Home About IUP Magazines Journals Books Archives
     
A Guided Tour | Recommend | Links | Subscriber Services | Feedback | Subscribe Online
 
The IUP Journal of Financial Risk Management
Basel II Capital Requirement Sensitivity to the Definition of Default
:
:
:
:
:
:
:
:
:
 
 
 
 
 
 
 

The paper is motivated by a disturbing observation according to which the outcome of the regulatory formula significantly depends on the definition of default used to measure the Probability of Default (PD) and the Loss Given Default (LGD) parameters. Basel II regulatory capital should estimate, with certain probability level, the unexpected credit losses on banking portfolios, and so it should not depend on a particular definition of default that does not change real, historical and expected losses.This paper provides an explanation of the phenomenon based on the Merton default model and tests it using a Monte Carlo simulation. Moreover, it develops an analytical method to model LGD unexpected risk and to combine it with the PD unexpected risk. The developed formula and, in particular, its simplified version could be used to improve the current regulatory formula.The analysis, at the same time, provides a different insight into the issue of regulatory capital sensitivity to the definition of default. Finally, the paper performs a structural model-based simulation to test the hypothesis according to which scoring functions developed with a soft definition of default have weaker predictive power than the ones developed with a hard definition of default.

 
 
 

where UDR is a regulatory estimate of the unexpected default rate, PD is the estimated average default rate, and LGD is the loss given default. The resulting percentage of capital requirement (C) is multiplied by EAD, the exposure of the receivable at default (in most cases equal to the actual outstanding amount), to get the account level capital requirement in absolute terms. The calculation is done at account level, but the result should reflect only the account's contribution to unexpected risk within a well-diversified portfolio. The UDR is given by a regulatory formula that depends only on PD (other key parameters of the formula, i.e., correlation and probability level are set by the regulator as given constants, with a possible adjustment depending on PD in the case of correlation). In other words, the capital requirement C is a function of PD and LGD,

The logic of Formula 2 is to give an estimation of unexpected systematic loss relative to the expected loss at 99.9% probability level. This goal is mainly achieved through the first part of the formula, UDR(PD) - PD, that models the difference between the unexpected default rate value and the expected mean default rate. On the other hand, the LGD parameter is defined in principle as a conservative estimate of the long-term average loss rates on defaulted loans. The regulator in addition, requires the banks to rather vaguely reflect the economic downturn conditions, dependence on PD or other factors to capture the relevant risks. Nevertheless under normal circumstances, the long-term average is considered as satisfactory. In that case, the possibility of an unexpectedly high loss rate is not captured by the formula at all.

In practice, the parameters PD and LGD are obtained from historical data on homogenous (in terms of product, segment and rating) sets of receivables. The values strongly depend on the definition of default, which determines the number of loans marked as defaulted. The definition of default must satisfy certain regulatory conditions, but the banks have a significant discretion in its implementation. Assume that a bank suffered a total loss L on its historical portfolio P and recorded a number of defaults, ND. Some of the defaulted receivables might have been recovered fully, while others contributed to the total loss L. If = V/N (where V is the total portfolio P volume and N is the total number of receivables) is the average exposure of the observed receivables and = L/ND is the average loss on defaulted receivables, then the loss given default parameter can be actuarially estimated as LGD = /. Similarly, PD can be estimated as ND/N. Consequently, we get that the product PD.LGD = L/V is the average percentage of observed loss rate, LR = L/V, on the portfolio, which does not depend on the definition of default.

 
 
 

Financial Risk Management Journal, capital requirement, well-diversified portfolio, LGD parameter, CreditMetrics model, terms of product, fundamental financial data, development of scoring functions, clients marked as defaulted, financial difficulties, Regulatory capital requirements, credit risk, The credit economic capital, credit Value at Risk, Portfolio Manager