Exchange Rate Forecasting Using Multiscale Vector Autoregressive Model

In the increasingly diversified and globally integrated market environment, the accurate forecasting in the exchange markets needs to take into account the heterogeneity at both individual and cross correlation level during the modeling process. In this paper, we propose the Heterogeneous Market Hypothesis based exchange rate modeling methodology to model the micro market structure. We further propose the implementation algorithm under the proposed methodology to forecast the exchange rate movement. The wavelet based multi resolution denoising algorithm is used to separate and extract the underlying data components with distinct features, which are modeled with time series models of different specifications and parameters. Empirical studies in both Chinese and European markets confirm the significant performance improvement when the proposed model is tested against the benchmark models.