Is it possible to predict sovereign debt default/crises? - A study to identify leading economic indicators while developing an empirical model to predict Sovereign Debt Default

Aneja, Vandit (2011) Is it possible to predict sovereign debt default/crises? - A study to identify leading economic indicators while developing an empirical model to predict Sovereign Debt Default. [Dissertation (University of Nottingham only)] (Unpublished)

[img] PDF - Repository staff only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (5MB)


There are economic and social costs attached to a country experiencing debt default. As the global economy is integrated, the presence of default in one country can lead to a default in another country with which it has economic and trade associations. This risk of contagion along with adverse economic and social impact of default makes debt default prediction an important topic of research. Correct and early prediction of debt crises allows for the possibility of remedial action and hence the reduction of economic hardship.

The scope of this management project is to analyse the statement whether it is possible to predict a sovereign debt default/crises. If it is possible to predict, then what are the economic, financial or political indicators that would be required in default prediction model. This research paper follows a step-by-step procedure in order to come to the conclusion of the question “Is it possible to predict sovereign debt default/crises?” A number of previous attempts by authors to predict sovereign default via elaborate and complex empirical models and early warning systems are investigated. A closer look at the different forms of sovereign debt default definition associated with these empirical models forms a starting point for the analysis. The analysis would also take into account the different sets of explanatory variables used in various studies to predict default. The paper will also overview the different model estimation techniques that have been used by authors. A detailed analysis of selected TWO predictive empirical models is made. The two models would be verified as per their original model in sample data. Then, they would be analysed for their applicability to contemporary emerging and advanced countries. The focus is to find whether a sovereign default can be predicted and it would be attempted by the construction of a logit regression model. The model would be based on a specific definition of the sovereign debt default and other feasible criteria such as underlying model complexity and the number of economic indicators available for countries that can be utilised in the model. A thorough analysis of the model and its implications with underlying assumptions and limitations would be the conclusive element of this research. The construction and analysis of the predictive model would also be supplemented with a discussion of used methodological approach and relevant issues faced.

On through analysis of the debt default prediction models in previous studies and the one developed in this paper, it is concluded that the models may be statistically significant predictive models, but may not be economically significant and viable models in practice. Any predictive model is better than random guessing but its usability is greatly undermined in a professional sphere where there is rapid flow of information. A model based on quantitative and qualitative assessments are better tools to predict economically distressed countries than the simple theoretical prediction or early warning system models.

In the end the conclusion is that predicting sovereign debt default is not easy. Many authors in the past have tried to develop and are currently trying to develop a correct prediction model. The paper concludes by pointing out an area of future research which can lead to development of a comprehensive debt default probability estimation model. The model needs to combine various non-model and qualitative factors with traditional quantitative solvency, liquidity and macroeconomic factors. Maybe this model can correctly answer the question “Is it possible to predict sovereign debt default/crises?”

Item Type: Dissertation (University of Nottingham only)
Depositing User: EP, Services
Date Deposited: 10 Nov 2011 14:54
Last Modified: 12 Oct 2017 18:22

Actions (Archive Staff Only)

Edit View Edit View