A CREDIT RISK MODEL FOR ALBANIA
Hilda Shijaku
Bank of Albania
Kliti Ceca
Bank of Albania
ABSTRACT
Our methodology takes into account existing methods of stress testing for credit risk for the system as a whole, and adapts them to the Albanian financial intermediation features and data. The model starts by testing the presence of a system of equations (stationary VAR) capturing the joint behaviour of financial and real variables. The equation is modelled to contain some dynamics allowing for the persistence of shocks over more than one period given the backward looking nature of the chosen proxy for the default rate. In the second step, by using Monte Carlo simulations we propose generating the distribution of losses of the default rates. We relax the assumption of zero error terms in the reduced form estimates in step one and simulate univariate/multivariate normally distributed vectors of error terms via Monte Carlo methods, where the information on the variance/covariance matrix is obtained from the reduced form equations in step one. Next, we introduce a shock to the nonfinancial variable (growth rate, interest rates, exchange rates etc.) and similarly compute a range of values for the default rate in the stressed scenario.
JEL classification: C51, E58.
Keywords: stress testing, credit risk
Acknowledgments: We would like to thank the participants of the 3rd SEE Workshop for a useful exchange discussion. Special thanks are due to our discussant Faidon Kalfaoglou for his valuable comments and suggestions. The views expressed in this paper are those of the authors and do not necessarily reflect those of the Bank of Albania and the Bank of Greece. We alone are responsible for the remaining errors and omissions.
Correspondence:
Hilda Shijaku
Financial Stability Department
Kliti Ceca
Department of Research
Bank of Albania, “Sheshi Skënderbej”, Nr.1, Tirana, Albania.
Email: hshijaku@bankofalbania.org and kceca@bankofalbania.org