This dataset is one of the early outputs of a substantial effort launched at EDHEC to collect private information and document the investment characteristics of infrastructure investments, and create a global database of infrastructure cash flows spanning several decades.
In a context were data paucity remains a concern and complete time series of cash flows covering the decade-long life of investments remain rare, the authors also propose a novel approach to modelling and predicting the “trajectories” of DSCRs in infrastructure projects. The paper documents the existence of homogenous “families” of cash flow dynamics in infrastructure projects that are best explained by the contractual characteristics of these investments as well as the initial financial structuring choices made jointly by project sponsors and creditors.
Hence, this paper also addresses the question of the path-dependency found in project dynamics and the resulting serial correlation of returns, especially in cases where achieving full diversification might be challenging. Borrowing from statistical methods usually applied in the physical sciences, the authors show how new information arriving sequentially, which is typical of long-term investment in infrastructure, can be integrated to infer the conditional parameters of cash flow distributions and the related credit risk measures for subgroups or even individual investments.