Loss prediction model for building construction projects using insurance claim payout

Abstract

The amount of risk and the occurrence of damage in construction projects have increased as construction projects are becoming larger and more complex, increasing the demands for an effective risk assessment model. Therefore, quantitative risk analysis is needed to develop a financial risk assessment model using the risk indicators for construction projects. To address this need, authors analyzed the damage records of actual construction sites in South Korea, in order to identify the critical risk factors of damage and investigate the relationships among the risk indicators to develop a loss prediction model. Authors used claim payout records from an insurance company to reflect real financial loss as the dependent variable. As independent variables, authors adopted four risk categories based on our literature review and data analysis: natural hazards, geographic information, construction method and ability, and construction information. Our loss prediction model for construction projects, along with our findings, offers fundamental guidelines for construction companies, construction project owners, and insurance companies hoping to model and predict financial loss for building construction projects.