Could modeling save the oil industry? An oversimplification perhaps, but uncertainty and risk expert Sam Savage and ORTEC’s John Poppelaars argue that it’s the companies who embrace today’s more sophisticated decision-making tools that will do best in the future.
Oil exploration is more difficult than ever. Most low-hanging fruit was picked long ago. Governments increasingly shun large oil companies, using local companies for exploration. This is forcing the majors towards fields requiring more money, technology and innovation to explore, increasing risk exposure in their portfolios.
One recent study shows many E&P projects fail to deliver promises on schedule, cost and first production year. A trend we expect to continue or increase as projects become more complex.
The answer? Better modeling capabilities that improve risk based decision-making and provide consistent evaluation, communication and aggregation of key uncertainties across projects. Facilitating high quality, evidence-based decisions that underpin sound growth strategies. Or, as former Shell CEO Jeroen van der Veer puts it, “We need more brains per barrel.”
Challenges of a changing landscape
For most major oil companies, there is an increasing need to renew their portfolio as previously comfortable production levels flatten. Heralding large investment plans and moves towards riskier ventures like deep sea or arctic exploration. Increased demand for natural gas and trends towards more fuel efficient and hybrid/electric cars will further reduce demand for oil over time, but at what pace is uncertain. While political turmoil in the Middle East, concerns about new environmental accidents, uncertain supply from places like Brazil, Russia and Africa, plus a sense of diminishing oil resources all increase the uncertainty of supply.
Moreover, as the size of oil projects grow, increasingly they are pursued through joint ventures requiring decision-making processes able to deal with multi-stakeholder issues. Companies need dynamic strategy development and investment capabilities, including portfolio optimization and integrated option planning: they simply can’t afford to bet on just one possible outcome. But existing operating models, investment strategies and optimization practices can’t cope with this changing landscape. Why not?
Oil exploration is more difficult than ever. Most low-hanging fruit was picked long ago. Governments increasingly shun large oil companies, using
Inadequate old models
Historically, Exploration & Production (E&P) project portfolio choices have been based on key metrics such as average/expected net present value (NPV), hydrocarbons and capital expenditure. These were then ranked by NPV and selected from the top until the budget was exhausted. Thus boiling the uncertainty of each project down to a single average number for each metric, then adding them up to estimate the total average NPV, hydrocarbons, etc. Easy; but it masks the uncertainty and risk resulting from portfolio effects, since it assumes projects have no common factors.
Recent practices better, but…
In the past 20 years oil companies have started using methods that account for uncertainties in investment decision-making. For example, What-If analyses, where a project’s NPV is evaluated under a set of predefined conditions. But while these can help evaluate specific scenarios, they don’t look at uncertainty holistically.
Probabilistic analysis of key variables like production curves, operational capacity or CAPEX/OPEX project components is also widely used. Though more advanced, it is easy to get buried in numbers and endless assumptions. And since it only focuses on a few variables, you still don’t know which uncertainties are the real value-drivers.
Essentially, single average estimates of future outcomes lead to the “flaw of averages.” Considered in isolation, you would never buy fire insurance because on average it loses money. It is only a wise investment if you have a house in your portfolio to go with it.
In E&P, portfolio effects are driven by uncertainties in global prices, markets and geopolitical events. And recent methods of risk aggregation capture the portfolio effects of those uncertainties to allow for risk-return optimization. While interactive graphical interfaces make results far easier to understand.
More computer grunt
In studying the performance of 20 oil companies active in the North Sea, Aberdeen University proved a strong correlation between decision-making sophistication and overall company performance.
For decades Wall Street has used models based on the work of Nobel Laureate Harry Markowitz that optimize risk return tradeoffs using a covariance matrix based on historical stock behavior to model interrelationships between assets. But because no covariance matrix exists for E&P project portfolios, structural economic models must be applied, requiring computer horsepower only recently available.
Today we can optimize general models of E&P portfolios using scenario and robust optimization . Giving oil companies an instrument that makes the profitability of new ventures and the company clearer. It also helps create a quantitative view of risk exposure. Leading to fact-based decisions rather than hunches; better asset/budget allocation (increasing the predictability of future revenues); short-term competitive advantage; and ultimately increased long-term shareholder value. In short, growth through more brains per barrel.
Dr. Sam L. Savage is Executive Director of Probability Management.org, Consulting Professor at Stanford University and Fellow of the Judge Business School at Cambridge University. He is author of The Flaw of Averages: why we underestimate risk in the face of uncertainty and consults extensively to industry on uncertainty and risk. He has a PhD from Yale University in computational complexity.