IX. Overview of EAD Estimation Concepts
Walter Gruber and Ronny Parchert
1 PLUS i GmbH
1. EAD Estimation from a Regulatory Perspective
1.1. Definition of Terms
The exposure at default (EAD) is defined as the expected amount of a receivable
at the time of a defaul麦片的功效与作用
t. In order to describe the borrower related risk, the EAD has
to be calculated without considering provisions.1 Bank-internally available provi-
sions rve to cover the equity in the balance sheet in ca of loss. Possible
loss already have reduced the risk bearing capacity of the bank at the moment of
risk identification by the realization of provisions.
The definition shows that in a first step, the EAD is determined by the exact time
of default. If obrved economically, the expected EAD changes depending on
whether the default horizon consists of, e.g. one or of two years. According to
regulatory prescriptions, the EAD must not be lower than the book value of a bal-
ance sheet receivable.2 Therefore a regulatory necessity to estimate future EAD
for such positions is not given.
Credit Conversion Factors (CCF) have to be estimated for off-balance sheet trans-
actions and credit approvals. They describe the percentage rate of undrawn credit
lines (UCL) that have not yet been paid out, but that will be utilized by the bor-
rower until the default happens. Therefore the EAD for an undrawn amount is de-
fined as:
EAD (1)
CCF
UCL
Concerning credit lines that have not completely been paid out (balance sheet re-
ceivables); the EAD for a one year time horizon is defined as illustrated in Figure
1.
1 See BIS (2004), 308.
2 See Bundesbank (2005), 104.
178 Walter Gruber and Ronny Parchert
Figure 1. Differences between exposure and credit approval
1.2. Regulatory Prescriptions Concerning the EAD Estimation Regulatory requirements concerning estimations of loss parameters and related to that also the EAD are mainly given in the regulatory framework of Bal II. Within the equity provisioning requirements that are defined here, three parate approaches for the calculation of risk weighted asts are distinguished. As far as the standardized approach (SA) and the foundation internal ratings-bad ap-proach (FIRB) are concerned there is no tolerance regarding the estimation of the EAD/CCF. This is due to the fact that
the CCF in relation to class of receivables is prescribed by regulatory bodies. Specific minimum requirements are defined concerning the advanced internal ratings-bad approach (AIRB):3
x“EAD for an on-balance sheet or off-balance sheet item is defined as the ex-pected gross exposure of the facility upon default of the obligor. For on-balance sheet items, banks must estimate EAD at no less than the current drawn amount, subject to recognizing the effects of on-balance sheet netting as speci-fied in the foundation approach…”
x“Advanced approach banks must have established procedures in place for the estimation of EAD for off-balance sheet items. The must specify the esti-mates of EAD to be ud for each facility type. Banks estimates of EAD should reflect the possibility of additional drawings by the borrower up to and after the time a default event is triggered. Where estimates of EAD differ by facility type, the delineation of the facilities must be clear and unambiguous.”
3 See BIS (2004), 47
4 ff.
IX. Overview of EAD Estimation Concepts 179 x“Advanced approach banks must assign an estim
ate of EAD for each facility. It must be an estimate of the long-run default-weighted average EAD for similar facilities and borrowers over a sufficiently long period of time, but with a mar-gin of conrvatism appropriate to the likely range of errors in the estimate. If a positive correlation can reasonably be expected between the default frequency and the magnitude of EAD, the EAD estimate must incorporate a larger margin of conrvatism. Moreover, for exposures for which EAD estimates are volatile over the economic cycle, t中国青年作文
he bank must u EAD estimates that are appropriate for an economic downturn, if the are more conrvative than the long run av-erage. For banks that have been able to develop their own EAD models, this could be achieved by considering the cyclical nature, if any, of the drivers of such models. Other banks may have sufficient internal data to examine the im-pact of previous recession(s). However, some banks may only have the option of making conrvative u of external data.”
x“The criteria by which estimates of EAD are derived must be plausible and in-tuitive, and reprent what the bank believes to be the material drivers of EAD.
The choices must be supported by credible internal analysis by the bank. The bank must be able to provide a breakdown of its EAD experience by the factors it es as the drivers of EAD. A bank must u all relevant and material infor-mation in its derivation of EAD estimates. Across facility types, a b
ank must review its estimates of EAD when material new information comes to light and at least on an annual basis.”
x“Due consideration must be paid by the bank to its specific policies and strate-gies adopted in respect of account monitoring and payment processing. The bank must also consider its ability and willingness to prevent further drawings in circumstances short of payment default, such as covenant violations o一直的意思
r other technical default events. Banks must also have adequate systems and proce-dures in place to monitor facility amounts, current outstandings against com-mitted lines and changes in outstandings per borrower and per grade. The bank must be able to monitor outstanding balances on a daily basis.”
Apart from that specific minimal estimation, periods are defined concerning class of receivables. Regulatory prescriptions clearly show the high qualitative and quantitative demands banks have to meet when using the AIRB. In practice, the utilization of a bank’s internal model to bring the estimation method for the EAD into unison with the other loss parameters probability of default (PD) and loss given default (LGD), is independent of the question whether the model is util-ized for internal or for regulatory requirements or for both.
1.3. Delimitation to Other Loss Parameters
Next to a clear definition of the specific parameters and the best possible data quality, a uniform definition of default is most important for a methodologically correct internal estimation of loss parameters for the credit risk. The first primary loss parameter is the PD which describes the probability of default of a borrower
180 Walter Gruber and Ronny Parchert
within a predefined period, usually one year. The statistical methods for estimat-ing the PD are described in Chapters I, II, and III.
The estimation period of the PD is identical with the estimation period of the EAD, in general one year. The difference between tho two parameters lies in the data ba needed for estimation. When estimating the PD, it is necessary to fore-cast how many of the customers who are alive today will default over the next pe-riod. Therefore, the overall portfolio has to be considered. Concerning the CCF es-timation, the data ba is reduced and only tho credit lines where the default took place within the period of obrvation can be regarded ex post. In an eco-nomic point of view con团队简介范文
cerning amortization effects, all receivables have to be taken into account for the EAD estimation.
LGD describes the rate of the defaulted amount of receivables (EAD) that leads to a loss for the creditor. It is the third major component of credit risk and expected loss (EL), which is computed as EL = EAD PD LGD. The LGD estimation, similar to the estimation of CCF, depends on the defaults that have already taken place. Only on the basis of the defaulted receivables, can it be measured empiri-cally which part of the default-volume will lead to an economic loss for the bank. The estimation of the empirical LGD suffers from the relatively limited amount of data and additionally from the long duration of the estimation period. Whereas PD and EAD/CCF with an estimation horizon of one year already assume a very long period compared to market price risk parameters, LGD estimation periods on av-erage reach 3 to 5 years. The experience within banks shows that the administra-tion of defaulted engagements - including the realization of collateral – takes a long time. Due to this, the backtest of LGD estimations becomes increasingly dif-ficult or nearly im网络卡怎么办
possible. Figure 2 shows the relations between EAD-CCF-LGD in a scheme.
The figure shows that LGD always refers to EAD. Concerning estimations, a sim-ple motto might be derived from this fact. Everything that takes place until the de-fault happens has to be considered in the EAD. All payments after this event only influence the LGD.
IX. Overview of EAD Estimation Concepts 181
Figure 2. Relation between EAD-CCF-LGD
1.4. EAD Estimation for Derivative Products
If we examine 升旗仪式
derivative products like interest rate swaps, caps, floors, swaptions, cross currency swaps, equity swaps, or commodity swaps; two kind of counter-party risks have to be considered: ttlement and pre-ttlement risk. Settlement risks occur if the payments are not synchronous: This is for example the ca if bank A has paid a EUR cash flow in a cross currency swap to bank B before it has received the USD cash flow. Here the risk consists in missing the USD cash flow. If ban
k B defaults, a loss of the amount of this cash flow would occur. Settlement risks are mostly of short term character.
Much more imp橘子老虎
ortant are the pre-ttlement risks. Characteristic for pre-ttlement risks is the following situation: Bank A expects in the future a rising in-terest rate. For hedging the loan portfolio, bank A makes a payer swap with bank B. This transaction eliminates the interest rate risk but creates a counterparty risk. If bank B defaults during古情诗
the lifetime of the interest rate swap, bank A has to look for a new counterparty to make the same payer swap 猪跟什么属相配
with this new counterparty. If in the meantime, the interest rate has moved up, the replacement with an identi-cal swap will only be possible by paying an upfront payment to the new count青云志剧情
er-party. Pre-ttlement risks are of a long term nature becau they may occur dur-ing the whole life time of the derivative product.
When calculating the EAD for derivative products, it has to be noted that the EAD consists of two parts:
