No abstract

The recent crisis underlined that proper estimation of distress-dependence amongst banks in a global system is essential for financial stability assessment. We present a set of banking stability measures embedding banks' linear (correlation) and nonlinear distress-dependence, and their changes through the economic cycle, thereby allowing analysis of stability from three complementary perspectives: common distress in the system, distress between specific banks, and cascade effects associated with a specific bank. Our approach defines the banking system as a portfolio of banks and infers its multivariate density from which the proposed measures are estimated. These can be provided for developed and developing countries.

No abstract

The concept of risk-based capital requirements enjoys widespread support. Effective implementation, however, requires that risk be measured accurately both across borrowers and across time. Under the New Capital Accord, the cornerstone of this risk measurement process is the rating of the borrower. In this paper we use the ratings assigned by individual Mexican banks to examine how measured credit risk for these banks has changed since the financial crisis in the mid-1990s. We then examine the implications of these changes in risk for regulatory capital under the proposed changes to the Basel Capital Accord. We find that measured risk increased after the crisis and then fell as the recovery took hold. In turn, despite the limitations of the data, we find that the proposed internal ratings-based approach would have generated large swings in regulatory capital requirements over the second half of the 1990s, with required capital increasing significantly in the aftermath of the crisis, and then falling as the economy recovered. Looking forward, if movements in actual bank capital were to show this same cyclical variation, then business cycle fluctuations might be amplified by developments in the banking industry.

No abstract