Random walk analysis and forecasting of the Philippine peso and US dollar exchange rates from 1979-2014

• Main Authors: Adriano, Joseleo Rey R., Lim, Jamie Ann L.
• Format: text
• Language: English
• Published: Animo Repository 2015
• Subjects: Foreign exchange rates > Philippines > Forecasting; Random walks (Mathematics); Statistics and Probability
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/14899
• Institution: De La Salle University

Daily Philippine Peso (PhP) - US Dollar (USD) exchange rates from 1979-2014 were examined using random walk analysis and forecasting. This study aims to determine the existence of random walk and find highly accurate predictive models. It was shown that the exchange rates do not follow a random walk by autoregressive integrated moving average (ARIMA) modeling and the obtained best fit ARIMA model is ARIMA(0,1,1). A variance ratio test was also used to analyze the random walk of specific intervals and discovered that 1993-1994 is a relevant time period in this test. In forecasting these exchange rates, linear and nonlinear models were obtained through ARIMA and Artificial Neural Network (ANN). Accurate out-of-sample forecasts were generated using these methods.