Modelling risk allocation in privately financed infrastructure projects using fuzzy logic

Risk allocation (RA) plays a critical role in privately financed infrastructure projects. Project performance is contingent on whether the adopted RA strategy is efficient. However, no mechanism was specifically designed to facilitate the risk allocation decision-making (RADM) process. Two theoretical frameworks based on the transaction cost economics (TCE) theory and on both the TCE and the resource-based view (RBV) of organizational capability, respectively, were thus adopted in this article. As conventional modeling techniques are not suitable for modeling RADM processes, which involve ambiguous and qualitative information, fuzzy inference systems (FISs) were developed, illustrated, and evaluated to model these frameworks. An industry-wide survey and rounds of expert consultation were conducted to collect data and generate fuzzy rules. It was found that both FISs are capable of reliably explaining the RADM process. In particular, the FIS based on both the TCE and the RBV theories performed more accurately and thus is more suitable for forecasting efficient risk allocation strategy.