Author
Soytaş, Mehmet Ali, Volkan, E.
Publication Date
2016
Publication Place
-
Okan University
Subject
External debt bankruptcy risk, Hotz-Miller Estimation Technique, Endogenous bankruptcy risk, Foreign debt, Sovereign default risk, Hotz-Miller Estimation Technique, Endogenous default risk, External deb
Type
Periodical
Language
Turkish
Digital
Yes
Manuscript
No
Library
Özyeğin University
Library Asset ID
1307-7112
Record ID
d4843ef4-6f36-4696-927d-ebba7d79e153
Library Location
Economics
Date
2016
Sample Text
This project reveals the necessary theoretical implications for the adaptability of the Hotz-Miller estimation technique (Hotz and Miller 1993), which is frequently used in the applied microeconometrics literature, to foreign debt bankruptcy models. Foreign debt bankruptcy models are solved numerically using the "calibration and value function iteration" technique, based on the fixed-point theorem. This theory and solution method limits the performance of the models in the quantitative field due to the calculation-based technical constraints it creates and deprives them of the feature of quantitative prediction. While the Hotz-Miller technique estimates the structural parameter values of discrete choice recursive competitive partial (general under certain conditions) equilibrium models using the real probability of the endogenous decision, which we call discrete choice, and other necessary real data, it saves the model from the fixed point theorem, which imposes great constraints on its solution. In this article, it is theoretically shown that by applying reverse engineering to this technique, the preliminary bankruptcy risk and bankruptcy probability functions corresponding to the structural parameter values estimated from the business cycle models of the countries can be estimated. Using this theoretical framework, the qualitative results obtained in the foreign debt bankruptcy literature can be produced exactly by using the model country's income and external debt level in the bankruptcy probability function. With this technique, the impact areas of the models are increased and external debt general equilibrium models become open to development., This project adopts the Hotz-Miller estimation technique (Hotz and Miller 1993), often used in applied micro-econometrics literature, into sovereign default models. Using the fixed-point theorem, sovereign default models are solved by "calibration and value function iteration" method, which due to its computational constraints, greatly limits the models' quantitative performance and completely foregoes its quantitative projection ability. Hotz-Miller technique estimates the structural parameter values of recursive competitive partial (under some conditions general) equilibrium models with discrete choice, given the actual probability of the (discrete) endogenous choice and actual data. In this paper, by reverse engineering the technique, given the structural parameter values estimated from business-cycle models, it is theoretically shown that the ex-ante default probabilities can be estimated. The estimated default probabilities are functions of income and debt level, characteristically displaying a one-to-one similarity to those in the default literature. Additionally, the simulation results, such as average default probability, risk spread, debt-to-output ratio and a number of business cycle statistics, can be obtained from model estimates. In this respect this paper contributes to default literature with a new estimation technique. Moreover, with this technique, not only the computational constraints of the fixed-point theorem will be bypassed but also the quantitative inference ability of sovereign default models will be improved.
Cilt
53
