Exchange Rate  Forecasting and Model Selection in Pakistan (2000-2010): Exchange Rate Forecasting and Model Selection in Pakistan (2000-2010)

Khurram Saleem Malik Khurram Saleem Malik

doi:10.2112/jbe.v3i1.31

Authors

Khurram Saleem Malik .

DOI:

https://doi.org/10.2112/jbe.v3i1.31

Keywords:

Forecasting exchange rates, Purchasing power parity, AR model, ARCH model, ARMA model, Time series decomposition model, Combined forecast model

Abstract

Emerging markets like Pakistan are becoming an attractive place for foreign
investors. Foreign investment crucially depends on expected exchange rate
movements. This study attempts to determine, whether or not, the exchange
rate in Pakistan can be forecasted using different exchange rate models. The
specific objective of this study is to determine the best model. This is done by
analyzing forecasting performance of various univariate and multivariate
exchange rate models. Seven models including the Autoregressive (AR),
Autoregressive moving average (ARMA), Autoregressive conditional
heteroscedasticity (ARCH), Decomposition of time series, Purchasing power
parity (PPP), Dornbusch Frankel sticky price monetary (DB) and the
Combined forecast models, are all estimated using monthly data over the
period January 2000 to June 2010. ARCH model is found to be the best
model for forecasting exchange rate in Pakistan for the selected time period
followed by combined forecasting and autoregressive (AR) models.

Exchange Rate Forecasting and Model Selection in Pakistan (2000-2010)