| The purpose of this dissertation is to propose a simple and general approach to solving real options problems. Most real options work to this point has been focused on complicated and specialized models. The focus here is to make real options modeling more robust and accessible. Although real options is the focus of this dissertation, many of the contributions are useful for financial modeling as well.; The main application areas are options on observables, which are general variables whose value can be observed. Observables can be classified as non-assets, market assets, or private assets. Non-assets are variables that are incapable of being traded but whose outcome is directly observable. Examples of non-assets include a company's market share, the total world demand for a product, a site's number of internet hits, and a competitor's decisions. Non-assets with no quantitative value can be given arbitrary values (e.g., a loan is either approved(1) or denied(0)). Market assets are items of monetary value that have a well-known price and are freely traded in an efficient marketplace like The New York Stock Exchange or NASDAQ. They could be stocks, bonds, or commodities; or, they could be derivatives of any of those. Private assets are items that either are not traded or are infrequently traded so that their current market value cannot be directly observed. We will estimate or compute their monetary value in this dissertation. The value of private assets may depend on the value of other private assets, of non-assets, and of market assets. Examples of private assets include private companies, projects, and intangible assets like a brand name.; The contributions of this dissertation supplement the theory and improve the lattice techniques used to value real and financial options. More specifically, the contributions supplement the theory by handling non-asset underlying variables, by including learning effects, and by finding risk-neutral probabilities with new sets of information. The contributions also improve the lattice techniques by introducing and testing new lattices that are accurate, robust, simple, and able to handle both multiple underlying variables and learning effects. |
