Software Investments under Uncertainty Modeling Intangible Consequences as a Stochastic Process

Sammanfattning: Software systems are today a part of more or less every organization. The varieties of software used in organizations are ranging from simple log-keeping applications to advanced decision support systems. The task of a priori valuation of software investments has attracted a lot of research for a long time. One of the main themes of this research has been which types of consequences software investments result in and how these consequences can be incorporated in the a priori valuation of the investment. Much of this research has stated the problem as how to incorporate intangible consequences in the valuation since intangible costs and benefits are assumed to represent a large part of the consequences from a software investment. These consequences are therefore highly relevant in the appraisal of software investments. This thesis is concerned with the question of how intangible consequences can be incorporated in the a priori valuation of a software investment. To answer this question, this thesis presents a theoretical model for the valuation of a software investment based upon a discounted cash flow model in continuous time. The general model argued for in this thesis is that usage results in consequences which must be translated into cash flows to be incorporated in a discounted cash flow model. The software usage is chosen as the underlying value creating function since it is the basic underlying function that creates all consequences specific to the software investment. This thesis develops a stochastic cash flow model to incorporate the uncertainty and characteristics of when the intangible conse quences have an effect on the cash flow by adopting a Brownian motion into the valuation model. To find an analytic model for the problem, the expectations of the future cash flows is transformed into risk-neutral expectations. This allows us to use the risk-free rate of return as a discount factor in the model.