Publication: Modelos de medición del riesgo de crédito
Loading...
Files
Official URL
Full text at PDC
Publication Date
2017-01-17
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad Complutense de Madrid
Citations
Exportar
Abstract
El posible impago por parte de prestatarios y contrapartes eleva a un primer plano el riesgo de crédito para las instituciones financieras. Existen diferentes técnicas para estimar la probabilidad de insolvencia empresarial durante un determinado periodo de tiempo futuro: modelos basados en los precios de los bonos (denominados modelos reducidos), modelos basados en datos históricos, modelos basados en los precios de las acciones (denominados modelos estructurales) y modelos actuariales. El objetivo de este trabajo es realizar un análisis crítico y comparativo de los modelos de estimación de la probabilidad de impago, identificando sus ventajas y limitaciones, de forma que nos permita determinar cuándo es preferible utilizar uno u otro modelo. Asimismo, analizamos las herramientas técnicas (derivados de crédito) y legales (mecanismos contractuales) existentes para eliminar o mitigar el riesgo de crédito. Los modelos reducidos y los modelos basados en datos históricos se basan en la calidad crediticia de una compañía. Sin embargo, las calificaciones crediticias se revisan con cierta periodicidad que puede resultar escasa. Esto hace que se cuestione si los precios de las acciones pueden proporcionar información más actualizada para estimar probabilidades de impago, dando origen a los modelos estructurales. Las probabilidades de impago implícitas en los precios de los bonos son probabilidades riesgo–neutro, dado que se asume un escenario sin riesgo, mientras que las probabilidades extraídas de datos históricos (frecuencias relativas) son reales. Las probabilidades de impago riesgo–neutral son significativamente más altas que las probabilidades reales. Esta discrepancia se explica por varias razones...
The possibility of default of borrowers and counterparties puts credit risks at the forefront of financial institutions’ concerns. There are various techniques for estimating the probability of corporate insolvencies during a set future timeframe: models based on bond prices (so-called simplified models), models based on historical data, models based on stock prices (so-called structural models) and actuarial models. The objective of this research is to carry out a critical and comparative analysis of models for estimating the probability of default, identifying their advantages and shortcomings and allowing us to determine when it is preferable to use one or another of them. Thus, we analyse existing technical (credit derivatives) and legal (contractual mechanisms) tools for eliminating or mitigating credit risks. Simplified models and those models based on historical data are underpinned by the credit quality of a company. Yet, credit ratings are revised at regular intervals which may turn out to be too infrequent. This leads us to ask whether or not stock prices may provide more up-to-date information for forecasting probabilities of default, giving rise to structural models. The probabilities of default implicit in bond prices are risk–neutral since they assume a risk–free scenario, while probabilities extracted from historical data (relative frequencies) are real. Risk-neutral probabilities of default are significantly greater than real ones. There are several reasons which explain this discrepancy...
The possibility of default of borrowers and counterparties puts credit risks at the forefront of financial institutions’ concerns. There are various techniques for estimating the probability of corporate insolvencies during a set future timeframe: models based on bond prices (so-called simplified models), models based on historical data, models based on stock prices (so-called structural models) and actuarial models. The objective of this research is to carry out a critical and comparative analysis of models for estimating the probability of default, identifying their advantages and shortcomings and allowing us to determine when it is preferable to use one or another of them. Thus, we analyse existing technical (credit derivatives) and legal (contractual mechanisms) tools for eliminating or mitigating credit risks. Simplified models and those models based on historical data are underpinned by the credit quality of a company. Yet, credit ratings are revised at regular intervals which may turn out to be too infrequent. This leads us to ask whether or not stock prices may provide more up-to-date information for forecasting probabilities of default, giving rise to structural models. The probabilities of default implicit in bond prices are risk–neutral since they assume a risk–free scenario, while probabilities extracted from historical data (relative frequencies) are real. Risk-neutral probabilities of default are significantly greater than real ones. There are several reasons which explain this discrepancy...
Description
Tesis inédita de la Universidad Complutense de Madrid, Facultad de Ciencias Económicas y Empresariales, Departamento de Economía Financiera y Contabilidad I (Economía Financiera y Actuarial), leída el 11/12/2015