Risk analysis and high speed rail projects in France: introducing economic slowdown into appraisal methodologies

Abstract In France the high speed rail (HSR) network has been regularly extended (More than 2000km in 2015). But France is now facing at once a structural economic slowdown and a growing scarcity of public funds. What are the impacts of this “new deal” on HSR projects appraisals? How to introduce these systemic risks in CBA? How the public decision maker can take into account the risk? Can we guess that the state, investing in a large number of different projects, has indeed a complementary portfolio and therefore not risky? The paper makes a distinction between non-systemic risks (when the risk of the project is not correlated with macroeconomic trends) and systemic risks (related to macroeconomic variables like GDP). These systemic risks are exactly the kind of risks that HSR projects have to face because of the economic slowdown. In the new methodology proposed in France by the new guideline (2014), different options are proposed to introduce the macroeconomic risks, especially the likelihood of a long period of very weak economic growth. Considering the systemic risks, and following one of these options, we introduce in CBA a QUOTE β parameter, defined by taking into account the macroeconomic risk. The result is a change in the discount rate. A first stimulating result has been obtained by computing not the IRR, that is to say the rate of discount leading the NPV to a value of zero, but the rate of economic growth of GDP (between 2016 and 2024) cancelling the NPV of a given project (South West HSR project). A second stimulating result lies in the comparison of NPV and IRR according to different values of the rate of discount. Taking into account the risk in the new methodology leads to increase the rate of discount and therefore to reduce the NPV of the project, even if the IRR is higher in relation with the fact that the new methodology with risk is including a longer period of operating (70 years instead of 50).