Forecasting Exchange Rates Using Time-Varying Parameter Model

• Main Authors: TSAI, WEN-HAO, 蔡文豪
• Other Authors: LIANG-SHUH
• Format: Others
• Language: en_US
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/a8fgq7

碩士 === 東吳大學 === 國際經營與貿易學系 === 105 === Taiwan has been relying on foreign exchange intervention policy to decide or protect various domestic financial products or industries, so exchange rate fluctuations will affect the Taiwanese enterprises. In order to accurately predict the exchange rates, we provide daily data on the endogenous variables to predict exchange rates, and these endogenous variables are the exchange rates decision of one of the important factors. We propose a Bayesian vector auto-regressive model with time-varying parameters (BVAR-TVP) to examine the short-term predictability of exchange rates of Taiwan. An important contribution of the paper is the application of the BVAR-TVP model is that investors could have made excess profits if they had followed trading strategy based on the signals generated by the model’s one-day-ahead exchange rates forecasts. We employ criteria and statistical tests to assess the exchange rates predictability. The predictions show that the predicted results prove that Bayesian time varying model can predict the exchange rates of Taiwan and forecast the situation after the profit.