Essays on International Asset Portfolios and Commodities Trade

Thesis advisor: Christopher BaumThesis advisor: Fabio GhironiDo events in the natural gas market cause repercussions in the crude oil market? In light of the enormous impact that price movements in the two largest U.S. energy markets have on the economy, it is important to understand not just the individual markets but also how they relate to one another. On this front, the literature presents a puzzle: while economic theory suggests that the oil and gas markets are interlinked through a bi-directional causal relationship, empirical research has concluded that the oil market affects the gas market but not vice versa. The first chapter of this dissertation improves on the previous studies in two ways: by using high-frequency, intraday oil and gas futures prices and by analyzing the effect of specific news announcements from the weekly oil and gas inventory reports. The results dispel the notion of one-way causality and provide support for the theory. The reaction of the futures volatility and returns is asymmetric, although this asymmetry does not follow the "good news" vs. "bad news" pattern from stock and bond markets; the response depends on whether the shock is driven by oil or gas inventory gluts or shortages. The two-way causality holds not only for the nearby futures contract but also for contracts of longer maturities. These findings underscore the importance of analyzing financial markets in a multi-market context. The second chapter of this dissertation asks whether volatility and trading volume evolve in a unidirectional or bidirectional, contemporaneous or lagged relationship in the crude oil and natural gas futures markets. This question is important because it affects trading and government regulation but previous studies have come to conflicting conclusions. Their main shortcoming is the low frequency of data used in the analysis. This chapter improves on the previous studies in three ways: by using high-frequency, intraday oil and gas futures prices and volume, by including trading not only during the day but also during the night, and by analyzing not only the nearby futures contract but also contracts with longer maturities. For the nearby contract, Granger-causality tests show that past values of volume help explain volatility which agrees with the Sequential Information Arrival Hypothesis. Past values of volatility have explanatory power for volume only when absolute return is used as the volatility measure; when the conditional variance from GARCH models is used as the volatility measure, the causality in this direction disappears. These results change when low-frequency daily data is applied. It is also shown that the volatility-volume relationship differs for contracts with longer maturities. These findings are relevant for regulations, such as trader position limits recently adopted by the U.S. Commodity Futures Trade Commission. The third chapter of this dissertation investigates whether the production structure of firms affects international optimal portfolios, risk-sharing, and response of terms of trade (TOT) to shocks. The answer to this question would enhance our understanding of the home equity bias, yet it has not been addressed in the theoretical literature. This chapter studies the question in a two-country dynamic stochastic general equilibrium model with endogenous portfolio allocation. It shows that the optimal portfolio includes more home equity as the production structure changes from exporter-only, i.e., firms operating in their home countries and serving foreign markets by exports, to multi-national-company-extreme (MNC), i.e., firms hiring labor in both countries and producing locally in both countries. This shift occurs because changing the firms' production structure eliminates exposure to technology differences and allows the home household to accomplish the same diversification with less foreign equity. The production structure also has implications for the effect of technology shocks on the TOT. Under the exporter-only setup, a shock to technology causes a standard TOT deterioration, whereas under the MNC-extreme setup, a shock to technology leads to a TOT improvement. By producing testable predictions, this chapter underscores the need to take firms' production structure into account when analyzing international optimal portfolios, risk sharing, and response of the TOT to technology shocks. This is especially important since empirical research has generated conflicting results.Thesis (PhD) — Boston College, 2012.Submitted to: Boston College. Graduate School of Arts and Sciences.Discipline: Economics