In this paper, we develop a structural credit risk model that relies on cash flow data to derive credit risk metrics. The model is useful for illiquid assets for which a time series of prices is not observable.

Our methodology is designed to require a parsimonious dataset of observable inputs, and provides a clear link between an asset’s fundamental characteristics and its risk profile. The model is flexible enough to value debt instruments with path-dependant cash flows, such as mortgages and floating rate loans, and can incorporate various debt covenants, such as debt refinancing, and restructuring options, as well as cash sweeps, dividend lockups, and reserve accounts.

The implementation of the model is illustrated with project finance debt, which is highly illiquid, and suffers from a serious lack of price data. We show that the dynamics of the debt service cover ratio (DSCR) along with the debt repayment profile and the debt covenants is sufficient to implement our credit risk model.

For reasonable parameter values of the DSCR dynamics, the model reproduces stylised empirical regularities regarding the probabilities of default for two generic types of infrastructure projects.