Applied and Efficient Modeling in Natural Resources: Case Studies of Mining and Oil and Gas
Commercial interests in natural resources present particular challenges to the modeler, such as price volatility, weather-dependent demand cycles, changing regulations, technological developments, and many other factors. The complexity of modeling these challenges can be daunting, and in fact, undue attention to getting the model “right” can overwhelm the usefulness of the final product. Good model design is instrumental to building simple models of complex systems that can deliver immediate and valuable insight to decision makers.
Good model design allows translating the essential needs of the decision-maker into an efficient and effective tool. The design of the model will determine the time required building it and running it, but also how data-intensive the model will be. Model design also has implications for risk management and decision making.
In this talk, we discuss several case studies from our client work where we had to strike a balance between model realism and structural simplicity. We will emphasize moments when we were able to exploit realistic assumptions or fundamental statistical theory to simplify the model and still obtain robust results. We will also discuss commonly-seen mistakes that can result in systematically misleading models.