18. Stochastic forecast of the Slovak public debt

In this work we present a device for stochastic forecasting of the public debt that helps us achieve two objectives. First, it provides an independent, purely model-based, forecast of the public debt that cross-checks the official forecast. And second, estimated probability distribution allows us to quantify uncertainty of the forecast. At the core of the model, the standard debt accumulation equation is used to generate a large number of random forecasts. These are then used to extract the alternative – median – forecast along with confidence intervals for the central forecast. The debt equation summarizes stochastic behaviour of debt determinants – growth of nominal GDP, interest rates, primary surplus and exogenous financing of the debt – and is estimated in three steps. In the first step, growth of nominal GDP and interest rates are obtained in a VAR model. Independently in the second step we deal with the exogenous financing component. Finally, in the third step we utilise the outcome of the two preceding steps and simulate the fiscal reaction function together with the debt accumulation equation to generate a set of random forecasts of the debt.

The analysis was reviewed by Council for Budget Responsibility and National Bank of Slovakia. Their comments and author’s reactions are attached below.