Grants and Contributions:

Grant or Award spanning more than one fiscal year. (2017-2018 to 2022-2023)

Real option analysis (ROA) is recognized as a superior method to quantify the value of real-world investment opportunities where managerial flexibility can influence their worth, as compared to standard discounted cash-flow methods typically used in industry. A comprehensive ROA of an oil, gas or mineral mining project can improve the allocation of capital and managerial decision making and the methodology is currently used, to some degree, in the commodity extraction sectors. However, realistic models that try to account for a number of risk factors can be mathematically complex, and in situations where many future outcomes are possible, many layers of analysis may be required. Typically, managers are usually unable to understand the models and dismiss results that seem unintuitive to them. A recent empirical study showed that roughly 35% more value can be extracted from mining operations if managers had made better timing and capacity decisions.

The focus of this proposal is the development of an “intelligent” real options valuation (IROV) methodology geared towards practical use with specific emphasis on mining finance. The overall objective of the proposed research is to develop a methodology and tool that allows managers and engineers to apply ROA in complex, real world settings in the context of the valuation of mining projects. The methodology will be based on sound financial principles utilizing ROA, and will account for both systematic risks associated with economic factors, as well as idiosyncratic risks associated with mining, such as size of deposit and ease (cost) of extraction. While the methodology must account for real world complexities, it must be general and broadly applicable. Finally, the methodology must be such that its fundamentals are easily understood by managers and engineers so that its use is readily adopted.

A key innovation of the model being proposed is the idea of fitting optimal decision making boundaries to optimize the expected value, based on simulated stochastic processes that represent important uncertain factors associated with the mine. Ultimately, the model will consist of hyper-planes that can be used by managers to decide when to build, expand, mothball or abandon. Two approaches to determine these boundaries are proposed: 1) utilizing standard curve fitting and optimization techniques, and 2) utilizing machine learning where an algorithm is utilized to optimize managerial decision making, in the same way as algorithms are utilized to challenge opponents in games such as chess. To the best of my knowledge, the methodology proposed here has not been presented and may provide a key aspect in the practical implementation of ROA. From a training perspective, the research is multi-disciplinary and will provide trainees the opportunity to learn aspects of financial modelling, mining, data analysis, optimization and machine learning.