Prediction and Optimization of Export Opportunities Using Trade Data and Portfolio

The modern portfolio theory targets to achieve a safe investment while extracting maximum profit. Its use in exploring export opportunities is undocumented. Traditionally, the gravity model of trade is widely used to calculate trade flows while the prediction of trade flow was based on application of time-series prediction algorithms on historical trade data. The proposed research introduced the risk involved in the trade opportunity as a quantitative factor determined by product complexity and gravity model of trade, while predicting the optimal export commodities to maximize profit and minimize risk. Improvement in trade prediction accuracy using portfolio optimization methods as compared to other previously documented methods is also reported. The results indicate MSE of 0.161 and 0.239 using Black Litterman model and CAPM against 1.226 and 1.026 using the traditional Holt and Grey models respectively. The results are supplemented by the level of risk attached to each commodity, to classify the optimal products for export investment.