Vector error correction and multiple regression analysis to find the effect from gross domestic product, exchange rate, consumption & treasury bill rate on inflation in Sri Lanka

Abstract

Inflation is the most important macro-economic variable mainly used for economic management. There is a widespread agreement that high and volatile inflation can be damaging both to individual businesses and to consumers and hence, to the economy as a whole. When inflation is highly fluctuating and unable to correctly predict, individuals do not tend to invest money for various projects. The objective of this study is therefore to find out the effect of economic variables Gross Domestic Product (GDP), Exchange Rate (USD), Rice Price (RCP) and Treasury bill rate (INTRST) on CCPI using quarterly data on each variable for the period 1979-2005 in Sri Lanka. The Colombo Consumer Price Index (CCPI) is used as a proxy variable for inflation. Unit root test confirmed that neither series are stationary in its levels nor first difference of all the series. Therefore Vector Error Correction methodology was carried out to find a suitable model for inflation (CCPI) and identified that series of GDP, USD and RCP granger cause CCPI but INTRST does not do so. The final model developed is: ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) According to the above model lag 2 of CCPI, lag 2 of GDP, lag 1 and 3 of RCP show a positive correlation with CCPI while lag 3 of CCPI and GDP and lag 2 of RCP show a negative correlation with CCPI. It is therefore confirmed that GDP and RCP are the significant factors for inflation in Sri Lanka, and USD and INTRST have no significant effect on inflation. By using the selected model CCPI values are forecasted and the Mean Absolute Percentage Error (MAPE) of the fittede model was found less than 5%. Therefore above model is recommended as the suitable model for forecast CCPI.