Exchange RatePrediction Using Singular Spectrum Analysis

The effects of foreign exchange rates on economic variables in every countries show the importance of modeling and prediction of exchange rates.In this paper, singular spectrum analysis (SSA), which is a non-parametric technique for time series analysis, are used for modeling and predicting daily exchange rate US dollar when compares with Iranian Rials (USD/IRR) during June 2013 to Sep. 2015. ARIMA model is used as a benchmark to assess the performance of SSA. In order to find the best ARIMA model, R package auto.arima is used. In addition, errors in sample and out of samples for short, medium and long term forecasts are considered to compare the capabilities of models with together. Results indicate that SSA is able to be used for modeling exchange rate data.