Stochastic approach to debt sustainability analysis applied to Serbia

Zdravković, Aleksandar (2014) Stochastic approach to debt sustainability analysis applied to Serbia. In: Economic integrations, competition and cooperation. Faculty of Economics Rijeka, Rijeka, pp. 606-619. ISBN 978-953-7813-19-2

euconf_2013-676-689.pdf - Published Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (269kB) | Preview


This paper examines the public debt sustainability of Serbia, based on integration of financial gap analysis approach and stochastic modeling and forecasting of relevant macroeconomic and fiscal variables. Within this analytical framework, sustainability is interpreted as whether underlying policies can be sustained under plausible macroeconomic conditions without endangering solvency (Debrun, Celasun and Ostry, 2006). Conventional debt sustainability analysis is conducted as a simple accounting exercise, based on deterministic forecasts of variables that are included into debt accumulation equation and arbitrary scheduled bound tests. However, because debt sustainability is a forward-looking concept, it cannot be assessed with certainty (Wyplosz, 2011). Stochastic approach to debt sustainability as an alternative to conventional debt analysis takes into account the high degree of uncertainty surrounding medium-term debt trajectories, which cannot be captured by simple bound tests as these are limited in number (ECB, 2012). The recent research in this area (Garcia and Rigobon, 2004; Debrun et al., 2006; Kawakami and Romeu, 2011) mainly uses a Vector Autoregression (VAR) modeling as a basic framework for econometric estimation of the relationships among interest and exchange rates, inflation and primary balance and their forecasting and simulation. In addition, impulse response analysis is based on calibrated shocks obtained by Cholesky decomposition of variance-covariance matrix of the regression residuals. We apply stochastic approach to Serbian monthly data, run simulations of debt-to- GDP ratio and compare the results with IMF and Serbian Government mid-term projections. Our projections of debt-to-GDP ratio in two years ahead based on VAR approach gives the similar forecast relative to those given by IMF, while projections based on AR(1) approach seem to overestimate debt-to-GDP ratio with increase of forecast horizon. Yet, our forecasts strongly suggest that projection of Serbian government of debt-to-GDP ratio is too low and consequently misleading.

Item Type: Book Section
Additional Information: COBISS.SR-ID: 512268898
Uncontrolled Keywords: debt sustainability assessment, debt-to-GDP forecast, Vector Autoregression model, stochastic simulations, Serbia
M Category: H Social Sciences > H Social Sciences (General)
Depositing User: Mrs Aleksandra Bradic-Martinovic
Date Deposited: 07 Nov 2016 11:02
Last Modified: 07 Nov 2016 13:41

Actions (login required)

View Item View Item