The Oilsand Producer’s Dilemma

Perhaps the most remarkable thing about Deloitte’s game the-ory analysis is that we’re even talking about it at all. Game theory has always been a rather inaccessible and misunderstood phenomenon. For decades, it was the purview of esoteric academics and the business and political elites who employed them to figure out, for example, how to grow their businesses in competitive markets, or how to fend off nuclear annihilation (or at least how to win the war once the missiles started flying).

Conceived by Hungarian mathematician John van Neumann in the 1930s and ’40s, game theory is a rather complex branch of applied mathematics originally developed to help understand the behaviours of firms, markets and consumers. Over time, it has been adopted by analytically inclined academics in a variety of fields, from political science and evolution to psychology and philosophy.

It gets its rather fun-sounding name from the fact that it describes, mathematically, various scenarios (or “games”) that involve two or more “players,” each with an array of strategic choices (or “moves”) they will make to reach some preferred outcome (usually to “win,” often as defined by maximizing economic utility). It sounds complicated, and for most of us it is, especially when there are large numbers of players and moves, which can provoke millions upon millions of potential outcomes.

The advent of fast computers and increasingly sophisticated software programs has made game theory application, like video games, more accessible and increasingly popular, especially with corporate executives charged with making multi-billion dollar investment decisions for their shareholders. In fact, something of a boutique game-theory industry seems to have grown up to help the world’s biggest corporations make better strategic decisions.

At least that’s the rumour. Until Deloitte released “The Producers’ Dilemma” in late 2007, it was virtually impossible to find a real-world example of how a corporation used game theory to inform its risk-management and decision-making strategies. In 2005, Fast Company scoured 40 of the most promising journals and assembled 30 of the world’s most respected game theorists, looking for examples of game theory being applied on the playing field of big business. It found nothing.

The reason? The game theory process involves a range of strategic details that corporate decision-makers are not interested in sharing publicly. “We’re dealing with the very highest level of the company,” says Gerry Sullivan, the mathematically inclined chemical engineer turned game theory analyst who worked on the Deloitte project. “And all of our work is very confidential.”
Sullivan’s company, Priiva Consulting Corporation, has applied game theory to about 200 real-world business cases, the results of which will never be seen by more than a few sets of very classified eyes. A former professor at the University of Waterloo, he now travels all over the world, using game theory to help Fortune 500 companies make better decisions about mergers and acquisitions, competitive positioning, alliance development, labour negotiations, government regulations, litigation and environmental issues.

“It’s a lot of fun,” says the amenable Sullivan, as if he were talking about chess or Monopoly rather than a multifaceted mathematical approximation of economic reality. “It’s what I really, really enjoy doing.”

Gathering the information necessary to run the analysis must have been rather fun. The whole thing was a simulation. Cooper put together a team that included Sullivan, some of his Deloitte colleagues, Brant Sangster, former senior vice-president of oilsands for Petro-Canada, and an assortment of other industry insiders. Together, they spent two days pretending they were the executive leadership of the major oilsands producers. They used publicly available information and their collective experience to make some educated guesses about what the options and preferences of, say, Suncor’s Rick George were likely to be as he (and his counterparts at EnCana, Canadian Natural Resources, Shell, Imperial Oil and Petro-Canada) made long-term, big-budget investment decisions. Sullivan then plugged this information into his proprietary software and ran the game-theory analysis.

The industry players’ major options were to continue to invest in oilsands production, to delay investment or to invest in Alberta-based “value adds” like upgraders and refineries. The major options for both levels of government were to impose heavier legislation, maintain lighter legislation, increase royalties and/or introduce incentives to encourage certain kinds of development.

“Then we figured out what’s the most important thing any player could do, from the context of that player,” says Sullivan slowly, aware that the complexities of game theory are often best explained using hypothetical examples. “So, for example, say we look at company ‘A.’ Of all the things that any of the players could do, what’s the most important one for company ‘A.’ Well, maybe it’s not what company ‘A’ can do itself, maybe it’s that one of their competitors could invest heavily in another project that would compete with them. Maybe they don’t want it to happen, so that appears at the top of their preference tree. Then we rank all those options for each player. From that we can presume which outcomes they’d like and which they wouldn’t like, and whether or not they can do anything about it. All of that gets inputted into our proprietary game theory software that allows us to predict where the stable outcomes are.”

In the end, the analysis generated a whopping 33 million potential outcomes. Most of them, however, were either redundant or
impossible, so they were discarded. The remaining outcomes were grouped into two classes, which represented the two most likely scenarios for the future of the oilsands industry. The first is the “natural” outcome, what Cooper has dubbed “maintaining pace.” Left unchecked, oilsands producers will continue to invest in projects at a pace that the model suggests will make the industry economically unsustainable.

“The natural market forces are coming into play around the supply and demand for labour, the challenges on the infrastructure, the costs of excavation, the challenges on the steel, and transportation,” says Cooper, “and that will have its own governing effect on the pace of development.”

The paradox is that the implications of the “natural” scenario, like the prisoner’s dilemma, are less than ideal for the producers themselves. This “gold-rush mentality” could dramatically increase building costs (labour, contractors, construction material such as steel) such that projects face cost overruns and serious delays.

The accompanying social issues – housing, inflation, crowded roads and hospitals – could create a public uproar that will force the provincial government to intervene with regulations and/or royalty increases that will slow the pace of development. Add to that public concern about the environmental implications of the oilsands and the long-term future begins to look rather gloomy for investors.

This is where the oilsands situation parallels the “prisoner’s dilemma.” Like the two prisoners, sequestered away in their separate cells, each conspiring to act rationally and in their own perceived self-interest, the game theory model predicts that oilsands producers will behave in ways that, paradoxically, will result in a less desirable outcome than could otherwise have been achieved had they worked together to slow the pace of development.

But that, says Cooper, is the cost of doing business in the free market – and it’s better than the alternative. “In some parts of the world there’s a government-driven economy, where they make choices for the industry around who does what and first,” he says, sitting up a little straighter in his chair, the timbre of his otherwise measured voice loudening to drive the point home. “That is not the economy we live in nor do we want to live in. This is not a managed economy. You make investments in this marketplace based on what’s likely to transpire without being concerned about someone else managing your choices.”

Unlike the doomed prisoners, however, who are separated by concrete walls and iron bars, and so cannot conspire together to implement the best strategy, there is another option for oilsands producers. This is the second, more stable scenario, what Cooper calls “maintaining peace.” This is the “best achievable” outcome, which Sullivan says the players can get to only by “judicious choice.”

In this scenario, the industry as a whole would have to agree to slow the pace of oilsands production, and the governments would have to provide the industrial players with other investment opportunities, say, by implementing incentives that encourage the construction of upgrading and refining facilities in Alberta.

Despite a history of collaboration among oilsands producers, self-directed collaboration is unlikely to happen. “Co-operation on strategic issues is difficult to achieve because, well, there are penalties for delay of game,” says Cooper. “There is an expectation that you will operate in the best interests of the shareholders. You’re not rewarded for delaying strategic investments that could yield return on investment.”

Although the report didn’t come out until after the provincial government raised royalties on oil and gas, the analysis actually had been completed by August, two months before the announcement was made. This and a second announcement – made four months later by Suncor, which committed another $20 billion to increase bitumen production and build an upgrader to convert that bitumen into higher-value crude oil – seem to corroborate two of “The Producers’ Dilemma’s” conclusions: one, that government was going to be increasingly “forced” to intervene in the regulation of the increasingly lucrative and economically strategic oilsands industry, and two, that until either intervention or market forces substantially increased the costs of production, oilsands producers would continue to invest at levels that will likely become unsustainable.

These announcements, say both Cooper and Sullivan, validate both the integrity of the results and the value of using game theory as a predictive tool that corporate executives can use to help them manage risk and make strategic decisions. Deloitte’s analysis looked even more prescient when, in the midst of the provincial election campaign in February, a coalition of industrial, environmental and government concerns known as the Cumulative Environmental Management Association issued a letter to the Alberta government calling for a moratorium on land-lease sales through 2011. CEMA’s stated objective was to conserve dwindling wild lands in the area south of Fort McMurray, but the initiative – backed by oilsands incumbents Suncor, Imperial Oil, Petro-Canada, Husky Energy and the Canadian divisions of Royal Dutch Shell, Devon Energy and ConocoPhillips – would also effectively slam the gate to further competition for labour and materials. It is this kind of collaboration that could lead to the more stable outcome predicted by the Deloitte analysis.

Others aren’t convinced of game theory’s predictive power, though. While it can be a “useful way of putting a framework around the situation the industry is looking at,” says Alan MacFadyen, professor emeritus at the University of Calgary and a expert in the economics of the oil industry, the outcomes shouldn’t come as a surprise. “You almost don’t need game theory to understand that result. It’s just logical.”

The real value of game theory may have to wait until round two. Cooper and Sullivan are going to run a second analysis, and this time it won’t be a game. Instead, it will be based on the actual preferences of the major industry and government players, who will provide their input under the protection of anonymity. And while the details of the analysis – players, options, choices and preferences – will be kept confidential, the predicted outcomes will be made public in May. Let the games begin.

Current Issue