Συλλογές | |
---|---|
Τίτλος |
Credit risk measurement: applying historical stress tests on a bank’s bond portfolio |
Δημιουργός |
Χρηστάκου, Πολυτίμη |
Συντελεστής |
Οικονομικό Πανεπιστήμιο Αθηνών, Τμήμα Λογιστικής και Χρηματοοικονομικής Χαλαμανδάρης, Γεώργιος |
Τύπος |
Text |
Φυσική περιγραφή |
87p. |
Γλώσσα |
en |
Περίληψη |
Market history has demonstrated that periods of intense economic turmoil usually reveal the financial system’s vulnerabilities as well as the weaknesses regarding the approaches followed for the management of the risks entailed in the banking system. One of the market’s theories that relates to these shortcomings is the “too big to fail” notion denoting that large banking corporations are so highly interconnected that their failure would endanger the national or global economy. Unforeseen banking crises usually lead to sovereign crises around the world which are often followed by one or more government debt restructurings provoking severe losses in the banking portfolios. Below, we present a brief analysis of the recent Greek financial crisis until the agreement concerning the Private Sector Involvement(PSI) on March 9th, 2012. Following, we offer a review of additional three financial crises of historical significance; Russia’s Argentina’s and Uruguay’s. Thereafter, follows a thorough examination of the fundamental notions regarding risk management. We will focus on the credit risk management and what Basel framework suggests that Banks shall hold against their bond portfolio. Furthermore, we explore the Value-at-Risk measure and the stress testing approach since these are worthwhile tools for the identification of potential risks and weaknesses in the banking portfolios. We also present the basic models prevailing in the banking market for the measurement and management of credit risk. Extensive reference is made to thecomponents of economic capital as defined by the Basel framework and the international literature. Of particular interest is the Markov property which relates to the construction of transition matrices. These matrices depict the probabilities that a corporation or government stays on the current rating status or changes to another (i.e. downgrade or upgrade). In this study, our ultimate aim is to perform severe historical stress tests in order to estimate the bank’s bond portfolio worst losses. To this end, we employ the methodology proposed by S. Varotto (2011), i.e. we increase the investment horizon from one year, as proposed by current regulation, to three years. The idea behind this suggests that it will take more than one year for a bank to close its non-performing exposures and stop losses or to raise new capital to meet further losses. We include the default risk accompanied with migration risk by employing Moody’stransition matrices in order to derive the cumulative and thereafter the conditional probabilities of default given by the no default in an earlier period. Our additional objective is to compare our results with the realized losses. To this end, we select a Greek bank’s representative portfolio three years before the Greek debt’s restructuring (PSI, 9/3/2012) and we perform three historical stress tests of different severity. Thereafter, we compare our estimated losses to the realized losses. We will see that, depending on the historical scenario applied the losses estimated differ substantially, due to the recovery rates employed. We understand that the diversification of a portfolio may lead to the absorption ofsubstantial losses occurring due to country or other risks and we also realize that applying severe scenarios is not sufficient for the implementation of a stress test; plausibility and suggestiveness of risk-reducing are also significant factors that a risk manager should take into consideration. |
Λέξη κλειδί |
Stress tests Basel III Accord Value at Risk (VaR) Credit risk Basel II Accord Risks Sovereign defaults |
Ημερομηνία |
31-07-2013 |
Άδεια χρήσης |
https://creativecommons.org/licenses/by/4.0/ |