Tea industry is a major thrust industryin Sri Lanka which has a significant contribution Sri Lankan economy. Therefore, it is vital to study the behavior of tea export earnings. The study attempts to forecast tea export earnings by category namely Bulk tea, Tea bags, Tea packets and Total exports. The data was obtained from Sri Lanka Tea Board from January 2005 to October 2019. The Vector Autoregressive model was adopted initially by consideringlong run and short run relationship among category-wise tea export earnings via the Johansen cointegration technique. To further explore the dynamic comovement among variables, Vector Error Correction model was used. Residual analysis was carried outwith Residual plot, Correlogram, Residual portmanteau test and the results indicated that model was satisfactory. The analysis revealed that category-wise tea export earnings are co-integrated. Hence, there is a long-run equilibrium relationship between them. Further, bulk tea earnings positively relate to tea packets earnings whereas tea bags earnings negative relate to tea packets earnings in long run.The studyproduced an out-of-sample forecast to analyze and compare the statistical results to determine the precision of the fitted model. Thus, it can be concluded that the fitted VEC model can be used to predict tea export earnings by category in Sri Lanka with significant accuracy.

Keywords : Tea Export Earnings; Co-integration; Vector Error Correction Model.