School of Business publications portal
This portal is no longer updated. Aalto University School of Business Master's Theses are now in the Aaltodoc publication archive (Aalto University institutional repository)
School of Business | Department of Finance | Finance | 2014
Thesis number: 13603
Basel II correlation estimates in light of empirical evidence and literature
Author: Vänttinen, Aleksi
Title: Basel II correlation estimates in light of empirical evidence and literature
Year: 2014  Language: eng
Department: Department of Finance
Academic subject: Finance
Index terms: rahoitus; financing; rahoitusmarkkinat; financial markets; pankit; banks; luotto; credit; valvonta; control; riskienhallinta; risk management
Pages: 68
Key terms: Basel, asset value correlation, Internal-ratings based approach
Basel II rules put forward by Basel Committee on Banking Supervision introduced a more risk-sensitive approach to banks to calculate regulatory capital requirements. This more risk-sensitive framework includes an option for banks to apply for permission to use Internal-ratings based approach in which the bank can estimate some parameters by itself without resulting to supervisory estimates. The framework for calculating the capital requirement for exposures in corporate portfolio includes a correlation function in which the bank's own estimate of probability of default would be the only variable.

However, the correlation function used in Basel's internal ratings-based approach is not based on a strong theoretical foundation but instead relies on few simple assumptions according to which the function was formed and the calibration was done and later revised. The paper aims to analyze some of the limitations the function induces to the framework by first going through the existing critique in the literature and then calculating asset correlations implied by loss distributions of different industries and sectors and drawing from sensitivity analysis of risk-weighted assets and loan margin.

The resulting asset value correlations calculated from the empirical dataset in this paper are close to what previous studies on default data have found to be the empirical level. The dataset is based on the ratio of gross impairment to loans and guarantees. Furthermore the feasibility of the results is assessed through three hypotheses which all assist in showing the feasibility of the chosen approach. The results show that asset value correlations and both expected and unexpected losses are different among industries.
Master's theses are stored at Learning Centre in Otaniemi.