Portfolio Optimization: Systems vs. Sequential Approach

Abstract

Abstract Traditional investment decisions in the oil and gas industries ignore dependencies and interactions between parameters such as reserves, production and economics. In a previous paper1 we have shown that for a single project the system stochastic approach better captures the impact of dependencies and interactions among the components of decision parameters compared to the sequential approach and improves decision quality. In this paper we build on the previous work and investigate the impact of system stochastic and sequential stochastic modeling on the portfolio efficient frontier. We use a stochastic integrated model, which captures uncertainty for a mix of five hypothetical offshore oil field development projects. The model combines reserves, production, capital and operating costs into an integrated probabilistic economic evaluation. The output of the stochastic model is used as input into the Markowitz Mean-Variance model and varying our company's participation level in the five projects to compute the efficient frontier. The results show that the two approaches yield substantially different efficient frontiers. As expected the difference between the systems and sequential approach increases with increasing standard deviation of the portfolio NPV with the expected value of portfolio being greater in the systems approach. In our example, the difference in the computed expected NPV ranged from 8 % at lower level of risk to 15% at higher level. In conclusion, the system approach is superior in capturing intra-project dependence and its impact on the portfolio level is significant compared to the sequential approach. Both the systems approach and the portfolio methodology complement each other in capturing both intra-project and inter-project dependence at the portfolio level.