Justification of investments in new information technology is one of the many challenging issues facing managers today. Many tangible and intangible factors have to be assessed and weighted. Although qualitative factors play an important role in IT investments, the evaluation of quantifiable costs and benefits should at least be a part of any valuation. The most common financial justification method used, offered by the capital budgeting theory, is the Net Present Value method (as one of the discounted cash flow methods). The NPV method, however, poses several problems. One of them being how to deal with future possible investments (options) enabled by an actual investment, for example, in IT infrastructure. In 1972 Black and Scholes developed a model to determine the right price of an option in the financial markets. Later this model was adapted to applications in industry settings. In 1985 Brennan & Schwarz adapted the model to value projects for the oil industry. Also the insurance, timberland and mining industries started to use option theory.

In the 90’s attempts were made to apply option theory to IT investments. In 1991 Dos Santos proposed the Black and Scholes model for application to IT investments in order to value projects which can only exist when another investment has previously been made; the so called ‘second-stage’ projects. Two years later Kambil, Henderson and Mohsenzadeh introduced the option perspective to establish a linkage between many categories of IT investments and business value.

In the following we refer to the models mentioned above as (complex) option models. The question addressed in this paper concerns the practical applicability of option pricing in valuing IT investments. In other words, what advantage does option pricing, compared to the NPV method, offer to management?

The development of theory with regard to investment analysis spans several decades. In the 50's Anthony laid the foundation of what is actually known as ‘capital budgeting’. Starting from a financial perspective several methodologies have been developed to guide decision making in this area. Well known and widely applied, they have become the so called discounted cash flow methods. These methods assess the extent to which investment proposals support the financial objectives (eventually shareholder value) of the firm.

A basic characteristic of the DCF methods is that investments are represented as a set of negative and positive cash flows. In order to value a project we must know the dollar value of the initial investment, the outlays and revenues during its lifetime, and the salvage value of the investment, hence an IT project. Positive and negative cash flows at different moments are not comparable as such. For that reason discounting is introduced. Two methods prevail: the net present value method (NPV) and the internal rate of return (IRR). The former discounts cash flows, using a time value of money as the discount rate; the latter seeks the discount rate that equals positive and negative cash flows. For an investment to be acceptable respectively the NPV should be greater or equal to zero, or the IRR should be equal or greater than the time value of money. A problematic issue is finding the right time value of money. Different theories have been developed. Bierman and Schmidt (1992) argue that the time value of money is determined by the cost of capital, being the weighted average of the cost of acquiring the different capital components on the balance sheet of the firm. In the following we will concentrate on the net present value.

The NPV method has received a lot of criticism from many authors. Major problems concern the ability of the method to value intangible benefits and costs, the estimation of future cash flows, the possibility to properly value management flexibility, and the determination of the appropriate discount rate. As we only focus on quantifiable factors considered in an investment analysis in this paper, the first subject is beyond our scope.

Generally, the NPV method uses a series of discrete cash flows per period, usually per year. The investment outlay is assumed to occur at the beginning of the first year, the subsequent cash flows are assumed to be received or paid at the end of each period. This is a simplification as e.g. revenue will be collected throughout the year. Using one estimate per period also raises the question of how high this estimate should be. As future cash flows cannot usually be predicted with a hundred percent certainty, some probability distribution applies. However, as is the case in many economic decisions, objective probabilities are impossible to generate. The decision makers have to rely on subjective probabilities, which are the personal estimates of those involved in the decision making process. Often a distinction is made between an optimistic, a pessimistic and a neutral prediction per cash flow, each of the predictions is granted a probability to occur (the sum of all probabilities per cash flow being equal to 100%). A possible appropriate estimate of the periodical cash flow will be the expected value (the statistical mean) of the distribution function. It should be noted that ‘the statistical mean’ is not equal to the cash flow with the highest probability, which is often used as an estimate (see Figure 1) (of course, in the case of a normal distribution, the statistical mean will be equal to the cash flow with the highest probability of occurrence).

The use of the 'most likely' cash flow will result in a wrong net present value (Palm et al. 1986).

The 'expected' cash flow, calculated as the statistical mean, should be used.

Second, the discount rate is problematic. Besides choosing the right basis for calculating the ‘time value of money’, its relation to the project risk is a problem. In order to accommodate for project risk a ‘risk adjusted discount rate’ is often used, which is the summation of a risk-less market rate (e.g. returns on bonds) and some risk premium. Applying a single risk premium assumes a particular risk profile for the whole project. Different stages in the project lifetime and different cash flows may be connected to different risk profiles.

A third important problem poses the concept of management flexibility. As is stated by different authors (Brennan et al. 1985, Dos Santos 1991) the NPV method does not properly take account of management flexibility. Consider, for example, in our previous example the existence of the possibility (an option) to start a second project, like installing a new highly productive spreadsheet on-top-of the graphical user interface. The option to use this application under the new interface adds value to the investment in the GUI. The traditional DCF method does not incorporate this (extra) information. Management flexibility has been the most important reason to introduce a new method: option pricing. Many authors like Dos Santos and Kambil et al. see this as the key aspect for introducing option theory. In the next paragraph option pricing will be discussed.

The option theory is a theoretical model which is commonly used in the financial world to determine the price of an option on the derivative market. In 1972, Black & Scholes (1973) developed a model to determine the price of an option in the financial markets. Brennan & Schwartz (1985) adjusted this model to value projects for the oil industry in 1985. Currently, in a wide range of industries, such as insurance (Marcus et al. 1984), timberland (Zinkhan 1991) and mining (Palm et al. 1986) option pricing is being introduced for investment analysis purposes.

In the world of IT, the use of the option theory based on the Black & Scholes model, was proposed by Dos Santos (1991) to value second-stage projects. In 1993, Kambil, Henderson, and Mohsenzadeh, introduced the options perspective. For them option pricing is a critical first step in establishing linkage between many categories of IT investments and business value.

In the following we will explain the basic properties of the option model, both in the finance and the IT world.

An option gives the holder the right to buy or sell a share of stock at a specified price (Van Horne 1980). This price is known as the exercise price, or strike price. A call option gives you the right to buy a stock, and a put option gives you the right to sell a stock. In this article we will focus on call options.

For example, one could buy a call option on a stock at $25 on October 31 of the current year, which is the expiration date. If the price of the stock exceeds the strike price on October 31, we will buy the stock, or exercise the option. If the stock price is lower, we will not exercise the option.

Black and Scholes were the first to develop a model to value options, which has now been widely accepted and enhanced.

The first thing we notice as we look at the Black and Scholes formula is that it is rather complex. Bookstaber states that "The number of people who use these models, exceeds the number who understand them." (Bookstaber 1991). Other writers describe the option pricing model as being "complex and un-intuitive to many of its users." (Brenner et al. 1994). Although more sophisticated models can reflect a more accurate future, the ability to understand and communicate the results of a model, is probably its most valuable property.

If you would like to understand the content of the option pricing theory we advice you to read the next two exhibits. If an explanation of the overall concepts will do you can skip to the next section.

The valuation of IT investments has always been a problem. Especially investments in IT infrastructure are difficult to value; future cash-flows are very uncertain and difficult to identify. IT infrastructure investments enable follow-up or second-stage investments. This can be seen in the following formula.

NPV(total)= NPV(infrastructure)+NPV(second-stage) (5)

The option theory explicitly focuses on the second NPV calculation. It assumes that the traditional NPV does not include this part of the calculation, or is not able to calculate the correct value for it. By incorporating the NPV of the follow-up investment the management flexibility concept in the option pricing is being introduced. Management flexibility implies that management still does have a choice to start the second stage project or not to start the project at all.

The option pricing model confronts management with three problems:

• estimation of the input values for the variance and NPV of the second stage project is hard;
• the model is too simplistic because too many assumptions are being made;
• the model is too complex to communicate.

The option pricing model requires two important input parameters; the variance of the NPV of the second stage projects, and the NPV of the second stage project itself. Managers are used to thinking in decision points, not in continuous distributions of cash flows. Therefore, managers will find it very hard to answer the question "what is the standard deviation of the rate of change of the development costs or the revenues?". It will be even harder to estimate the correlation coefficient between the rate of change of revenues and the development costs.

Estimation of the NPV of the second stage project remains the same old problem for management; predicting cash flows and determination of the appropriate discount rate. The option model does not solve the problems with the DCF, it only creates more.

The option formula as presented is too simplistic to have 'real life' value. The original Black and Scholes formula has assumptions that will not hold, such as constant interest rate, no transaction costs, the stock pays no dividend. Researchers, especially in finance, have worked to adjust the base model, in order to relax these assumptions. Relaxing the assumptions, however, increases the complexity of the model (Markland 1992), and makes it even more difficult to use.

Another problem with the option pricing model is that the formula is hard to understand. As we stated earlier, other authors labelled the model as complex and un-intuitive to many of its users. Communicating results from such a model will pose problems for managers who will have to understand the nature of these results. A related risk of using complex option models is that the attention to the real challenging issues is lost; for IT investments predicting future costs and benefits for the NPV calculation are the real problem, not the volatility of that NPV until a certain decision point. Therefore, the option model adds relatively low value to the total valuation problem.

Although complex option models are hard to use, we suggest to incorporate option thinking in a more practical way. By using decision trees (manageable) decision moments can be identified and valued. The following exhibit shows that option thinking can be used to construct decision trees using the NPV method and compares the results with the traditional NPV method.

In an investment model all factors are subject to uncertainty: duration of the project, distribution of cash flows, and related to this, the discount rate. Management can deal in different ways with this uncertainty. It can shorten the duration of the project, make a conservative estimation of the (distribution of) cash flows, or use a higher discount rate. None of these approaches are perfect. Option pricing learnt us to incorporate the value of an option to an investment. It still leaves us, however, with the same problems as in the NPV method. The complex option models do not add much value to IT investment analysis. Even worse, using complex option pricing models provides management with a very difficult task to estimate the input parameters. And above all, these methods are hard to understand and to communicate.

The NPV method, on the other hand, is a well known model. With this familiar method management deals with uncertainty by using relatively easy to understand decision trees and estimated cash flows.

The complex option models as proposed may be a dead end, but they have taught us two important lessons:

• to leave negative NPV’s of follow up investments out of the valuation when there is an option available to management;
• to explicitly recognise the freedom of choice management has concerning second stage projects.

• Bierman H and Smidt S (1992), The capital budgeting decision: economic analysis of investment projects, Prentice Hall.
• Black F and Scholes M (1973), The pricing of options and corporate liabilities, Journal of Political Economy, Vol. 81, No. 3, May/June, 637-654.
• Bookstaber R M (1991), Option pricing and Investment strategies, Probus Publishing Company, Chicago, third edition.
• Brennan M J and Schwartz E S (1985), Evaluating natural resource investments, Journal of Business, Vol. 58, 135-157.
• Brenner M and Subrahmanyam M G (1994), A Simple Approach to Options Valuation and Hedging in the Black-Scholes model, Financial Analysts Journal, March-April, 25-28.
• Cox J, Ross S and Rubinstein M (1979), Option pricing: a simplified approach, Journal of financial economics, Vol. 7, 229-263.
• Dentskevich P and Salkin G (1990), Valuation of real projects using option pricing techniques, Omega international journal of management science, Vol. 19, No. 4, 207-222.
• Dos Santos B L (1991), Justifying Investments in New Information Technologies, Journal of Management Information Systems, Vol. 7, No. 4 , Spring, 71-90.
• Kambil A and Henderson J C and Mohsenzadeh H (1993), Strategic management of information technology investments: an options perspective, Idea Group Publishing, Chapter 9.
• Marcus A J and Shaked I (1984), The valuation of FDIC deposit insurance using option-pricing estimates, Journal of Money, Credit, and Banking, Vol. 16, No. 4 , November, Part 1, 446-460.
• Margrabe W (1978), The value of an option to exchange one asset for another, The Journal of Finance, Vol. 33, No. 1, 177-186.
• Markland J T (1992), Option Theory: A new way forward for exploration and engineering economics?", SPE Oil and Gas economics, Finance and Management conference, London, April 1992.
• Palm S K and Pearson N D and Read J A Jr. (1986), Option pricing: A new approach to mine valuation, CIM bulletin, 61-66.
• Van Horne J C (1980), Financial Management and Policy, Prentice Hall, fifth edition, Englewood Cliffs.
• Zinkhan F C (1991), Option pricing and Timberland's land-use conversion option, Land Economics, Vol. 67, 317-325.