Hedging and optimization of energy asset portfolios

Abstract

This thesis includes three papers on hedging and optimization of energy asset portfolios. The regulatory scheme for natural gas (NG) prices in Mexico is described and the behavior of international and domestic gas prices and the peso-dollar exchange rate from January 2012 to June 2017 is analyzed. Statistical analysis reveals that volatility in the daily growth rate of international NG prices exceeds daily fluctuations in the exchange rate. Based on this knowledge, the behavior of First-Hand Sales prices is modeled, and two price hedging strategies are proposed, one through futures and the other through swaps. Given how First-Hand Sales prices are calculated, the optimal futures hedge should consider the acquisition of gas futures one and two months prior as well as contemporary exchange rate futures. Based on a hedging strategy that includes NG futures and using an MGARCH VCC (MGARCH stands for Multivariate Generalized Autoregressive Conditional Heteroskedasticity and VCC for Variable Conditional Correlation) model, conditional variances were estimated with lags of 20 and 40 days between the prices of NG Futures. Dynamic hedges of NG were calculated assuming theoretical futures prices of the US dollar in Mexican pesos. By applying backtesting, it was found that the forecasts of optimal hedge ratios improve with short prediction periods and proximate observed data. The dynamic hedging model proposed can be extended to other fuel markets. The importance of hedging NG prices derives from the size of the market and the extent of the risks to which the market participants are exposed. Using the share price data of six energy companies of Latin America and other regions and two crude oil futures, this thesis proposes the integration of hedging portfolios and the calculation of efficient frontiers under different risk measures. The original financial series are transformed into new ones to improve the risk measurement. With the new series obtained through simulation with the support of the Extreme Value Theory and t-copulas, different conditional risk measures are calculated. These conditional risk measures are used to solve the hedging and optimization problems. Non-linear integer programming techniques are used to obtain these solutions. The programming codes used to generate the new series and solve the hedging and optimization problems are presented in the annexes. Due to the economic value and the volatility of energy markets, hedging strategies and portfolio optimization are useful tools to reduce non-desired levels of risk or to avoid unnecessary costs.