Dynamic Risk Management

Mastering the "Missing Link" in infrastructure Finance.

There is a paradox at the heart of investing in infrastructure. On the one hand, investors are typically attracted to infrastructure assets because they are seeking stable cash flows over long time horizons. On the other hand, greenfield infrastructure projects represent huge and often risky bets—bets that can go spectacularly bad.

It’s no wonder, then, that infrastructure funds in recent years have found it easier to come across interested investors than to unearth investments that suit their investment strategies —even as global infrastructure needs continue to outstrip the capacity of public sources to fund them.

Today, more than ever, infrastructure investors need tools to bridge the gap between their risk appetite and the actual level of inherent risks of projects requiring massive capital outlays against time-distant revenue streams. In Oliver Wyman’s work with large infrastructure projects, we have found that there are a number of tools for dynamic risk modeling that are often underused, but that can be a valuable resource for project sponsors, lenders and equity investors alike.

The Untapped Potential of Risk Management

Infrastructure projects, be they roads or rail lines, ports or airports, power lines or waterworks, all share certain characteristic features. These typically include:

  • High upfront investment requirements
  • “Chunky” capacity, with significant scale economies
  • Building ahead of demand (often uncertain or speculative demand)
  • Uncertain cost to create capacity
  • Uncertain timing of revenue
  • High leverage (typically 60-80 percent gearing)
  • Extraordinarily high sensitivity to financing costs

Numerous academic studies have come to the conclusion that greenfield infrastructure projects systematically disappoint their backers: cost overruns, schedule delays and revenue overestimates seem to be the norm more than the exception. It is no exaggeration to say that mastering risk—understanding, quantifying and managing it—is the key capability in successful infrastructure investment.

In this environment, sophisticated investors have learned to appreciate the value of dynamic financial modeling (e.g., Monte Carlo simulation) in assessing the likely performance of prospective investments. Unlike traditional static financial modeling, a stochastic risk model recognizes that key drivers of financial results (capital costs, operating costs, volumes, prices, timing of cash flows, etc.) are inherently uncertain and can interact in unexpected ways. Instead of assigning a discrete value to these variables in a spreadsheet, the Monte Carlo method models key variables in the form of a probability distribution function. This can be further extended to include the dynamic interactions between simulated outcomes of risks.

The output of such an analysis is a much richer view of the financial prospects of the investment. Instead of looking at, say, the results of three or four scenarios, a decision maker can see the consolidated results of thousands or tens of thousands of simulation runs. And while a traditional financial model might answer the question, “What is the sensitivity of cash flows to a 1 percent change in interest rates?”, it cannot reliably answer questions such as, “What is the probability that this project will meet its IRR target?” or “What is the probability that the project will remain in compliance with all its financial covenants?” The stochastic risk modeling approach, however, can answer those questions, which is one reason it has become the acknowledged gold standard for financial analysis of infrastructure investments.

In our experience, however, many project sponsors and investors do not capture the full value that stochastic risk modeling offers. Value is typically left on the table in two ways: the risk model itself may be faulty or incomplete, or the risk model is too often abandoned after the initial investment decision has been made.

The first major pitfall in dynamic risk management is getting the model wrong. When it comes to stochastic risk modeling, there is wisdom in the old adage that a little knowledge is a dangerous thing. The very precision of the outputs (“In 95 percent of the cases, the project will meet its IRR target”) can lead to a false sense of confidence if the appropriate care has not been taken in constructing the underlying model. The recent proliferation of easy-to-use spreadsheet add-ons such as Crystal Ball and @Risk may have encouraged a tendency toward overreliance on unreliable models.

There are many ways to go wrong in modeling risk (just ask anyone who invested in collateralized mortgage obligations), but one example serves to illustrate this point. Imagine a project in which the net present value is sensitive to two variables: the price of crude oil and the dollar exchange rate. With the help of a spreadsheet add-on and a few databases, it’s a simple exercise to generate a probability distribution function for both variables based on historical ranges. After running a Monte Carlo simulation of the project, the expected NPV of the project might look like the figure on the left in Exhibit 2 at left.

But this analysis implicitly assumed that the oil price and dollar exchange rate are independent of one another, when in fact they are correlated. After modifying our model to account for the correlation between the two variables, our expected NPV might look more like the figure on the right. What once appeared to be a sure thing is revealed to have a nontrivial chance of failure. Across many projects in diverse industries, Oliver Wyman has seen our belief confirmed that there is no substitute for a disciplined modeling approach, rigorously applied by skilled practitioners.

The second major pitfall in dynamic risk management is getting the model right, but not doing the right things with it. A common shortcoming is the disjunction between the risk analysis that goes into the concept, design, and finance phases and the risk management approach that guides the engineering, procurement, construction and operating phases.

Oliver Wyman’s approach to risk analytics looks at the variability of cash flow versus plan (“cash flow at risk”) as the primary metric. While this metric usually makes intuitive sense to project sponsors and investors, it stands in contrast to the engineering-driven approach to risk analytics that often prevails in a contracting and construction environment. To be sure, there can be value in the tools used in engineering-driven risk management, such as comprehensive risk registers, heat maps and the like. But this approach falls short of the needs of senior management. While notionally comprehensive, it fails to distinguish the merely important from the absolutely critical. And it leaves senior decision makers without the tools to understand potential trade-offs in risk and reward.

Dynamic Risk Management: Fewer Risks, More Rewards

Our experience has shown that investors and sponsors who incorporate a dynamic risk management approach can avoid these pitfalls and extract substantial additional value from their investment. The benefits of a more robust risk management approach are numerous, and accrue to infrastructure funders, operators and users alike.

Focuses on the right risks. The dynamic risk management framework gives management visibility into the impact of risks on the bottom line. In one recent engagement, the project sponsor intuited that the major risk to cash flow was demand risk, and was prepared to sacrifice substantial revenues to mitigate that risk through take-off agreements. Oliver Wyman’s risk analytics showed that risks related to internal execution were far more important, leading the client to devote more resources to those risks.

Supports a wide range of management decisions. Armed with the right analytical tools, management can compare and contrast the value created by investing in different risk mitigation measures for different risks. Funding strategies, hedging strategies, sourcing strategies and technology choices are among the tools that become more effective with a reliable understanding of cash flow at risk. In one recent example, we used a stochastic risk model to quantify a heretofore underappreciated supply risk. The client subsequently modified its technology strategy to focus on a more expensive, but more secure source of raw material.

Supports efficient allocation of risk. Infrastructure projects increasingly involve multiple investors and stakeholders, for example, through public-private partnerships and customer-supplier co-investment. Efficient allocation of risk can be a significant lever of value creation, not to mention a vehicle for making deals possible that might otherwise founder on stakeholder resistance. In a recent deal involving a major expansion to a transportation asset, risk analytics revealed that the infrastructure developer faced substantial exposure to steel price inflation—an exposure that could not be conveniently hedged. Faced with the prospect of paying for the steel risk through a price premium, the infrastructure users found it more efficient to accept the risk themselves, as they had some upside risk exposure to steel prices. The natural hedge was a win-win for the developer and the users.

Prioritizes value improvement opportunities. Dynamic risk management is not just about avoiding downside risks, but also enabling upside opportunities. By comparing multiple investments across the dimensions of risk and return, companies often can find “free lunch” opportunities: higher return for the same level of risk.

Lowers financing cost. Dynamic risk management is ultimately about making risk transparent—to sponsors, operators and investors. Bank regulators, following the capital adequacy standards in the Basel II and Basel III accords, are pushing lenders in the direction of greater reliance on dynamic risk evaluation. Project sponsors increasingly find that they need to have access to dynamic risk models to access the widest possible capital pool.
Based on concrete project experience, Oliver Wyman has identified a set of factors that underpin the success of risk management in infrastructure projects or, indeed, any large capital project:

• Adopt a cash flow at-risk framework, and apply it consistently throughout the project life cycle.
• Get the model right. Take a rigorous approach to constructing a pyramid of risks that describes the network of interrelated risk drivers.
• Calibrate the model carefully. Pay attention to the choice of statistical distributions, to the impact of “tail risks,” and the correlation of risks.
• Anchor the responsibility for risk management in the organization. Decision making processes and governance should adhere to the risk framework.

The universe of infrastructure investment opportunities grows larger every day. But comparatively few opportunities have the “ideal” risk profile investors seek. If every infrastructure investment had known capital costs, predictable revenues and stable margins, there would be no need for sophisticated risk management techniques. Until then, savvy sponsors and investors will need the best tools at their disposal to master risk.

Dynamic Risk Management