Predicting contract surety bond claims using contractor financial data

This paper describes the results of a study to predict contract bond claims based upon contractor financial data. Financial statements of 87 contractors (36 claim and 51 nonclaim) were obtained from four companies underwriting construction-contract surety bonds. A claim contractor was defined as one that defaulted on a bond, many times requiring the surety to pay a loss. A nonclaim contractor was defined as one that has not defaulted on a surety bond. A predictive model was developed using discrete choice modeling. Variables identified in the model are cost monitoring, underbillings/sales, total current liabilities/sales, retained earnings/sales, and net income before taxes/sales. This model predicts the probability of experiencing a claim in the accounting period following the period in which the financial statement was prepared. Considering each variable, recommendations are presented for reducing the probability of claim. Additionally, questions related to the variables have been developed for consideration by the management of construction companies. The model can assist construction-industry professionals by providing them with a quantitative tool to evaluate contractors.