Meet the Expert
Michael Walls PhD
Professor and Director - Division of Economics and Business, Colorado School of Mines
- Professor and Director of the Division of Economics and Business at the Colorado School of Mines specializing in the areas of strategic decision making, business strategy, and risk management, with a particular emphasis on applications in the petroleum sector.
- Main research interests are in the areas of petroleum valuation, corporate risk management, and the integration of decision analysis and portfolio management in the corporate context.
- Relying on technical developments in the areas of decision analysis, finance and business strategy. provides clients with the ability to improve their decision quality and create value.
- Clients include: Amoco, Anadarko Petroleum, BP Exploration, Cabot Oil and Gas, ExxonMobil, Hess Corporation, Occidental Petroleum, Penn-Virginia Oil Company, Petrobras Petroleos Brasiliero, Phillips Petroleum, Schlumberger, Texaco
Meeting Packages from $600
Your Meeting Package Includes:
- All 9 Best Practices
- Pre-Meeting Discovery Process
- One-on-One Call with Expert
- Meeting Summary Report
- Post-Meeting Engagement
Risk Management and Strategic Decision Making for the Petroleum Industry
Professor and Director - Division of Economics and Business, Colorado School of Mines
Common Problems
- An established culture of deterministic rather than probabilistic thinking can mask uncertainties.
- Traditionally, oil and gas managers tend to think deterministically instead of probabilistically. As a result, their analysis does not account for the high degree of uncertainty that accompanies each oil and gas venture. It's never a "sure thing." Each new project is always a bit of a gamble.
As they analyze any given project, they look for the potential IRR (internal rate of return) or the NPV (net present value). Neither is easy to establish when there are so many unknown variables in the mix. The pay zone, the size of the reservoir, the level of daily production, the product price – there are so many uncertainties that must be taken into account when a company is looking at a project and evaluating whether or not to invest in it.
The problem with the single-parameter (deterministic) analysis is that it tends to mask the uncertainties. There's nothing to be gained by that and much that can be lost. It’s much better to think probabilistically, and discover whatever can be known or at least predicted with some statistical confidence. This offers a better way to compare each opportunity and see which package is actually likely to be best for the company.
A deterministic analysis might deliver the expectation that a project would make $60 million. A probabilistic analysis will look at the same opportunity and deliver more detail. For instance, it could show that while the expectation is $60 million, there's a potential upside of $100 million and a potential loss of $30 million. There's no way to be certain as to what will actually happen. Through probabilistic thinking, however, it's possible to get a better picture of what could happen and the difference might be crucial. A small firm might be enticed by the $60 million expectation and the $100 million upside but that $30 downside might constitute too much risk. It may make more sense for that particular company to obtain a joint venture partner or walk away from the deal. - Firms fail to systematically assess relevant uncertainties and risks as they make strategic decisions.
- Oil and gas companies are very data driven. On any given project, their risk assessment is driven by the data that they have available to them. In the context of drilling wells, a company often has a particular area where they like to drill. They will tend to make decisions based on the offset data they have from the surrounding wells they already have there, along with their seismic data. By incorporating a basic statistical analysis, however, they can actually quantify the risks and uncertainties associated with the data they have.
There's a subjective side to this. Companies don't always have much data on which to base their estimates. In those cases, there are still ways to encode their prior knowledge into a probability distribution. This is a statistical technique that can bring more clarity to the assessment. There are many ways the data can be processed or manipulated in order to get to a more accurate answer. Risk assessment is very important, but you can't do good risk management unless you've got good input on the assessment side.
For companies operating in the international environment, there are added risks. There’s risk on the exploration and production side, and there's political risk like the risk of nationalization, for instance. If you look at Exxon right now, they made a big investment in Russia, so all the stress between our two countries has got to have them worried. It's very hard to quantify and predict this sort of risk, but companies certainly can (and must) make an effort at it, at least through scenario analysis and other methods. - Companies evaluate each venture on a stand-alone basis rather than considering how the project might affect the firm’s overall portfolio.
- Companies often do their project evaluation on a one-off basis. This is not ideal. Even if a company is doing its analysis probabilistically, it’s important to consider each project within the context of how that project may affect the firm's overall portfolio of activities.
Most of us know that there should be a mix of different stocks in order to create some diversification in a stock portfolio and that this helps to mitigate risk. The same holds true for oil and gas firms. Even when all their assets are in oil and gas projects, they still can achieve significant diversification. The percentage of oil projects versus gas projects will have a big impact on the overall risk and return. Since oil and gas prices do not move together but are independent, this allows for diversification. Companies can also create diversification though a mix of exploration versus development, and domestic versus international. It’s important for firms to look at their whole mix of activities and investments, and do the same sort of portfolio analysis they would do with a stock portfolio.
This also applies to capital asset evaluation. Some companies are doing this more often now, while others are doing what they like to describe as portfolio analysis but the process is actually something else. - There is often poor communication about risk and risk-taking among the various managers within the firm.
- Many firms lack a common language with which to communicate about risk and risk-taking. This is sometimes related to them not thinking probabilistically, but it can also go beyond that issue. Even within the firms that do a considerable amount of probabilistic risk analysis, the different divisions often do not know how to communicate with one another about risk.
Managers are often looking for guidance from senior management as to how much risk they should be taking. If those senior managers don't communicate to their team in a pretty straightforward way, it can lead to some very inconsistent risk-taking behaviors. One of the benefits of probabilistic analysis is that it establishes a common language and an analytical process that the entire company can follow. Firms that start putting these techniques into play see their communication across divisions automatically improve. Their executives are forced to talk in a more rational way.
It is not necessary that everyone participating in the process understand all its intricacies and inner workings. This can be an arena where it’s not the senior managers but their direct reports who are likely to be more sophisticated. Some senior managers resist the use of these more advanced decision-making techniques simply because they don't understand how they work. That's unfortunate, because it's really not how they work that's most important. It's the fact that they do work. - A lack of consistency in risk-taking can lead to a poor performance overall.
Managers will sometimes risk a huge amount in one project, and then suddenly become more risk averse with their next project.
There is much to be gained by being more consistent. When a firm is consistent in its risk-taking, then it can look back and say, “Okay, so this is the sort of tolerance for risk that we utilized this year. Let’s see how we did. Maybe we should increase our tolerance next year, or maybe we should decrease it.”
If the firm has no a mechanism to measure how much risk they're willing to take, it's very difficult to at the end of the year to make any sense of what happened. There are a couple of ways to establish a company’s risk tolerance.- Historic Risk Tolerance Analysis. This looks at the last three years of decisions, reconstructs the opportunities and what kind of risk tolerance the company had when they decided to participate. While this sounds good, it's really not as informative as one might expect. Usually the data is muddled and there are often elements that came into play that were not related to the risk-taking. For instance, the project may have been in a geographic area the company simply wanted to enter and may have been willing to overpay for the lease position in order to do that. Or they may have decided to get into fracking for the first time. Such decisions may not reflect their true tolerance for financial risk. Historic risk tolerance analysis is therefore only moderately accurate.
- A Decision-Science Questionnaire. The decision makers at a company answer a simple questionnaire that provides them with a set of projects consistent with what they have in their current inventory and asks them to make choices about them. Because the questionnaire is designed to offer situations very similar to those the executives actually deal with on a day-to-day basis, the process tends to produce a better assessment as to what kind of financial risks the company is willing to take.
- Historic Risk Tolerance Analysis. This looks at the last three years of decisions, reconstructs the opportunities and what kind of risk tolerance the company had when they decided to participate. While this sounds good, it's really not as informative as one might expect. Usually the data is muddled and there are often elements that came into play that were not related to the risk-taking. For instance, the project may have been in a geographic area the company simply wanted to enter and may have been willing to overpay for the lease position in order to do that. Or they may have decided to get into fracking for the first time. Such decisions may not reflect their true tolerance for financial risk. Historic risk tolerance analysis is therefore only moderately accurate.
- Failure to analyze the value of information in a systematic way can lead to the purchase of unnecessary or overly expensive data.
- Companies in the upstream end of the business are often looking to buying seismic data. This basically maps the sub-surface and helps managers determine where to drill. Not surprisingly, this information can be very expensive. You can easily spend $5-10 million on a couple lines of data.
This data is never perfect. It can never say for certain what's going to happen. Much depends on the interpretation and the quality of the data. This can vary widely but there are ways to compute the expected benefit of the information and once this is known, it's much easier to ascertain the value of the information. This sort of analysis is not new. It's used by a lot of companies, but given the benefit, it's surprising that it's not universally used.
Larger companies can spend hundreds of millions of dollars on information. While there’s a tendency for scientists to think more information is always better, this is not necessarily true. Having more information might make the decision maker feel better, and it can create a delay in making a tough decision. The bottom line is the value of the information. Is it quality data? What is it really worth? - Failure to understand the value of portfolio analysis can lead to inconsistent capital allocation.
- "Portfolio analysis" is a term that gets bantered around a lot. Some companies are actually doing it, while others are just ranking projects based on a variety of different metrics and are not using the right methodology. They are not capturing the tendencies and correlations. Proper portfolio analysis means looking at the appropriate tradeoffs between risk and return and identifying the inner dependencies between the assets in a way that leads to the highest level of diversification for a particular level of return. When it comes to allocating capital, it’s important to look at the mix of assets and where money can best be allocated going forward. The impact can be huge in either direction – a big gain or a big loss.
There are companies that know how to do this. Anadarko and Chevron Texaco are very good at it. But even some of the biggest companies don't apply this technique in the correct manner. A company that is not probabilistic at the outset will be a long way from implementing proper portfolio analysis. It can take a while to get to that level of sophistication. - Failure to adopt a common methodology for assessing risk can lead to an isolated bad decision that affects the entire organization.
- Many companies are broken into silos that operate as fairly independent business units. Often, one hand doesn’t know what the other is doing and yet these units can ultimately have a big impact on each other. If one business unit makes a bad decision, it can seriously affect the financial stability of the entire corporation.
We’ve seen this at Exxon and BP, where individual units took risks that led to big oil spills. Their bad decisions ultimately had a significant effect on the company’s bottom line. It makes sense for a corporate entity to have a system in place that fosters probabilistic decision making and better inter-departmental communication. - Failure to integrate the quality of risk assessment into the firm’s post-audit review makes it hard for a company to learn from its mistakes.
- It’s a good idea to incorporate the risk assessment process into the post-audit review. This allows a company to assess how well they were able to predict the uncertainties. If the geo-scientists project the reserves on a well to be somewhere between 300,000 and 1 million barrels of oil, but then the oil comes in at just 100,000 barrels, there's clearly a need to evaluate the assessment they did and see what went wrong.
In the oil and gas industry, however, companies very often fail to do this. If they did such an assessment, they could give helpful feedback to the analysts and geo-scientists and engineers. They could make sure that everyone learned from whatever didn’t work out and continuously improve their assessment process. Some firms are doing this. Some even have, for lack of a better term, a team of “risk police” who go from division to division, business unit to business unit, to encourage a consistent risk assessment effort.
In the industry in general, however, this is not yet done and would be hard to implement. It would mean making a fundamental change in the corporate culture. - Firms only reward good outcomes when they should also reward good decision-making practices.
- It is entirely possible for a manager to make a successful decision that's not the result of good decision-making techniques. By the same token, a manager could make a very good decision that is not ultimately a successful one. If a manager makes the decision to participate in a venture based on a full range of analysis and after using the most sophisticated available approach to the data, but then it turns out that the oil or gas just isn't there, it doesn’t mean that the decision itself was a bad one. There's been an undesired outcome but the decision-making process itself has been good.
In the long run, studies show a quality decision-making process will produce a better overall performance. Companies are likewise making the decision process part of the performance metrics for their managers, regardless of the outcome.
Risk Management and Strategic Decision Making for the Petroleum Industry:
Common Problems
Expert Topic