Unveiling the Potential of PPP Theory: A Practical Approach to Short-Term BDT-USD Exchange Rate Forecasting in Bangladesh

Abstract

This study explores the viability of the PPP (Purchasing Power Parity) theory in predicting short-term exchange rate movements in Bangladesh. We aim to construct a forecast model, exclusively based on PPP theory, to accurately estimate the 30, 60, and 90-day forward exchange rates of BDT-USD (Bangladeshi Taka - US Dollar) with minimal error. Drawing from approximately 9 years of monthly data, we utilize monthly nominal CPI values from Bangladesh and the USA to compute six inflation differentials across various periods (30, 60, 90, 120, 150, and 180 days). To determine the lagged impact of inflation on exchange rates, we employ a straightforward correlation matrix with associated p-values. Among these sets, the one exhibiting the highest correlation (along with the lowest p-value) with the percentage change in the 30-day forward rate is identified as the very variable having the highest impact on the next 30-day forward rate. This process is repeated for the 60 and 90-day forward rates, leading to three distinct equations for forecasting each duration. Finally, error-adjustment variables are incorporated in these equations. Our model relies on five readily available data points to forecast forward exchange rates. Results indicate that this model accurately forecasts the 30-day forward exchange rate with a ±0.58% error margin and 98.84% accuracy, with statistical robustness at a 5% significance level across the sample period. However, the performance diminishes when forecasting 60 and 90-day forward exchange rates. This study underscores the effectiveness of PPP theory in predicting up to 30 days of forward exchange rates in Bangladesh, highlighting its practical applicability in economics and finance.