Assessing Public Debt Sustainability in Mauritania with a Stochastic Framework
42 Pages Posted: 20 Apr 2016
Date Written: November 1, 2014
This work presents a stochastic framework for assessing public debt sustainability and applies it to the case of Mauritania. The sustainability assessment projects solvency and liquidity indicators -- public debt stock and gross financing needs relative to GDP -- for 2014-23. The analysis uses deterministic scenarios and stochastic simulations to analyze policy options and fiscal risks. The study relies on simple econometric models to generate forecasts of key macroeconomic variables driving the public debt dynamics and to compute debt-distress probabilities and debt thresholds. The study builds on basic techniques to determine optimal portfolios suitable as benchmarks for public debt management. A main result is that, if Mauritania maintains a strong growth performance and pursues sound policies to balance the budget and take advantage of concessional financing opportunities, it could reduce the public debt from 74 percent of GDP in 2013 to 30 percent by 2023, and the gross financing needs from 12 percent of GDP to 4 percent. Further scaling up capital spending is likely to deteriorate public debt sustainability because the estimated (marginal) growth-dividend is small. A more promising avenue would be to improve the quality of public investment and institutions, as opposed to the volume of capital expenditure. Different debt strategies can significantly affect the liquidity needs and the on-budget interest bill. But it is the fiscal policy geared toward balanced budgets that ultimately would permit Mauritania to improve the solvency indicators, and thus the public debt sustainability.
Keywords: Macro-Fiscal Policy, Economic Adjustment and Lending, Public Sector Economics, Government Policies, National Governance, Debt Markets, Macroeconomic Management, Youth and Governance, External Debt, Public Finance Decentralization and Poverty Reduction, Economic Conditions and Volatility
Suggested Citation: Suggested Citation