Print this article | Return to Article | Return to CFO.com
Applying Black-Scholes analysis to capital spending projects has one big flaw.
Ronald Fink, CFO Magazine
September 1, 2001
It's time to get real about real options. To be sure, this much- vaunted alternative to the conventional method of evaluating capital- spending decisions using net present value (NPV) is catching on with more and more senior finance executives. To one degree or another, CFOs and strategic planners at companies ranging from Anadarko and Enron to Times Mirror and Hewlett-Packard are self-confessed admirers of this theory, which posits that the evaluation of financial options can be applied to other investment decisions.
There's no question that many, if not all, of the potential investment decisions companies face include scenarios analogous to financial options. "We have portfolios [of real options] whether we recognize it or not," says Corey Billington, vice president of supply- chain services at HP. The issue, as Billington sees it, is how they're managed. "We need to use these capabilities better than our competitors," he says.
But real options work only if a company is truly prepared to cancel a project if the numbers look bad after the initial investment. The flaw in the theory is not its complexity, as some have said, but the fact that it ignores the psychological and political realities of capital investments.
Granted, the technique for valuing real options--based on the Black- Scholes method for modeling option prices--is more formidable than that used in NPV calculations, at least at first glance. With the latter approach to capital budget decisions, an appropriate discount rate is applied to a project's anticipated cash flows. If the resulting present value of those cash flows exceeds the cost of capital and the NPV of alternative investments, the company proceeds with the project. With real options, also called strategic options, a decision tree is plotted showing a series of different scenarios that could develop at various points throughout the life of the project. Then probabilities and discount rates are attached to each scenario and the cash flow is discounted back to the present. If the outcome is positive, the company makes at least a partial investment.
The most important difference between real options and traditional NPV is that while NPV analysis is basically an all-or-nothing approach, real options analysis lets a company make an initial investment and then proceed--or not--as the project develops. This "go/no go" flexibility enhances any project's potential appeal.
The trouble is, the flexibility created by a Black-Scholes analysis must be supported by corporate discipline for the analysis to hold any water. Traditional financial options, after all, are contracts with a specified expiration date, and their value tracks that of an underlying security. This explains why real options analysis can so readily be applied to projects in the energy business, where the use of futures contracts and exploration leases with definite terms is commonplace.
Elsewhere, however, capital investment decisions often do not involve contracts or have a security underlying them, so there's nothing to trigger expiration of the option or help assign a specific value to it. Sure, you can estimate precise outcomes for a project's scenarios over specific intervals, and thereby manipulate them to resemble financial options. You can also securitize the projected cash flows, as noted finance theorist Robert Merton has recommended. However, securitization may not always be effective, especially in pharmaceuticals and other industries in which there is considerable regulatory and other exogenous risk. As Nalin Kulatilaka, a finance professor at Boston University, and Martha Amram, a Palo Alto, California, consultant, point out in a recent article in the Journal of Applied Corporate Finance, "When the value and exercise of investment options cannot be linked to risks priced in the financial markets, the value of strategic options is better captured by other frameworks."
Pulling the Plug
In other words, real options analysis may rely too heavily on the assumption that a company possesses enough discipline to end a project without delay if the initial investment doesn't pan out. If that capacity is lacking, the option won't expire and the project's potential downside isn't zero, as the analysis assumes, but equal to the entire amount invested to that point. And the option's value will have been exaggerated by that amount.
So much for flexibility. Although "you can make any project look good if you build in enough options," notes Kulatilaka, a real-world approach must address two questions: When exactly do you shut it down, and is there a good mechanism in sight to do that? Adds Paul Greenberg, a principal at Cambridge, Massachusetts, consultancy Analysis Group/Economics, "You have to be willing to pull the plug."
Yet it's common knowledge that many companies continue to fund projects even when they aren't living up to expectations, if only because the managers involved have a vested interest in seeing them continue. For this reason, says Kulatilaka, an option's value often has more to do with a company's management and organization than with the methodology's application.
That helps explain why HP's use of real options, as extensive as it is, focuses on procurement and other aspects of its business where contracts are used. "We still have uncertainty," says Billington, "but we have a pretty good idea what it is."
Other companies are using real options applications in farther- reaching applications. General Motors and Ford, for instance, are developing applications for automobiles using the global positioning system (GPS) satellite network in combination with wireless technology.
Given the valuation discrepancies, most Wall Street analysts remain skeptical about real options analysis. Even a notable exception, Michael Mauboussin, managing director of Credit Suisse First Boston (CSFB), is careful to take into account a company's management and culture when valuing the options he deems it to possess. "Just because a company has certain options doesn't mean they'll be intelligently exercised," he says.
Laura Martin, a CSFB equity analyst, made much the same point as far back as July 1999, when she applied real options analysis to the cable industry. In her report, Martin concluded that Adelphi Systems's cable TV options were anywhere from four times to eight times more valuable than those of such competitors as Comcast, Cox Communications, and Time Warner, simply because more of Adelphi's assets were devoted to cable and the company was more highly leveraged.
Today, of course, the wherewithal of these and other companies to exploit their options has been reduced by the recent downturn in the capital markets. As Mauboussin puts it, "An option isn't of much value if you can't fund it."
In the end, BU's Kulatilaka suggests that real options analysis is more useful in conceptualizing projects than in evaluating them. That way, he says, "you use them to push the organizational change" required to make a project live up to its potential. But if such change is necessary in the first place, green-lighting a project based on real options analysis amounts to putting the cart before the horse. And CFOs at such companies can't be blamed if they prefer to give NPV the reins for at least a little while longer.
Ronald Fink (email@example.com) is a deputy editor at CFO.
WHOSE OPTION IS IT, ANYHOW?
The real value of a real option may have as much, if not more, to do with the nature of the holder than with that of the option.
The typical real options analysis, for instance, would assume that automotive applications based on GPS-wireless (also known as telematics) applications represented--at least at the outset--the same opportunity for both Ford and GM. And that each, therefore, possessed the same options and faced the same decision about investing in them. But because the companies have approached the technology so differently, the value of their options has also turned out to be different, according to Boston University finance professor Nalin Kulatilaka. GM's system, called OnStar, was reportedly designed, at least at the outset, to provide little more than navigational and other assistance to drivers. Ford's system, Wingcast, produced in a joint venture with Qualcomm, is designed to offer full Internet access and, say analysts, is based on technology that allows new features to be added more easily than GM's.
"GM looks at OnStar as another automotive feature," says Kulatilaka. A spokeswoman for GM disagrees, saying that the technological difference between the two systems is exaggerated, and that OnStar now offers virtually everything that Ford claims Wingcast will. OnStar "started with safety and security," she says, "but it has evolved since then."
Yet Ford's greater ambitions were clear from the start. Said Ford CFO Jacques Nasser, when the company announced its investment in Wingcast in July 2000: "We are...transforming the automobile into the next mobile portal." And, says Kulatilaka, "that's where the true value is."
Granted, GM is way ahead of Ford in rolling out its system. While GM began offering OnStar on 3 Cadillac models in 1996 and is now offering it on 32 of the company's 54 models overall, Wingcast has yet to be introduced. What's more, analysts worry that customers may not be willing to pay much, if anything, for even the most basic service, so cash flows may fall well short of projections.
"OnStar is less flexible than Wingcast," concedes Domenic Martilotti, an analyst at Bear, Stearns & Co., "but how willing are customers to pay for [Wingcast]?" He notes that OnStar is free for the first 12 months, and that while the company claims 60 percent of owners of OnStar-equipped cars renew for the annual subscription price of $199 to $399, GM hasn't said how many of those are Cadillac owners, who tend to be wealthier than other GM customers. Yet if, as some analysts expect, Ford ends up delaying or even ending its investment in Wingcast while GM continues to fund OnStar despite disappointing results, that would only support Kulatilaka's contention that Ford's option on telematics applications was more valuable than GM's because of Ford's greater discipline. -- R.F.