Financing research and development projects: Real options, spillover, and SWORD securities

This dissertation extends and applies real option theory to the external financing of corporate R&D projects. A real option model is developed which shows the interaction between R&D investment and financing decisions. The model is used to formulate a decision role for financing R&D projects. This is used in a clinical setting to evaluate SWORD securities (Stock Warrant Off-Balance Sheet Financing of R&D) issued by the biotechnology firm Centocor.; High-technology R&D projects are often both capital intensive and long term, and a heavy financial burden to firms. Companies have traditionally financed R&D projects internally, presumably to avoid the disclosure of proprietary information which may result in spillover costs if it falls into the hands of competitors. Many high-technology firms, however, do not have enough financial slack to internally finance every good R&D project, and are forced to externally finance or abandon them.; Real options are a recent development in capital budgeting theory, which are used to evaluate the option-like features found within many types of real investment projects. These embedded real options are the result of some flexibility or growth opportunity, which may be exercised by management over time as new information arrives to resolve some uncertainty. Real options cannot be measured using traditional discounted cash flow (DCF) techniques, like net present value (NPV).; R&D projects are particularly well suited to real option modeling because they contain a high degree of uncertainty, exist over long periods of time, have an output which itself is another option (either an option on a new business venture, or an option on the next stage of R&D), and proceed in stages where learning takes place over time to reduce some uncertainty. A continuous time real option model of R&D is developed showing the assumptions and limitations of the option pricing model. Critical issues to be resolved before applying a real option model to R&D include: the "spanning assumption," the existence of a dividend-like "convenience yield," and a "non-constant variance."; R&D projects are modeled as combinations of flexibility and growth options, in amounts which depend on the particular project. Flexibility options include the ability "to defer" or "to abandon" an R&D project. Growth options have two components: the expected growth option, which is an option on a new business venture; and unexpected growth options resulting from new discoveries. A real options perspective of R&D gives new insights about the interaction between investment and financing decisions.; Multiple real and financial options interact, which causes the interaction between investment and financing decisions for R&D projects. It is shown that: flexibility and growth options are independent of each other, making them value-additive; investment and financing flexibility are substitutes, making them non-value-additive; and spillover costs are the result of disclosing. growth options, not flexibility options. This is combined into a financing decision rule for R&D projects based on real option content. This is used to analyze SWORD securities issued by the firm Centocor.