Applications of Stochastic Modeling for Investment Decision-making under Market Uncertainties

Profit-seeking organizations make investment decisions with financial implications. These decisions are often complicated by (i) market uncertainties about investment payoffs, (ii) the possibility to select a portfolio of investments, and (iii) the presence of real options, which make it possible to postpone investments until some uncertainties are resolved. Quantitative support for these decisions builds on methodological contributions in stochastic modeling, financial modeling, and decision analysis, among others. This dissertation develops scenario-based approaches and decision models for several problem contexts, most notably for (i) the optimal harvesting of forest stands, (ii) the management of electricity contract portfolios, (iii) the investments in power plants, and (iv) the valuation of real options in new product development. These models support investment decisionmaking and, in some cases, they also help analyze policy impacts at the industry level. Each model is presented in view of later publication in a refereed journal. The methodological advances of the dissertation offer several novel insights into the above decision problems. In the context of the optimal harvesting of forest stands, this dissertation demonstrates that multi-level risk management over several time periods and at multiple confidence levels can reduce risks significantly without a major reduction in the expected return. In the management of electricity contract portfolios, the dissertation shows that correlation between price and load is important to model, or else risks may be underestimated, resulting in suboptimal decisions. In the analysis of the investments in power plants, it is concluded that uncertainties in the carbon dioxide emission policies may foster the development of more concentrated and less competitive electricity markets, because the new investments are more likely to be made by larger financially stronger incumbent firms than small, project-financed independent power producers. Finally, in the valuation of real options in new product development, the value of real options is shown to be non-monotonic with respect to increased competition, whereas the options for enhancing product development and delaying product launch are found to be typically most useful when the level of competition is weak.