Stochastic approach to risk assessment of project finance structures under public private partnerships

The dissertation provides a forward-looking methodology in order to quantify the risks of a portfolio of project finance structures. As an application of such methodology, the thesis values the contingent government risk in Public Private Partnerships (PPP) for a set of project finance structures in a pre-determined time frame. Since the early 1990s, there has been evidence, due to fiscal restrictions, that governments are utilizing PPPs to fulfill basic infrastructure requirements that were traditionally provided by the public sector. The PPP project's viability lies in the government's ability to bear particular risks. In order for investors and international financiers to be comfortable that the government fulfills its obligations under the PPP, a guarantee fund is envisioned. Such guarantee fund is expected to be collateralized and directed to make payments under particular scenarios that future PPP projects encounter in a particular time frame. The study provides a framework in estimating the asset requirements for the guarantee fund in a pre-defined time frame. Traditional project finance structures rely on sensitivities test in assessing the robustness of the underlying project; such risk assessment has significant limitations. Utilizing the Monte Carlo simulation to determine the level of contingent obligations the government may face over the life of a set of projects under the PPP program further expands the field of project finance in the area of risk assessment. A Payment at Risk (PaR) methodology together with the Portfolio Simulation Approach (PSA), traditionally employed in the banking sector, may be used to determine the size of the guarantee fund to cover the government's contingent obligations under their respective PPP programs for a pre-defined period.