Fiscal vulnerabilities depend on both the level and composition of government debt. This study examines the role of debt thresholds and debt composition in driving the non-linear behavior of long-term interest rates through a novel approach, a panel smooth transition regression with a general logistic model. The main findings are threefold. First, the impact of the expected public debt level on interest rates rises exponentially when the share of foreign private holdings exceeds approximately 20 percent of government debt denominated in local currency. Second, when the share of foreign private investors is 30 percent, an increase in the share of foreign private holdings of government debt could raise long-term interest rates once the public debt-to-GDP ratio exceeds 60 percent of GDP, offsetting the downward pressure on long-term interest rates from higher market liquidity. Third, out-of-sample forecasts of this novel non-linear model are more accurate than those of previous methods.
Detalhes
-
Autor
-
Data do documento
2020/06/11
-
TIpo de documento
Documento de trabalho sobre pesquisa de políticas
-
No. do relatório
WPS9276
-
Nº do volume
1
-
Total Volume(s)
1
-
País
-
Região
-
Data de divulgação
2020/06/11
-
Disclosure Status
Disclosed
-
Nome do documento
Debt Intolerance : Threshold Level and Composition
-
Palavras-chave
public debt; foreign investor; long-term interest rate; private investor; government bond yield; advanced economy; marginal impact; robustness check; forward interest rate; composition of government debt; revenue-maximizing tax rate; risk-taking behavior; local currency bond market; share of government bond; increase tax revenue; increase in interest rate; low real interest rates; government debt security; nominal interest rate; exchange rate risk; private foreign investor; foreign currency; domestic investor; net foreign asset
- Exibir mais
Downloads
COMPLETAR RELATÓRIO
Versão oficial do documento (pode conter assinaturas, etc.)
- PDF oficial
- TXT*
- Total Downloads** :
- Download Stats
-
*A versão do texto é um OCR incorreto e está incluído unicamente em benefício de usuários com conectividade lenta.