From driverless cars to an improved approach to measuring credit risk in

This paper is the first empirical study of the dynamics of debt service cover ratio (DSCR) in infrastructure project finance and shows that focusing on certain data points and using advanced statistical methods can greatly improve investors’ ability to understand credit risk in private infrastructure debt.

For this study, the authors rely on a new and unique dataset of cash flows spanning 15 years for hundreds of infrastructure projects in the OECD.

Co-author and Director of EDHECinfra Frederic Blanc-Brude said, “Most credit risk models of private infrastructure debt use static assumptions which are not suited for project finance because the risk profile evolves dramatically in time. The authors make use of concepts from robotics that are akin to how driverless cars position themselves in space to track key cash flow ratios in a mean/variance plane.

“Thanks to this, we can better understand and predict cash flow risk in private infrastructure debt”.

Ms Anne-Christine Champion, Global Head of Infrastructure & Projects at NATIXIS, said, ”This paper, based on a large and diversified sample of projects, demonstrates what we are experiencing in practice: the nature of the underlying revenue models (i.e level of contracted cash flow) is the most important variable to explain DSCR volatility.”

Ms Champion said, “With a now documented distribution of DSCR associated with a robust valuation framework, investors will be able to make informed asset allocation decisions on infrastructure debt. Those results are also highly valuable to calibrate prudential frameworks and to better align them with the risk profile of the underlying assets.”

Next steps with the EDHEC/NATIXIS research chair include the valuation of individual infrastructure assets and building reference portfolios that can be used as benchmarks of private infrastructure debt.