To print: Click here or Select File and then Print from your browser's menu

This story was printed from silicon.com, located at http://www.silicon.com/

Story URL: http://comment.silicon.com/0,39024711,11011897,00.htm

Strategic Vision: Giga Information Group on... Evaluating IT Investment
Traditional ways of measuring IT's impact on the balance sheet have been notoriously unreliable, leading many FDs to hold off on IT spend. But a new method called Real Option Pricing could change all that. In the first of a series of monthly columns, Giga Information Group VP, Will Cappelli, examines what it means

By silicon.com

Published: Wednesday 04 August 1999

Business managers have - finally - woken up to the idea that IT can be a strategic investment. Despite this, many companies continue to balk at spending significant sums on IT infrastructure.

One reason for this is that investments in infrastructural components - such as middleware and network management software - rarely show sufficiently high return on investment (ROI) when the numbers are evaluated using traditional net present value (NPV) techniques.

It's easy to criticise the assumptions that underlie the most widely-used NPV techniques. Indeed, such criticisms have frequently been voiced by frustrated IT and business managers when their cherished proposal fails to make it through the NPV filter.

Unfortunately, there have been few alternatives to NPV on offer. Those that have arisen have tended to be too controversial to serve as an alternative way of predicting the stream of costs and benefits likely to flow from a given IT investment.

A revolution, however, has begun to sweep through the practice of corporate financial planners and capital budgeteers, one which could have significant repercussions on the way IT investment is planned and justified.

Across the US and, increasingly, in Europe, NPV is being rejected in favour of a new method called 'Real Option Pricing' (ROP.) This is an important trend for IT decision makers because it so happens that ROP is far more likely to demonstrate the value of an IT infrastructural investment than traditional NPV.

The traditional NPV method begins by projecting the possible streams of costs and a stream of benefits associated with a given capital investment. It discounts the costs and benefits in accordance with the corporate internal rate of return and assigns probabilities to each of the possibilities. The average value of the discounted cost/benefit stream (weighted by the probabilities) is then taken as the value of the investment. If it's positive, then a go-ahead is warranted. If it's negative, then it should not be considered.

The big problem with NPV is that it fails to take two key factors into account. Firstly, information about the market changes over time and secondly, decisions made now can either allow a company to take advantage of the new information likely to be available in the future, or prevent them from taking advantage of that new information.

Now, putting a company in a position to be able to take advantage of the opportunities revealed by new information itself has a value which NPV models simply do not capture. Hence NPV will tend to underestimate the full stream of benefits - particularly with regard to investments where uncertainty is high but liable to reduce with time. (It will also underestimate the full stream of costs associated with not making such investments).

ROP addresses these issues by treating capital investments like financial options. After all, in a very strong sense, an option is nothing more than an investment that puts you in a position to take advantage of future changes in information about the market.

ROP had its origins in attempts to evaluate the potential of high-risk North Sea oil fields in the early 1980s. Outside of the oil industry, it mainly occupied the attention of academic financial economists until the early 1990s when Paul Henderson of Boston University first discerned it could be applied to IT capital investments. Unfortunately, Henderson's attempts at evangelisation fell largely on deaf ears, since ROP had yet to penetrate the practices of industry at large.

However, general recognition of an acceleration of the pace of business change has, over the last two years, greatly increased the visibility of ROP in corporate financial offices. Textbooks have been published. The Harvard Business Review, Business Week, and other popular business journals have all recently propounded the advantages of ROP over traditional NPV.

Finally, 20 years of academic and industrial work with the model have reduced what were originally a set of rather arcane mathematical formulae to relatively easy-to-construct spreadsheet models. In other words, the time has come to reintroduce ROP to IT business justifications.

ROP is not the Holy Grail, of course. It has its limitations. In particular, it presupposes (like NPV) that one can trace the causal relationship between the implementation of a given piece of hardware or software and changes in the nominal flow of cash to a given organisation. Nonetheless, if one has enough information to build an NPV model, one has enough to build an ROP model. And the latter is likely to be a far better guide to the total economic impact of an IT investment.

Will Cappelli is one of Giga's most experienced analysts, and specialises in the areas of outsourcing, service management, and network and systems management technologies. Will has 16 years of experience in the IT industry, serving in research, management and business development positions in a number of major IT consultancies and research houses, including Ovum, Meta Group and Gartner Group. He has published articles and books on a wide range of IT subjects, ranging from chargeback architectures to parallel processing.