Are "real options" actually used in the real world?

INTRODUCTION

In their highly regarded book published at the turn of this century, Copeland and Antikarov (2001) suggested that real options would dominate the capital budgeting process within the next decade.

As these authors and others, such as Trigeorgis (1993) and van Putten and MacMillan (2004), have pointed out, the users of net present value and other discounted cash flow methods may not consider the flexibility to revise decisions after a project has begun. The flexibility might include terminating the project, taking a more desirable route once initial results are in, greatly expanding the project if there is unexpected success, and so on. Such elements are particularly likely to be present in natural resource discovery, technology-related investments, and new product introductions.

But the list does not stop here. Almost every capital budgeting project contains a potential element of flexibility once it is put into place. There is a real option to change the course of action, and this real option has a monetary value just as a financial option does.

While many proponents of real options are critical of NPV analysis, it is not necessarily NPV analysis that is at fault per se, but rather the improper use of it in many cases. One can use NPV analysis "correctly" if all options (invest, do not invest, delay, etc.) are known at the start of the project such that they all can be evaluated. This is why decision trees were developed. The benefit of real options analysis is that it incorporates volatility, whereas decision trees (which rely on NPV analysis) only compute expected values.

CAPITAL BUDGETING STUDIES AND REAL OPTIONS

Capital budgeting techniques used in business have been studied extensively by Mao (1970), Gitman and Forrester (1977), Schall, Sundem, and Geigsbeck (1978), Scott and Petty (1984), Canada and Miller (1984), Block (2005), and others. The central message throughout is that major industrial firms are moving toward the normative in various areas of capital budgeting.

However, the topic of real options is not covered or scores poorly in terms of utilization. For example, as reported by Teach (2003), in 2000, Bain and Company conducted a survey of 451 senior executives covering 30 industries regarding their views of management techniques, and only 9% reported using real options. Also, Ryan and Ryan's (2002) survey of 208 CFOs found real options trailing the list of 13 supplementary capital budgeting techniques with a utilization rate of 11.4%. In contrast, 85.1% of the respondents used sensitivity analysis and 96% used traditional net present value analysis for basic capital budgeting. In both of these studies, the inclusion of real options was an afterthought with no follow-up questions or analysis.

The current article is specifically dedicated to real options with the intent of finding out who is doing what and why.

It is organized as follows. In the following section, the methodology behind the current study is described; then the presently existing practices using real options are covered, the resistance to real options is analyzed, and a summary and conclusion follow, along with the future outlook.

METHODOLOGY UTILIZED IN THE CURRENT STUDY

The author used the Fortune 1,000 companies as the initial database for the study. Two hundred and seventy-nine usable responses were returned by the top-ranking financial officer of the firm or his designate. The three-page questionnaire was based on a pilot study of 50 Fortune 1,000 companies to ensure clarity of meaning.

A follow-up telephone survey of 40 randomly selected non-respondents subsequent to the actual mail-out indicated no statistically significant difference between those who initially answered the questionnaire and those who elected not to participate.

The financial characteristics of the participating firms are presented in Table 1. Of equal importance is the industry of origin of the survey participants. The industry listings of those firms who participated in the survey are presented in Table 2. The 12 categories are the most representative of the two-digit SIC codes of the respondents. Not all categories are mutually exclusive, so the emphasis is on the firm's primary area of activity. The fraction of responding firms versus non-responding firms, broken down by two-digit SIC codes, was generally consistent across all categories.

PRESENTLY EXISTING PRACTICES

The question of primary interest is what percentage of the respondents use real options as part of their capital budgeting process. Forty respondents, representing 14.3% of the survey participants, indicated the use of real options. Out of the 40 users, 18 indicated major utilization, 13 indicated they used real options as a supplemental tool, and 9 reported using real options to shadow the results of more commonly used methods.

In terms of the applications of real options to different types of decisions by the 40 users, the answers are reported in Table 3.

While the categories are clearly not mutually exclusive, and many respondents indicated an overlap, it is apparent that the nature of the decisions indicated has a high degree of uncertainty. Clearly, a new product introduction has subsequent components in which a decision must be made to either add additional resources, penetrate different markets, cut back on expenditures, or even abandon the project. The same type of decisions relate to research and development, mergers and acquisitions, and other categories.

INDUSTRY OF USERS

In their work, Triantis and Borison (2001) found that there were tendencies among those who employ real options to represent certain industries. Out of the 40 users in this study, 37 came from the following: technology (13), energy (11), utilities (6), health care (4), and manufacturing (3). Interestingly enough, finance had only two users and transportation had one. More will be said about the low participation rate by the financial sector (investment banking, commercial banking, etc.) later in the article.

A chi-square independence of classification test indicated that a null hypothesis that there was no relationship between industry classification and use of real options could be rejected at an alpha level of .01. The results are unknown in the appendix.

METHODS OF UTILIZATION

For the 40 firms that utilize real options, a follow-up question pertained to what technique they used in their implementation. They were given four major categories as well as another category. The major choices were those previously cited by Triantis and Borison (2001) in their research. The responses are shown in Table 4.

Needless to say, the approaches are not mutually exclusive and many were used in combination. The emphasis here is on the primary method of utilization. Also, some respondents used slightly different names for one of the four approaches listed here, but the intent was the same and was categorized as such.

While the binomial approach is the most frequently used, it was thought by the respondents to be more simplistic than the risk-adjusted decision tree approach or Monte Carlo simulation.

Although the Black-Scholes option pricing model is the key to valuation for financial options, the same cannot be said for real options. The Black-Scholes model requires knowledge of five variables: the price of the underlying asset, the price at which the option can be exercised profitably, the amount of time before the expiration date, the risk-free rate of interest, and the volatility of the underlying asset. Many of these variables are simply unavailable for real option analysis. For example, the time period before expiration can be easily determined for a financial option but is not discernable for a real option, where flexible decision-making allows for many different potential exercise dates, depending on how the project is progressing.

To determine if there were industry preferences among the four different methods of utilizing real options, a chi-square test was once again run in which the null hypothesis was that there is no relationship to the technique used with real options and industry classification. The null hypothesis could be accepted at almost any reasonable level of significance. There is no discernable relationship between technique used and industry classification. The results are shown in part b of the appendix.

In their research, van Putten and MacMillan (2004), Trigeorgis (2005), and others have stressed the importance of treating real options as a component of expanded net present value rather than as a stand-alone approach. The NPV is determined and the value of the real option is added.

Total value = passive NPV + real option value + other adjustments

Thus, it is possible to have a project that has a negative net present value but a positive total project value because of the presence of the real option. By failing to account for the real option value, an incorrect decision may be made. This is particularly likely to happen in projects in which there is a high degree of risk but also a large amount of flexibility. The classic NPV approach, if not properly used, may fail to pick up the value of being able to change the plan in the early, middle, or late stages of the project. The classic NPV is likely to have the greatest value in the early years of the project because of the time value of money, whereas the real option component may be more valuable in the later life of the project because uncertainty may be at its highest level and flexibility at its most valuable point.

Thus, classic, stand-alone NPV, if not properly used, may undervalue potential rewards in the research for new drugs in the pharmaceutical industry, the exploration for natural resources in the energy industry, and so on. Eighty two percent of the respondents employing real options indicated that they use the valuation of real options as a component (add-on) to the classic NPV approach rather than as a stand-alone item. The classic NPV is retained but modified to reflect how managers think and act in the real world.

Print Get the Mag Weekly Updates