Abstract/Summary

This thesis examines various drivers of commodity trade flows and it quantifies their predictability and impact on the prices of key industrial commodities. The first empirical chapter discusses how short-term change in the position and gradient of the crude oil market forward curve can contribute to the formation of speculative supply shock. The conditions for causality between the forward curve position, its slope steepness, and oil supply are examined. Evidence of causality and, therefore, of a speculative supply shock is detected, resulting in the development of a structural model of the global crude oil market that allows for a speculative supply shock as a result of a forward curve shift and steepness. The second empirical chapter investigates the dynamics between price and cross-border trade flows of the EU electricity market. First, I study the impact of the relative strength of economic activity and distance between two countries on their net cross-border electricity flow. The effect from changes of electricity flow between two markets on flows between another pair of markets is also examined. Lastly, I investigate the relationship between crossborder electricity flows and electricity prices. Evidence of causality between flow and price, flow and flow and the gravity of the trade coefficient and flow, is also documented. VAR/VEC model framework is employed to identify and trace the shocks introduced to the system of inter-connected markets, a short-term electricity trading model is proposed. The third empirical chapter of the thesis examines the role of global foreign exchange markets in the formation of supply and demand shocks for key energy, metal, grain, and shipping commodity markets. Difference in the predictive power of the currencies of exporters and importers is investigated and an S&D model based exclusively on foreign exchange signals is proposed. The results provide evidence that currencies of importers have higher explanatory power than the currencies of exporters - a major departure from the established consensus in the literature. Additionally, the currency-based S&D model is found to possess a stronger predictive power over the price of commodity compared to the predictive power of each of its constituents, which improves the explanatory power of the proposed VEC model.