Evaluating the dynamic performance of energy portfolios: Empirical evidence from the DEA directional distance function

In recent years, the complex global energy commodity market has led to increased uncertainty for energy investment returns and tremendous challenges for investors to design appropriate energy portfolios. Therefore, we employ four popular portfolio methods to determine energy portfolios based on daily fossil-fuel futures prices during 2006-2015. Moreover, we use the DEA window analysis method and DEA directional distance function to comprehensively evaluate the dynamic performance of these energy portfolios, based on the efficiency perspective. The empirical results indicate that, first, to increase investment returns, the mean-variance method that exclusively emphasizes the maximization of returns shows the best performance; however, in order to decrease the volatility and risk of investment returns, the mean-variance method that aims to resist risk more than make profits and the bootstrap-historical simulation Value-at-Risk (VaR) method have the best performance. Second, when the global financial crisis broke out in the second half of 2008 and energy markets experienced a sharp downturn in the second half of 2014, the DEA efficiency of energy portfolios appeared relatively lower than those of their neighboring periods, regardless of which portfolio method is employed. Finally, the average DEA efficiencies of energy portfolios are all higher than those of single-energy investment throughout the sample period, except when the equally weighted method and the information entropy-comprehensive index method are used; meanwhile, the mean-variance method that prefers profit-making to risk-resisting possesses the highest average DEA efficiency among various methods concerned.