Forecasting Value at Risk of Foreign Exchange Rate by Integrating Geometric Brownian Motion

Abstract

Foreign exchange is one of the most important financial assets for all countries around the world including Malaysia. After recovering from the Asian financial crisis, Malaysia tried to build a strong currency in order to maintain the economic performance. The study focuses on Malaysia foreign exchange rate and foreign exchange risk between ten currencies, which are CNY, SGD, JPY, EUR, USD, THB, KRW, IDR, TWD and AUD. Unpredictability of the foreign exchange rate makes the traders hard to forecast the future rate and the future risk. The study implements the parametric approach in the Value at Risk (VaR) method and the geometric Brownian motion (GBM) model. The objectives of the study are to integrate the VaR model with the GBM model in order to compute or forecast the VaR. By using parametric approach, the study successfully computes the VaR of foreign exchange rate for different confidence levels. The GBM model is suitable to forecast the foreign exchange rate accurately using less than one year input data and using the log volatility formula. Lastly, the study verifies the feasibility of the integrated model for a one month holding period using the data shifting technique. In conclusion, the prediction of future foreign exchange rate and foreign exchange risk is important in order to know the performance of a country and to make better decision on investment.