Implicit Public Debt Thresholds: An Empirical Exercise for the Case of Spain

Abstract

We extend previous work that combines the Value at Risk approach with estimation of the correlation pattern of the macroeconomic determinants of public debt dynamics by means of Vector Auto Regressions (VARs). These estimated models are used to compute the probability that the public debt ratio exceeds a given threshold, by means of Monte Carlo simulations. We apply this methodology to Spanish data and compute time-series probabilities to analyse the possible correlation with market risk assessment, measured by the spread over the German bond. Taking into account the high correlation between the probability of crossing a pre-specified debt threshold and the spread, we go a step further and ask what would be the threshold that maximises the correlation between the two variables. The aim of this exercise is to gauge the implicit debt threshold or «prudent debt level» that is most consistent with market expectations as measured by the sovereign yield spread. The level thus obtained is consistent with the medium-term debt-to-GDP ratio anchor of 60% of GDP.