Forecasting exchange rate based on macroeconomic variables: a GVAR approach

In this thesis, we conduct this study to forecast real effective exchange rate based on macroeconomic variables such as unemployment rate, real interest rate, current share price, industrial production, terms of trade and terms spread along with a global variable real commodity price index. We employ Global Vector Autoregression model (GVAR) developed by Pesaran (2004) and analyse the global impact of the country specific domestic variables on various exchange rates. Our empirical analysis with the help of impulse response reveals that positive shock to the UK and US share price has significant impact on Canadian dollar. Furthermore, a positive shock to the real commodity price index reflects some striking result, narrating that it has significant impact not only on the exchange rate but also on the real interest rate,terms of trade and industrial production. Using newly assembled data and hitherto applied methodologies; we believe that this study imparts valuable insights into the role of the macroeconomic variable on the foreign exchange rate market.