型拟合剔除异常点后的汇率波动,所得到的不同到期期限的远期汇率的条件方差表明,期限较长的远期汇率的条件波动较大,期限较短的远期汇率的条件波动较小,说明不同期限的远期汇率可能存在信息不对称,即期限越长,信息的不确定性越大.
Setting the USD/JPY forward exchange rate as an example,we discriminate abnormal fluctuations of the forward exchange rate by the Markovian Regime Switch algorithm,then compare abnormal fluctuations with macro-economic events,and explain the influence of the macro-economic factor on volatility of the exchange rate.Furthermore,according to the result from self-correlation and time variation in volatility of the exchange rate,we fit volatility of the exchange rate by GARCH(1,1) model.Then conditional variance of the forward exchange rate for different maturities show that the forward exchange rate of the longest maturity have the maximal volatility,the forward exchange rate of near and the middle maturities always have the minimal volatility,which mean that there are possible asymmetric information among the forward exchange rate of different maturities,that is,the longer the duration,the more uncertain information.