Jump to Content Jump to Main Navigation
Gleeson on the International Regulation of Banking, 3rd Edition by Gleeson, Simon (30th August 2018)

Part II Commercial Banking, 10 The Internal Ratings-Based Approach

From: Gleeson on the International Regulation of Banking (3rd Edition)

Simon Gleeson

From: Oxford Legal Research Library (http://olrl.ouplaw.com). (c) Oxford University Press, 2015. All Rights Reserved. Subscriber: null; date: 17 September 2019

Subject(s):
Banks as a lender — Regulation of banks — Credit risk — Default and credit — Financial regulation — Basel accords

(p. 187) 10  The Internal Ratings-Based Approach

10.01  The internal ratings-based (IRB) approach permits a bank to use its internal models to derive risk weights for particular exposures. There are two available bases for the IRB: foundation (F-IRB), which permits the bank to model probability of default (PD), but relies on regulatory standard figures to determine loss given default (LGD) and exposure at default (EAD), and advanced (A-IRB), in which all three of these are modelled. The A-IRB approach models PD, LGD, EAD, and maturity (M). Both IRB approaches model both expected loss (EL) and unexpected loss (UL), and IRB banks are expected to recognize the EL derived from their models in their capital calculations—consequently a bank using an IRB approach will generally have a different total capital level from that which it would have if it were an SA (standardized approach) bank.

10.02  As noted above, in practice the IRB approach is not used to attribute an individual probability of loss to each exposure. Banks use their IRB models to create an internal rating system. Basel sets out criteria for these rating systems; in particular (p. 188) they must take into account two distinct dimensions: the risk of borrower default (PD) and transaction-specific factors (factors which potentially affect LGD), although these may be merged in respect of specialized lending. Banks’ rating systems must be appropriately and fully documented.

10.03  The building blocks of the IRB are the asset categories. Each asset category uses a different methodology to ascribe risk weightings to exposures, and in many cases the attribution of a particular asset to one class rather than another will have an impact on the amount of regulatory capital it requires. The asset classes are shown in Table 10.1.

Table 10.1  Basel II Exposure Sub-classifications

Classification

Sub-classifications

Corporate exposures

Specialized lending (SL), comprising:

Project Finance (PF)

Object Finance (OF)

Commodities Finance (CF)

Income-producing Real Estate (IPRE)

High-volatility Commercial Real Estate (HVCRE)

Sovereign exposures

Bank exposures

Retail exposures

Exposures secured by residential properties

Qualifying revolving exposures

Equity exposures

Eligible purchased receivables

Retail receivables

Corporate receivables

A.  Basel III and IRB

10.04  One of the most important aspects of the Basel III settlement is that it prohibits the A-IRB approach from being used in respect of large corporates and banks. The logic of this is fairly straightforward—exposures of this kind have very low default rates, which means that there is very little default data which can be used either to construct or to verify models; consequently the higher the quality of the credit, the lower the level of confidence that can be placed in the model. Thus these exposures are required to be risk weighted using the foundation IRB or standardized approaches. This means that the risk requirements for these exposures will almost certainly rise—the regulatory weightings prescribed for the relevant classes of activity under F-IRB and standardized are set to capture a broad average of corporate risks within specified rating levels, whereas the exposures under discussion here are very low-risk assets. This may therefore disincentivize banks from holding such assets. The A-IRB prohibition is in respect of:

(p. 189)

  • •  exposures to corporates belonging to a group with total consolidated revenues of EUR 500m or more; and

  • •  exposures to any bank, securities firm, or financial institution (including insurance firms).

B.  Corporate, Sovereign, and Bank Exposures

10.05  Under Basel II, all of these exposures are weighted using the same approach—this approach will therefore apply to the large majority of exposures of an IRB bank. Basel III imposes a different and higher weighting on some bank exposures, which is set out at para 8.38 ff.

10.06  The definitions of ‘corporate’, ‘sovereign’, and ‘bank’ (‘institution’ in the EU, to include investment firms) correspond with those used for the standardized approach. Thus sovereigns includes entities designated as sovereigns and public sector entities under the standardized approach, along with those multilateral institutions which attract a 0 per cent weighting under that approach. A corporate exposure for this purpose includes any exposure to a firm apart from ‘specialized lending’ (see later).

10.07  A key step in estimating any IRB parameter is preparing the reference data set (RDS) for that parameter. This involves a variety of challenges, including choosing the sample size, the length of time series, reliance on external data, the treatment of defaulted positions that have generated no loss, and the length of recovery processes (incomplete workout).

PD

10.08  The PD for all three types of exposure is calculated in the same way under Basel II, and is the probability of default within a twelve-month period. However, for bank and corporate exposures a 0.03 per cent floor is imposed, so that only sovereign exposures may be treated as absolutely risk-free in this respect. The PD of an obligor in default must be 100 per cent—ie the PD may not reflect the possibility of the default being cured. Basel III will raise this minimum to 0.05 per cent.

LGD

10.09  The primary difference between foundation and advanced IRB banks is that foundation IRB banks must use the prescribed figures for LGD, whereas an advanced IRB bank may model it. LGD is an important component of risk weighting. Given the fundamental equation that:

Part II Commercial Banking, 10 The Internal Ratings-Based Approach

(p. 190) it will be seen that any change in LGD changes the risk-weighted assets (RWA) proportionately—halve LGD and you halve RWA; double LGD and you double RWA. Thus the calculation and attribution of LGD figures are sensitive and important.

Foundation IRB

10.10  Under the foundation IRB approach, LGDs are prescribed by the regulator according to asset type. The table of applicable LGDs is shown in Table 10.2.

Table 10.2  Basel II F-IRB Prescribed LGD

Exposure Type

Prescribed LGD

Covered bonds

12.5%

Senior exposures

45%

Subordinated exposures

75%

Equity exposures

90%

The key issue here is the distinction between ordinary, senior, and subordinated exposures. In some circumstances (where there is express subordination in the constitutive documents) the point will be free from doubt, but issues such as economic subordination (senior unsecured debt owed by a borrower whose assets are pledged as security to other creditors) and structural subordination (senior unsecured debt owed by a company whose assets are held in a subsidiary) will create difficulties in implementation.

Foundation IRB collateral

10.11  Under the foundation IRB approach, collateral is recognized by an adjustment of the prescribed LGD figure. F-IRB banks may recognize slightly broader classes of collateral (notably receivables and certain specified commercial and residential real estate (CRE/RRE) and other assets), and give effect to that collateral by applying the haircuts specified in the comprehensive approach. This results in a revised LGD figure (LGD*) which can be expressed as

Part II Commercial Banking, 10 The Internal Ratings-Based Approach

where E* is the haircutted value of the collateral and E is the value of the exposure. Note that in the F-IRB approach, collateral does not affect EAD.

Basel III final—prescribed LGDs

10.12  Under Basel III, these figures will be amended as shown in Table 10.3.

Table 10.3  Basel III F-IRB Prescribed LGDs

Exposure Type

Prescribed LGD

Senior exposures to corporates

40%

Senior exposures to banks and financial firms

45%

Subordinated exposures to corporates and banks

75%

10.13  Also, Basel III refines the approach to the recognition of collateral in the F-IRB. A set of revised LGDs are provided for exposures secured by particular types of collateral as shown in Table 10.4.(p. 191)

Table 10.4  Basel III Prescribed Collateralized LGDs

Type of Collateral

LGD

Haircut

Eligible financial collateral

As applied in the comprehensive approach, adjusted for non-daily margining/revaluation

Eligible receivables

20%

40%

Eligible residential/commercial real estate

20%

40%

Other eligible physical collateral

25%

40%

Ineligible collateral

n/a

100%

These are then applied according to the following formula:

Part II Commercial Banking, 10 The Internal Ratings-Based Approach

where:

  1. (1)  E is the current value of the exposure (ie cash lent or securities lent or posted). In the case of securities lent or posted the exposure value has to be increased by applying the appropriate haircuts (HE) according to the comprehensive approach for financial collateral.

  2. (2)  ES is the current value of the collateral received after the application of the haircut applicable for the type of collateral (Hc) and for any currency mismatches between the exposure and the collateral. ES is capped at the value of E × (1 + HE).

    Part II Commercial Banking, 10 The Internal Ratings-Based Approach

  3. (3)  LGDU = the LGD applicable for an unsecured exposure.

  4. (4)  LGDS = the LGD applicable to exposures secured by the type of collateral used in the transaction.

10.14  Where currencies are mismatched, the haircut is the same as that applied in the comprehensive approach.

(p. 192) Advanced IRB

10.15  Where LGD is modelled by the bank, the LGD calculation should be based on the definitions of default and economic loss used by the institution, which should be consistent with the provisions contained in the CRD. LGD estimates should reflect the experience and practices of the individual institution. This means in practice that institutions cannot rely on industry-wide estimates without adjusting them to reflect their own position where necessary.

10.16  An advanced IRB bank’s LGD model should generally take collateral into account. Advanced IRB banks reflect collateral through their LGD modelling rather than through the collateral framework set out in Chapter 11.

10.17  Estimated LGDs are assigned to current facilities (both defaulted and non-defaulted) and used to calculate capital requirements for its exposures. Estimated LGDs are based on the realized LGDs for the applicable RDS. However, estimated LGDs are likely to differ from the historic average realized LGDs in the RDS because the former need to incorporate expectations of future recovery rates. This involves calculating a long-run forward-looking recovery rate for the facility grade or pool, taking both current and future economic circumstances into account. The institution should produce an LGD estimate appropriate for an economic downturn (‘downturn LGD’) if this is more conservative than the long-run average.

10.18  In addition to expected LGD, an institution must also produce a best estimate of estimated loss for defaulted exposures given current economic circumstances and exposure status. The difference between this amount and the estimated LGD derived from the RDS stems from the possibility of additional losses during the recovery period, and represents the unexpected loss capital requirement for the defaulted exposure.

Basel III final—LGD floors

10.19  Basel III imposes floors on modelled LGD numbers for corporate exposures. For unsecured exposures the minimum is 25 per cent. For secured exposures, the floor varies according to the type of collateral as set out in Table 10.5.(p. 193)

Table 10.5  Basel III Floor for Modelled LGDs

Type of Collateral

Floor

Financial collateral

0%

Receivables

10%

Commercial/residential real estate

10%

Other physical

15%

Downturn LGDs

10.20  Conditional expected loss is calculated by applying a function to the PD figure and by using a ‘downturn’ LGD figure. The key to the idea of the ‘downturn’ LGD is that the regulator was either unwilling or unable to apply a meaningful function to LGD—thus whereas UL PD is the result of a complex formula, banks are able to use their own unfettered discretion to estimate ‘downturn’ LGD. Thus the full formula is:

Part II Commercial Banking, 10 The Internal Ratings-Based Approach

One of the more difficult discussions around LGD is the issue relating to ‘downturn LGDs’. An institution is required to use LGD estimates that are ‘appropriate for an economic downturn’ if those are more conservative than the long-run average. The basis for this requirement is the idea that LGD for a portfolio of loans fluctuates according to the economic conditions, such that in a downturn LGDs would be larger overall—a principle which could be articulated as ‘a falling tide sinks all boats’. Possibly more importantly, it is likely that an economic downturn may both increase probability of default (PD) and increase LGD. Accordingly, LGD parameters need to be calculated assuming conditions where credit losses are substantially higher than the long-run average. Under such conditions default rates are expected to be high, so that if recovery rates are negatively related to default rates, LGD parameters must embed forecasts of future recovery rates that are lower than those expected during more neutral conditions. It is, however, open to an institution to demonstrate to its regulator on the basis of its data and models that in particular sectors recovery rates are expected to be independent of future default rates. If this can be done then there is no requirement to use higher LGD figures.

LGD—Loss data

10.21  The RDS for LGD should include only exposures to defaulted obligors, since it is only after a formal default that loss given default can be calculated (note that in this respect the RDS for LGD is different from that for PD, which must also contain undefaulted experience). It should also include factors that can be used to group the defaulted facilities in meaningful ways.

Ideally, the RDS should:

  • •  cover at least one full business cycle;

  • •  contain all defaults that have occurred within the considered time frame;

  • •  contain data for calculating realized LGDs;

  • •  include all relevant information needed to estimate the risk parameters;

  • •  include data on the relevant drivers of loss.

10.22  In some cases realized LGD may be zero. This will happen if, for example, a breach is cured with no material direct or indirect cost associated with collecting (p. 194) on the instrument, and no loss caused by material discount effects (for example, if the default was caused solely by the ‘90 day past due’ criterion, and payment obligations were subsequently completely fulfilled). It is also possible for realized LGD to be positive—for example, where a default is cured without loss but the effect of the default is to increase the rate payable by the borrower on the facility over the default period. However, estimated LGDs may never be less than zero. Where the RDS contains a large number of zero-loss LGDs, regulators are likely to be concerned that the institution is, for example, using an inappropriately early definition of default or that the RDS contains some facilities which are not true defaults (for example, technical defaults such as small outstanding charges on repaid loans).

LGD—Risk drivers

10.23  Expected LGDs should be allocated to facilities in accordance with risk drivers, and the RDS should, where possible, capture for each default the part played by relevant risk factors in the default. Consequently the analysis of risk drivers is important. The Committee of European Banking Supervisors (CEBS)1 suggested that the risk drivers for assessing LGD exposures can be grouped according to the following five categories:

  1. (1)  Transaction related, including facility type, collateral, guarantees from third parties, seniority, time in default, seasoning, loan to value (LTV), and recovery procedures.

  2. (2)  Borrower related, including borrower size (as it relates to transaction characteristics), exposure size, firm specific capital structure (as it relates to the firm’s ability to satisfy the claims of its creditors in the event that it defaults), geographic region, industrial sector, and line of business.

  3. (3)  Institution related, including internal organization and internal governance, relevant events such as mergers, and specific entities within the group dedicated to recoveries such as ‘bad credit institutions’.

  4. (4)  External, including interest rates and legal framework (and, as a consequence, the length of the recovery process).

  5. (5)  Other risk factors.

Each institution is responsible for identifying and investigating additional risk drivers that are relevant to its specific circumstances. Institutions should collect data on what they consider to be the main drivers of loss for a given group of facilities, and should include the most material drivers in their LGD estimation process. The institution’s judgements as to which risk drivers are most material should be appropriately documented and should be discussed with supervisors.

(p. 195) LGD—Estimation methodologies

10.24  Supervisors do not prescribe any specific technique for LGD estimation (or for that matter any other IRB parameters). However, institutions will have to demonstrate that the methods they choose are appropriate to the institution’s activities and the portfolios to which they apply. The theoretical assumptions underlying the models should also be justified, and the approach must be based on quantitative information—the CRD does not permit the use of estimates based purely on judgemental considerations.

10.25  The four main approaches to LGD estimation are workout LGD, market LGD, implied market LGD, and implied historical LGD.

  1. (1)  Workout LGD. This is created by analysing how the cash flows resulting from the workout and/or collections process, properly discounted, are calculated. Calculations for the exposures that are currently held by the institution have to be based on actual recovery data in order to produce a forward-looking estimate. The calculation should not be based solely on the market value of collateral; appropriate adjustments should be applied. The calculation of default weighted average of realized LGDs requires the use of all observed defaults in the data sources. Observed defaults include incomplete workout cases, although they will not have values for the final realization of LGD because the recovery process has not ended. The workout LGD technique can be applied using either direct estimates or a two-step approach. Under a direct estimate approach, the specific characteristics of the proposed facility are compared with the factors present in existing defaults, and a quantitative LGD estimate is derived for the specific exposure. Under a two-step approach, an average LGD is estimated for all exposures covered by the same facility grade or pool, and the facility is allocated to that pool.

  2. (2)  Market LGD. As an alternative to workout LGD, the RDS may instead be derived from the observation of market prices on defaulted bonds or marketable loans soon after default or upon their emergence from bankruptcy. This approach may be used where data is scarce.

  3. (3)  Implied market LGD. It is also possible to derive estimated LGDs from the market prices of non-defaulted loans or bonds or credit default instruments. These are referred to as implied market LGDs. Implied market LGDs may be used where no other data are available to produce reliable estimates, and if validation results show that these estimates are reliable. The market and implied market LGD techniques are unlikely to be acceptable to regulators unless the market from which the price is derived is deep and liquid and there are good reasons for the scarcity of data. According to CEBS it is unlikely that any use of market LGD can be made for the bulk of the loan portfolio.

  4. (4)  Implied historical LGD. The implied historical LGD technique is allowed only for the retail exposure class. It is derived by creating a realized loss figure for a retail portfolio by dividing the total loss by the total number of exposures (p. 196) in the pool, while the ‘average realized LGD’ is the same total loss divided by the number of defaulted exposures in the pool. This is accepted where institutions can estimate the expected loss for every facility rating grade or pool of exposures.

LGD for expected loss

10.26  Institutions that calculate their own estimates of LGD should use downturn LGDs both in calculating risk-weighted exposure amounts and in calculating expected loss for exposures that are not in default (if this is more conservative than the long-run average). However, for defaulted exposures, the CRD requires the use of an estimate of expected loss (ELBE) that should be the best estimate of expected loss, given current economic circumstances. In such cases, LGD is defined as the sum of ELBE and a measure reflecting possible additional unexpected losses during the recovery period. If downturn conditions are relevant to a certain type of exposures, then this should be taken into account in measuring the possibility of additional unexpected losses during the workout period. This treatment does not apply to exposures under the double default treatment, since, by definition, EL for these exposures is set at zero.

Recognizing double default

10.27  This is a special regime available for certain types of credit protection in respect of certain assets in certain circumstances. This approach—‘the double default approach’—applies where the protected asset is a corporate, insurance, government, or retail portfolio (ie not an exposure to a bank, securities firm, equity, securitization, or non-credit asset) and the protection provider is an institution, non-guaranteed export credit agency, or insurance undertaking which has ‘sufficient’ expertise in providing unfunded credit protection and: (a) is a regulated bank or equivalent and had at the time the protection was provided an external credit rating equal to step 3; or (b) had at the time the protection was provided an external credit weighting equal to step 2; or (c) has at the time of assessment an internal credit weighting equal to step 3.

10.28  The rationale for the double default regime requires a word of explanation. Where an exposure has the benefit of a guarantee or credit protection, the normal treatment is to substitute the credit of the guarantor for the credit of the underlying exposure. Where the guarantor is a better credit than the underlying, this results in an improvement in the risk weighting of the exposure. However, where the guarantor is a worse credit than the underlying, this means that the guarantee will have no consequences at all for the regulatory capital treatment of the exposure.

This is clearly not an entirely accurate reflection of reality, since in practice the existence of a second credit claim reduces the risk somewhat. However, it is in accordance with the additive structure of the Basel Accord. It was felt, however, that (p. 197) in some cases (notably those where products were ‘wrapped’ by specialist credit providers) this policy should be relaxed slightly, since in such cases the combination of two very high quality credits could be shown to have a significant impact on overall risk.

10.29  In such cases, the risk-weighted exposure amount calculated using the formula previously set out is adjusted. This operates as a two-stage process. In the first stage the exposure value for the underlying obligation is calculated using the PD of the obligor but the LGD of the protection provider. This yields a risk weighting for the exposure (RW). This figure is then reduced by reference to the following formula:

Part II Commercial Banking, 10 The Internal Ratings-Based Approach

where PDpp is the PD of the protection provider.

10.30  It should be clear from this that the application of the formula will only produce a benefit to the bank where the PDpp is less than (roughly) 0.53 per cent, since at any level above this the multiplier will be greater than one. This means that only protection providers of very high quality will confer any benefit under this regime.

C.  Exposure at Default

10.31  EAD is a more sophisticated equivalent of the credit conversion factor regime which applies in the standardized approach (see para 8.95 ff.), and in effect modelled EAD is substituted for credit conversion factor (CCF). However A-IRB banks may not use their internally modelled EADs in respect of commitments which attract a CCF of 100 per cent. An A-IRB bank which does not have permission to model EAD for a particular portfolio must use CCFs for that portfolio.

10.32  Crudely put, EAD is an estimate of how much higher an institution’s exposure will be to a borrower on default than it is at the time of the calculation. There are two ways in which EAD typically comes into play. One is where the exposure is in the form of a revolving or undrawn commitment, where the amount currently drawn by the borrower is less than that which could be drawn if the facility were fully utilized. The second is where the exposure arises in the form of a guarantee or similar arrangement, where the current obligation is zero but where the bank may at some point in the future incur an obligation to pay. More generally, banking book exposures will change over time for a number of reasons—additional drawings under existing facilities; or under new facilities to existing or even new borrowers; repayments (whether scheduled, voluntary or lender-induced) under facilities; or accrual/payment of interest and charges.

(p. 198) 10.33  It is a view widely held by regulators that bank risk systems potentially understate capital requirements on facilities where balances are not fully drawn. Hence the direct focus has been on increased drawings under existing facilities. Some simple rules were incorporated into the Basle I framework to address these, but those rules were primitive and (particularly as regards the nil weighting for exposures of under one year) had spawned structures created specifically to arbitrage those rules.

10.34  It is a core principle of EAD estimation that the EAD figure for an exposure may never be lower than the amount currently drawn—in other words, the institution may not estimate that there will be a net reduction of exposures prior to default. This is true even if there is good reason to believe that such a reduction will in fact take place. Partially as a consequence of this, EAD is not explicitly affected by the maturity of facilities.

10.35  The core requirement for EAD has been stated as that the EAD required for IRB purposes is the exposure(s) expected to be outstanding under a borrower’s current facilities should it go into default in the next year, assuming that economic downturn conditions occur in the next year, and assuming that, other than any changes resulting from the economic downturn conditions, a firm’s policies and practices for controlling exposures remain unchanged from what they are at present. As with other aspects of the IRB framework, the EAD estimates to be used for capital purposes are based on the realized EADs in the RDS of exposures that have gone into default in the past. This basic historical data needs to be adjusted to take account of, inter alia, changes in policies and practices, and to produce an orientation towards an economic downturn. Estimated EAD cannot be less than current drawings.

10.36  The exposure value for leases must be the discounted minimum lease payments. Minimum lease payments are the payments over the lease term that the lessee is or can be required to make and any bargain option (ie option the exercise of which is reasonably certain). Any guaranteed residual value fulfilling the criteria for unfunded credit protection should also be included in the minimum lease payments.

10.37  Where an exposure takes the form of securities or commodities sold, posted or lent under repurchase transactions or securities or commodities lending or borrowing transactions, long settlement transactions and margin lending transactions, the exposure value must be the value of the securities or commodities. Where a volatility adjustment is required to be applied by the collateral method used, that volatility adjustment must be added to the exposure. Exposures under derivative instruments are to be calculated in accordance with the rules applicable to derivatives.

10.38  For IRB purposes exposures must be measured gross of value adjustments. Where on-balance sheet exposures are calculated after taking account of netting, or where (p. 199) netting is used under a repo or lending or borrowing arrangement, the institution must calculate exposure in accordance with the credit risk mitigation netting rules

Netting and EAD

10.39  In the own estimates approaches, netting is relevant to the calculation of current exposures, but is also relevant to the calculation of future exposures. The circumstance in which this is most clear is where a bank operates a cash pooling arrangement for a group of companies and manages its exposures to that group on a net basis. For such an arrangement some borrowers can be significantly in debt provided that others have countervailing positive balances. In general, such exposures should be treated as net balances, and EAD calculated on a net basis, provided that the firm meets the general conditions for on-balance sheet netting.

10.40  The issue also arises in the case of securities underwritings. Clearly where a firm agrees to buy securities it has an exposure to the security issuer. The question is as to what recognition should be given within the system to the fact that by the time it receives the securities it will have made arrangements to sell them on to other investors. It is clearly objectionable to regulators to take the view that a firm should reduce its EAD on the basis of an intention to sell off its exposure. However, given the strong and enduring market practice in this area, regulators clearly take some comfort from the fact that underwriters invariably do sell off such holdings, and as a result regulators are prepared to permit IRB banks to alleviate their positions by recognizing this intention.

Commitments—when should a CF/EAD be applied?

10.41  This question can be restated as: when does an arrangement become a commitment? Under the Basle I regime a facility attracted a non-zero CF only if it was committed and had a drawdown period of one year or more. However, even then there was considerable scope for debate as to when a facility became sufficiently ‘committed’ to warrant a capital requirement being applied to it. To take a simple example, a bank is approached by an investor who is contemplating entering a contested auction for a particular asset. The bank agrees in principle that it would be prepared to lend to the investor to finance the asset on specific terms subject to credit review. At this stage the bank may be aware that the investor is in discussion with several other banks. The investor enters the auction, and makes the winning bid subject to diligence. It commences diligence, and simultaneously commences negotiations with the bank as to the detailed terms of the facility. At some point these terms are agreed and documented in a facility letter, subject to documentation. When diligence is completed, the loan agreement will be documented, and on completion the loan will be drawn. The question for the regulator is at what stage the bank should begin to recognize that it has a commitment. Banks have sometimes argued that since they are only formally legally committed at the (p. 200) documentation stage, no charge should be applied prior to that stage. However, both regulators and bank management recognize that industry practice is that reliance is—justifiably—placed on banks at a very early stage in such discussions, and that in practice a bank could not withdraw from an informal commitment of this kind without doing serious and lasting damage to its commercial reputation and standing. Thus, whatever the point at which the commitment is recognized, it must be at some point earlier than final documentation.

10.42  The issue for regulators here is getting the balance right between, on the one hand, not wanting to apply a capital requirement too early in the facility negotiation approach and where the prospects of its being completed may be quite low, and on the other of allowing too much liberalism such that the existence of conditions precedent has the effect of setting the regulatory capital charge to zero until the time that a drawing is virtually certain.

10.43  The general approach taken in the market is derived from some extremely old (and now repealed) guidance from the UK Financial Services Authority to the effect that ‘a firm should treat a facility as an exposure from the earliest date at which a customer is able to make drawings under it’. However, this is the last possible date on which it may be possible not to allocate a CF to a facility—and is not intended to set out the commencement date for the weighting. The broad principles to be applied seem to be as follows:

  1. (1)  An EAD/CF is required on a facility from the time that a borrower is advised by the bank that it has agreed the facility is to be made available. Where there is a strong possibility that the facility will not eventually be taken up, that may be reflected through a reduction in the EAD/CF applied.

  2. (2)  Internal indications of willingness to provide facilities in the future which have not been advised to the customer do not generally require an EAD/CF. However, where the bank has taken a firm decision to make funds available to the borrower, an EAD/CF is required from the time that that fact is recorded on its systems, although for a facility where this would ordinarily be confirmed to the borrower, this may be suspended until the communication has been made.

  3. (3)  An EAD/CF should not be applied where a facility is subject to a full credit assessment by the bank, resulting in a re-rating or a confirmation of the rating of the borrower.

10.44  As with LGD, estimation of EAD is intended to be oriented towards what happens in a downturn, either by the expected default weighted average over a long run, or the direct use of estimates appropriate for a downturn if the conversion factors are expected to be higher in a downturn. However, this is difficult, since EAD will be heavily influenced by lender behaviour—through transactional lenders taking action to cut their lines, through borrowers increasing utilizations, (p. 201) and by the impact on the cash flow of the borrower of the actions of other lenders.

Maturity

10.45  For F-IRB banks the default maturity for corporate exposures is two and a half years (apart from repo-style transactions, where it is six months). A-IRB banks are permitted to use a weighted average mechanism. Where an exposure is subject to a formal repayment schedule the maturity will be a weighted average of the maturities over the repayment periods (with a maximum of five years). In any situation where the position of the parties does not permit the institution to calculate this amount, the value of M to be used must be the maximum remaining time in years that the obligor is permitted to take to fully discharge its contractual obligations. The modelled figure is subject to a one-year floor, except for fully (or nearly-fully) collateralized capital market driven transactions, and repo-style transactions, which provide for daily remargining.

10.46  Time to default is primarily a factor for PD—the longer the repayment period the greater the chance of a default during it. However, it also has some relevance to EAD—for example, the EAD would be expected to be different if a borrower defaults in two years’ time, than if it defaults in one year’s time. However, there is no immediately intuitive association between time and quantum—it does not necessarily follow that a borrower which fails in two years’ time will owe more or less than one which fails in one year’s time. A bank could argue that the maximum exposure at the end of a two-year period should be reduced by the value of the facilities which would expire in that time—thus, the maximum EAD over a period should be the value of the currently existing facilities which would still be drawable at that time (therefore, for example, if an institution has credit lines of £200m open to a particular borrower of which £100m would have expired by the end of a two-year period, it could validly argue that the maximum commitment over the two-year period should be the lower figure). However, this would be to disregard the fact that banks conventionally grant new lines to replace existing lines; and to assume that existing lines would not be renewed would be unrealistic.

10.47  In general, the time horizon to be used for EAD calculation should be one year unless a firm can demonstrate that another period would be more conservative. However, where one year is the term used, firms need not hold capital against facilities, or proportions of facilities, that cannot be drawn down within the next year; and, where facilities can be drawn down within the next year, firms may in principle reduce their estimates to the extent that they can demonstrate that they are able and willing, based on a combination of empirical evidence, current policies, and documentary protection, to prevent further drawings.

Financial sector exposures under IRB—the asset value correlation multiplier

10.48  The asset value correlation (AVC) multiplier is implemented in order to embed in the regulatory system the assumption that all financial institutions are to some degree correlated with each other. The AVC multiplier does not attempt to reflect the extent to which this may be the case in practice, but is simply a levy on all exposures to financial institutions calculated roughly according to their existing creditworthiness. The AVC multiplier is effected by applying a correlation factor to the IRB model for all financial institution exposures.

10.49  It will be remembered that the IRB formula for exposures not in default is as follows:

Part II Commercial Banking, 10 The Internal Ratings-Based Approach

Part II Commercial Banking, 10 The Internal Ratings-Based Approach

Part II Commercial Banking, 10 The Internal Ratings-Based Approach

Part II Commercial Banking, 10 The Internal Ratings-Based Approach

For financial institutions a multiplier of 1.25 is applied to the correlation parameter of all exposures to financial institutions. Thus the normal R is replaced by R_FI as follows:

Part II Commercial Banking, 10 The Internal Ratings-Based Approach

(p. 202) The effect of this multiplier is very roughly to increase capital requirements by 35 per cent for exposures to the largest institutions.

10.50  The key to this is of course the definition of ‘financial institution’ for this purpose. Basel goes about this as follows:

  • •  Regulated financial institutions whose total assets are greater than or equal to US $100bn. The most recent audited financial statement of the parent company and consolidated subsidiaries must be used in order to determine asset size. For this purpose a regulated financial institution is defined as a parent and its subsidiaries where any substantial legal entity in the consolidated group is supervised by a regulator that imposes prudential requirements consistent with international norms. These include, but are not limited to, prudentially (p. 203) regulated insurance companies, broker/dealers, banks, thrifts, and futures commission merchants.

  • •  Unregulated financial institutions, regardless of size. Unregulated financial institutions are, for the purposes of this paragraph, legal entities whose main business includes: the management of financial assets, lending, factoring, leasing, provision of credit enhancements, securitization, investments, financial custody, central counterparty services, proprietary trading, and other financial services activities identified by supervisors.

It should be noted that the effect of this regime is to prescribe a premium for exposures to both large banks and all non-bank financial entities regardless of size. Exposures to smaller banks, however, are relatively advantaged within this structure, being exempt from the premium and treated as analogous to corporate exposures.

Highly leveraged counterparties

10.51  Prior to the crisis, it had been almost an article of faith amongst regulators that the next major financial market problem would arise out of the hedge fund sector. Hedge funds were perceived by regulators to be unregulated and highly leveraged, and although fund managers were generally (for UK although not for US funds) regulated and hedge funds in general averaged leverage ratios a fraction of those of the banks, they were still regarded as highly dubious. Although the hedge fund sector emerged blameless from the crisis, regulators retained this suspicious attitude into the post-crisis environment, and Basel III therefore saw the fruition of plans, which had been discussed well before the crisis, for increasing the risk charge attributable to these funds. Although hedge fund leverage is generally lower than that of banks, it is equally generally higher than that of most corporates, and the hook on which the exposure is hung is the designation of ‘highly leveraged counterparty’.

10.52  It is fair to note, however, that the policy basis of this proposal is wider than simply an attack on hedge funds. Bank regulators are clear-eyed about the fact that increases in bank capital requirements will drive financial business into the non-bank sector, and confidently expect non-banks to grow faster than banks as suppliers of credit to the real economy. This will of course mean that some of the financial risk which currently resides within the banking system will relocate outside the banking system. This is unproblematic if that risk is genuinely removed from the banking system. What concerns regulators, however, is that if non-bank entities fund themselves through bank borrowing, the banks will be as exposed to those risks as they would have been had they remained on their balance sheets. The basis of this proposal, looking forward, is therefore that any exposure to a highly leveraged counterparty should attract a higher risk charge, with the aim of decreasing exposures between the banking and the shadow banking system.

(p. 204) 10.53  The Committee argues that since exposures to hedge funds are in general margined, the rules mentioned (which have the effect of significantly increasing the capital requirements for margined transactions) will have the effect of increasing the risk charge applied to exposures to hedge funds. However, in order to supplement this, the rules on PD calculation will be amended so that ‘PD estimates for borrowers that are highly leveraged or for borrowers whose assets are predominantly traded assets must reflect the performance of the underlying assets based on periods of stressed volatilities’.

Basel III final—LGD and EAD—wrong-way risk

10.54  Basel is concerned about the existence of wrong-way risk. ‘Wrong-way’ risk is generally regarded as a phenomenon which arises when correlated exposures are treated as uncorrelated—thus credit protection purchased from an entity on that entity’s own risk would generally fall within this description. However wrong-way risk is by no means always this easy to identify, and the regulators would like to see more focus on the topic by the industry. Three measures are introduced to address the issue.

10.55  The first is a general requirement that ‘Banks should monitor general wrong-way risk by product, by region, by industry, or by other categories that are germane to the business’. The second is a requirement that any position which has any significant element of wrong-way risk should be disregarded for netting purposes, and should otherwise be treated as of no value—this is achieved in practice by requiring an LGD of 100 per cent to be used.

10.56  These measures, however, address only specific cases, and operate with the grain of existing bank practice. The third measure is much more important, and addresses a much larger issue. This is that in a crisis everything falls out of bed at the same time, and the risk manager finds himself in a world where everything is correlated with everything else.

10.57  The first part of this is the requirement for banks to conduct stress testing and for scenario analyses to be designed in such a way as to identify risk factors that are positively correlated with counterparty creditworthiness. Such testing needs to address contagion arising by region, by industry, or by other categories that are germane to the business.

10.58  Transactions with counterparties where ‘specific’ wrong-way risk has been identified—notably where entities are within the same group—need to be treated differently when calculating the EAD for such exposures. A bank is exposed to ‘specific wrong-way risk’ if future exposure to a specific counterparty is highly correlated with the counterparty’s PD. For example, a company writing put options on its own stock creates wrong-way exposures for the buyer that are specific to the counterparty. Procedures to assess specific wrong-way risk need to include (p. 205) both procedures for identifying this risk at the outset of the transaction, but also procedures for identifying when such risk may arise during the course of the transaction because of external factors (such as changes in the sources of funding or credit dependencies of the entity concerned).

10.59  Instruments where there is specific wrong-way risk may not be included within a netting set with other transactions. Thus if a bank has purchased credit protection from a counterparty on a variety of names including the counterparty’s own, the credit default swap (CDS) referencing the counterparty itself may not be included in the same netting set as the other CDS. Furthermore, for any such single-name CDS where specific wrong-way risk has been identified, EAD must be calculated on the assumption that the underlying issuer is in liquidation, and LGD for advanced or foundation IRB banks must be set to 100 per cent for such swap transactions. For banks using the standardized approach, the risk weight to use is that of an unsecured transaction. For equity derivatives, bond options, and securities financing transactions EAD equals the value of the transaction on a default of the underlying security. In as much as this makes reuse of possibly existing (market risk) calculations (for incremental risk charge) that already contain an LGD assumption, the LGD must be set to 100 per cent.

D.  Specialized Lending

10.60  The specialized lending regime addresses a problem which is inherent in the credit risk weighting system, that being that it weights credit exposures. For the vast majority of banking transactions this works appropriately. However, the approach breaks down to some extent where the exposure undertaken is an exposure to an asset—as is the case with asset finance and some project finance transactions. In these cases a PD/LGD approach can be implemented, but the implementation is considerably more difficult than with other asset types. This is partially due to a smaller universe of experiences, but partly also to the fact that the risk indicators approach does not apply in situations where asset valuations may be driven by a wide variety of factors only some of which can be modelled. Of course, institutions with a sufficiently detailed database of default data may be able to calculate PD and LGD for asset-based exposures, and in this case they will be permitted to do so. However, for institutions which do not have this capability but wish to engage in project and other asset finance lending, the specialized lending regime provides a mechanism by which they can do so.

10.61  The specialized lending regime may only be applied in particular circumstances. These are that:

  1. (1)  the exposure concerned is to an entity which was created specifically to finance and/or operate physical assets;

  2. (p. 206) (2)  the contractual arrangements give the lender a substantial degree of control over the assets and the income that they generate; and

  3. (3)  the primary source of repayment of the obligation is the income generated by the assets being financed, rather than the independent capacity of a broader commercial enterprise.

10.62  In general, the following transaction types will fit within the specialized lending criteria:

  • •  Project Finance. A method of funding where the lender looks primarily to the revenues generated by a single project for his return, and where the lender’s exposure is calculated by reference to the receipts of a single-project company and the collateral value of the project’s assets.

  • •  Object Finance. A method of funding the acquisition of physical assets, where the primary income of the borrower will be related to lease or rental contracts and where the borrower is a special purpose vehicle (SPV) or otherwise has no independent capacity to repay the loan.

  • •  Commodities Finance. Structured short-term lending to finance inventories or receivables of exchange-traded commodities, where the borrower has no independent capacity to repay the loan.

  • •  Income-Producing Real Estate. Financing the construction of specific buildings where repayment of the finance is to be provided through cash flows generated by the asset in the form of rent or sale proceeds.

  • •  High-Volatility Commercial Real Estate. Investment in types of income-producing real estate which the national regulator considers to be high-volatility exposures.

Creating and validating default and loss models for these exposures is considerably harder than for other types of corporate exposure. As a result, banks are required to use the standardized approach by default unless they can satisfy their regulator that they meet the criteria for modelling risk. However, in theory both F-IRB and A-IRB are available in respect of assets falling into this category.

10.63  In each of these cases, guidelines are provided (known as the ‘slotting criteria’) in the Accord which will result in a classification of the relevant exposure into a five-fold hierarchy of ‘strong’, ‘good’, ‘satisfactory’, ‘weak’, and ‘defaulted’.2 The analytical process involved is entirely judgemental, and the allocation of characteristics to categories is, in practice, one which relies predominantly on the expertise of the classifier. Once this analysis has been performed, the exposure is weighted as shown in Table 10.6.

Table 10.6  Basel II Risk Weights for Specialized Lending

Remaining Maturity

Strong

Good

Satisfactory

Weak

Default

More than 2.5 years

70% (50%)*

90% (70%)*

115%

250%

0%

Less than 2.5 years

50%

70%

115%

250%

0%

A firm may use the bracketed figures if it has explicit permission to do so from its supervisor.

10.64  Expected loss figures for specialized finance exposures should be as shown in Table 10.7.(p. 207)

Table 10.7  Basel III Expected Loss for Specialized Lending

Remaining Maturity

Strong

Good

Satisfactory

Weak

Default

More than 2.5 years

0.4% (0%)*

0.8% (0.4%)*

2.8%

8%

50%

Less than 2.5 years

0%

0.4%

2.8%

8%

50%

A firm may use the bracketed figures if it has explicit permission to do so from its supervisor.

Basel III final—specialized lending

10.65  Basel III makes major changes to the specialized lending architecture. Risk weightings are reallocated as shown in Table 10.8.

Table 10.8  Basel III Risk Weights for Specialized Lending

Classification

Strong

Good

Satisfactory

Weak

Default

Rating equivalent

BBB- or better

BB+ to BB-

BB- to B+

B to C-

N/A

PF, OF, CF, IPRE

70% (50%)*

90% (70%)*

115%

250%

0%

HVCRE

95% (70%)*

120%

(95%)*

140%

250%

0%

A firm may use the bracketed figures if it has explicit permission to do so from its supervisor, the exposures concerned have a remaining maturity of less than 2.5 years, and they are of high quality.

10.66  The EL allocations are also amended as shown in Table 10.9.(p. 208)

Table 10.9  Basel III Expected Loss for Specialized Lending

Strong

Good

Satisfactory

Weak

Default

Non-HVCRE SL

5% (0%)*

10% (5%)*

35%

100%

625%

HVCRE

5%

5%

35%

100%

625%

A firm may use the bracketed figures if it has explicit permission to do so from its supervisor, the exposures concerned have a remaining maturity of less than 2.5 years, and they are of high quality.

E.  Retail and Mortgage Exposures

Retail exposures

10.67  Retail exposures are approached in a different way from corporate, bank, and sovereign receivables, in that rather than each individual borrower being given a credit rating, exposures are assessed on a portfolio basis. This permits the institution concerned to recognize correlation effects within the portfolio, and as a result a diversified retail portfolio is likely to be treated as a better credit than an equivalent exposure to a corporate borrower.

10.68  There is no F-IRB approach for retail exposures—banks are required either to completely model their retail portfolios or to use the standardized approach.

10.69  The rationale for the existence of the retail exposures regime is to recognize the superior risk characteristics possessed by a highly diversified pool of non-granular exposures. This enables a separate capital calculation to be applied to that portfolio to take those characteristics into account. This means that the approach is only appropriate where exposures are managed as homogeneous pools rather than as individual exposures.

10.70  The criteria used within the IRB approach for an exposure to qualify as a retail exposure are broadly similar to those used in the standardized approach. The class of assets eligible with the IRB approach may be slightly wider—for example, the present value of retail minimum lease payments is eligible to be treated as a retail exposure. The IRB approach also requires that the exposures be managed as an aggregate and not as individual exposures.

10.71  Firms may group their retail exposures as they wish, and in particular may segment them by collateral type, by obligor risk characteristics, by transaction risk type, or by delinquency. A firm must assess, test, and validate its model and its loss data at least as effectively as it updates its corporate model, and the retail model must meet IRB standards. As noted in the section on models, the IRB model applied to retail exposures gives a lower figure than that for equivalent corporate exposures.

Specialized retail exposures

10.72  There are two specialized classes of retail exposures, both of which have their exposures calculated slightly differently. For exposures secured by real estate, the correlation formula is not used, but correlation is taken to be 0.15 in all cases. For overdrafts and credit card receivables (known for this purpose as qualifying revolving retail exposures), the same is true but the factor to be used should be 0.04.

10.73  The practical effect of these adjustments is that the risk weighting of qualifying revolving retail exposures is for most purposes substantially less than (p. 209) the weighting which would be given to ordinary retail exposures. However, it is notable that the effect of using a fixed factor rather than the formula means that at higher PD levels the risk-weighted figure is actually higher than it would be if the formula were used. Thus, the examples set out in Annex 5 of the Accord give the illustrative risk weights for the three different classes set out in Table 10.10.

Table 10.10  Illustrative Risk Weightings for Different Types of Retail Exposures

Risk Weights Assuming 45% LGD and 2.5yr Maturity

PD

Retail Exposures

Residential Mortgages

Qualifying Revolving Retail Exposures

0.05%

7%

6%

2%

1.00%

46%

56%

17%

2.00%

58%

88%

29%

5.00%

66%

148%

55%

10.00%

76%

204%

84%

10.74  The figures for mortgage lending look remarkably high in Table 10.10. However, this is because all of these figures are calculated using a common LGD of 45 per cent. In reality the LGD for a retail mortgage is considerably smaller than that for any other form of retail exposure, and as a result, when the mortgage formula is applied to a realistic LGD, the actual risk weightings achieved are significantly lower than those previously set out for general retail.

10.75  For these purposes the technical definition of a qualifying revolving exposure means a portfolio of exposures to individuals where outstanding balances are permitted to fluctuate based on the customers own decisions up to a limit. These balances must be unsecured and uncommitted, must revolve in practice as well as in theory, and must be subject to a maximum limit of €100,000 per individual exposure. The portfolio as a whole must exhibit a sufficiently high margin income that its expected net (after expenses) future margin income over the next twelve months will cover expected loss on the portfolio in that period by at least two standard deviations. Technically there are a number of types of retail facility which could be made to fit within this definition, but it is rare for any arrangement other than credit card receivables and overdrafts to be accepted as being within it.

10.76  The definition of default is slightly different for the retail portfolio. For exposures to enterprises managed within the retail portfolio, an exposure is treated as defaulted when it reaches 90 days past due—as with any other exposure. However, for true retail exposures this can be extended to 180 days past due if the relevant regulator considers it appropriate.

(p. 210) 10.77  A firm must maintain a database of retail exposures and use it to generate estimates of PD, LGD, and CCF calculated on long-run default experience. It is permissible to use third-party databases for this purpose, but only where there is a strong link between the firm’s assignment process and risk profile and that of the external data used. At least one of the data sources used (external, internal, or pooled) must include data over a period of at least five years, and if it covers a longer period then that period must be used, unless the firm can convince its regulator that the more recent data is a better predictor of loss.

10.78  For retail losses, further drawings may be reflected either in LGD or in CCF.

Basel III applies a floor to the LGD parameter for certain retail exposures. This is 5 per cent for mortgages, 50 per cent for QRRE, and 30 per cent for other retail, although this last may be reduced if appropriately collateralized.

F.  Eligible Purchased Receivables

10.79  There are two possible classes of eligible purchased receivables: retail receivables and corporate receivables. Purchased receivables, as with retail exposures, are assessed on a portfolio rather than an individual basis. There is in practice no difference between purchased and home-grown retail receivables. However, corporate receivables may be treated in accordance with this approach if they satisfy certain conditions.

10.80  In principle, if an institution acquires a portfolio of corporate receivables it is required to treat them as individual corporate exposures. However, if the portfolio satisfies certain conditions, it may be treated as a single asset using the top-down methodology employed for retail exposures. As with retail exposures, use of this approach significantly reduces the total risk exposure of the portfolio. The conditions are:

  • •  the receivables must be purchased from an unrelated third party, and may not be originated by the holder;

  • •  the receivables must have been created at arm’s length (thus, for example, intercompany debts are ineligible);

  • •  the maturity of the exposures must not be greater than one year (unless the exposures are fully secured by collateral);

  • •  the portfolio must be sufficiently diversified (this limit may be set by national regulators);

  • •  the claim must be on the whole pool. In general, any tranching of claims will have the effect of disallowing the use of this treatment and requiring the investor to use the securitization treatment (see later). However, the existence of a right of recourse to the seller will not have this effect.

(p. 211) 10.81  The approach to risk weighting corporate receivables involves the purchasing bank estimating a one-year expected loss amount. This expected loss figure is then decomposed into its components of PD and LGD. The exposure is then treated as an exposure to a single corporate borrower with those PD and LGD attributes.

Basel III

10.82  The effect of the Basel III restriction on the use of A-IRB approaches for certain corporate receivables (see para 10.04 above) means that the treatment of purchased corporate receivables is varied accordingly. Thus, if a bank purchases a pool of corporate receivables, it must use the F-IRB approach in respect of those obligors for whom A-IRB would not be permitted, although it may use A-IRB for the remainder. Since F-IRB is not permitted for retail exposures, all purchased retail receivables must be assessed using A-IRB.

G.  Equity Exposures

10.83  The regulatory capital system takes a broad definition of the notion of equity. Lawyers who are accustomed to consider ‘equity’ as that which is treated as share capital under the applicable company law are therefore required to refocus their attention on subordination as the primary determinant of equity status. For these purposes, equity means:

  1. (1)  non-debt exposures conveying a subordinated, residual claim on the assets or income of the issuer, and

  2. (2)  debt exposures the economic substance of which is similar to the exposures specified in (1).

This includes debt-like equity instruments (such as preference shares) and equity-like debt instruments (such as the most subordinated notes issued out of a tranched securitization or repackaging). It also clearly includes instruments whose effect is to pass on the economic equivalent of equities whilst remaining legally debt instruments. However, in some circumstances it may be remarkably difficult to identify precisely which instruments fall within this definition and which do not.

10.84  Equity exposures include both voting and non-voting interests. An instrument which is irredeemable, does not embody an obligation on the part of the issuer, and grants the holder a residual claim on the assets or income of the issuer is automatically characterized as equity. However, instruments which have the characteristics commonly encountered in innovative bank Tier 1 instruments (such as stock settlement or infinite deferral of payment obligations) may be characterized as equity at the election of the regulator. The substance over form approach is adopted both ways—preference shares and other instruments which, although legally equity, have the economic characteristics of debt are not caught by this definition.

(p. 212) 10.85  If a bank holds sufficient equity in a particular entity, that entity will be treated as part of the bank’s group. Thus the broad position in respect of any holding in a financial entity is as shown in Table 10.11.

Table 10.11  Treatment of Equity Holdings in Other Legal Entities

Financial Undertaking*

Non-financial Undertaking

Holding over 50%

Consolidated as a subsidiary

Consolidated as a subsidiary

Holding between 20% and 50%

Partially consolidated as a participation

Partially consolidated as a participation

Holding below 20% but above 10%

Material holding—deducted

Qualifying holding—deducted if the aggregate of such holdings exceeds specified limits

Holding below 10%

Weighted asset (subject to aggregate limit of 10% of bank’s own capital)

Weighted asset

‘Financial undertaking’ for this purpose means a bank, financial firm, or insurer.

10.86  The issue of equity held on bank balance sheets was historically of relatively little importance in the UK system, although in continental Europe the ‘relationship’ model of banking where banks held equity stakes in their customers was for a time more common. However, the major driver behind the increase in the amount of bank equity investment is the rise of private equity investment.

10.87  Within the standardized approach, equity investments are weighted at 100 per cent. The Accord suggests that regulators should have discretion to weight ‘higher risk portfolios’ at 150 per cent, but within the EU 100 per cent seems to have been adopted as the default weighting. The reason that this is important here is that many IRB banks have not progressed to any of the IRB approaches to weighting equity holdings, and remain on the standardized approach. The reason for this is not hard to discern—this is one of the relatively few areas in which all of the available IRB approaches are likely to produce a significantly higher capital requirement than the standardized approach.

10.88  Within the IRB approach there are three possible treatments for equity exposures. These are:

  1. (1)  the simple risk weight approach;

  2. (2)  the internal models approach; and

  3. (3)  the PD/LGD approach.

The first two of these are described in the Accord as the ‘market-based’ approach. This is also one of the areas in which EU (and therefore UK) regulation diverges significantly from the Accord.

10.89  In all of these cases the valuation of the exposure requires some thought. Broadly, the value is the value used in the accounts, whether the fair value or the lower of (p. 213) cost or net realizable value. However, hedging instruments may be recognized in calculating the value of the holdings as long as they have a maturity of at least one year and, unusually for the banking book, long and short positions may be set-off against each other to calculate the net exposure. Unfunded protection may also be recognized as reducing this exposure.

The simple risk weight approach for equity

10.90  The simple risk weight approach as set out in the Accord provides that institutions should apply a 300 per cent weight to equity holdings that are publicly traded and a 400 per cent risk weight to all other equity holdings.

10.91  Article 155 of the EU Capital Requirements Regulation (575/2013), by contrast, provided that equities should be weighted at either of 190 per cent, for private equity exposures in well-diversified portfolios, 290 per cent for exchange-traded equity exposures, and 370 per cent for all other equity exposures.

10.92  The expected loss figures to be used in calculating EL for equity are 0.8 per cent for private equity exposures in well-diversified portfolios and for exchange-traded equity exposures, and 2.4 per cent for all other equity exposures.

The PD/LGD approach for equity

10.93  The idea of using a PD/LGD approach for equity seems to lawyers to be absurd—equity, by definition, does not ‘default’, since it does not embed a payment obligation. Consequently it should not be possible to calculate either PD or LGD. The reason that the concept does make sense may be ascertained by referring back to the definition of ‘equity’ used for this purpose as set out in para 8.50. For regulatory purposes equity means the most junior capital within a structure, and does not refer to legal form. Thus the most junior piece of any financing will be likely to be considered to be equity for this purpose.

10.94  This approach employs the same approach as the foundation IRB approach for corporate exposures. The PD to be used must be estimated in the same way as that applied for debt exposures. Slightly higher minimum PD levels are imposed for equities under the EU Capital Requirements Regulation (575/2013):

  • •  0.09 per cent for exchange-traded equity exposures where the investment is part of a long-term customer relationship;

  • •  0.09 for non-exchange-traded equity exposures where the returns on the investment are based on regular cash flows not derived from capital gains;

  • •  0.4 per cent for exchange-traded equity exposures (including short positions); and

  • •  1.25 per cent for all other equities.

(p. 214) These figures compare with the 0.03 per cent minimum requirement for corporate debt exposures. The calculation of the RWA of the exposure is the same as for corporate and other exposures under the IRB formula.

10.95  The LGD to be used in the PD/LGD model is prescribed even for advanced IRB banks. It is set at 65 per cent for private equity holdings in sufficiently diversified portfolios, and 90 per cent in all other cases. A minimum risk weighting of 100 per cent is required for all public equities held as part of a long-term customer relationship and all private equity holdings not held for the purposes of achieving capital gain.

10.96  There is also an interesting quirk of the system as regards unfunded credit protection. Ordinarily, the purchaser of credit protection under the IRB approach substitutes an exposure to the protection seller—calculated as a function of the protection seller’s PD and LGD—for the exposure to the underlying asset. Broadly, the same approach applies where an IRB bank buys unfunded protection on an equity exposure. However, where protection is bought against an equity exposure, the LGD to be used on the protected portion is that which was initially applicable to the underlying equities—65 per cent for private equity in sufficiently diversified portfolios, 90 per cent for others. Thus the RWA calculation for an exposure subject to this form of protection is calculated using the PD of the guarantor and the LGD of the guaranteed exposure

10.97  The maturity assigned to all equity exposures should be five years.

The internal models approach for equity

10.98  The internal models method requires institutions to use their internal risk models to calculate the potential loss on the institutions equity holdings and to multiply that figure by 12.5 (the reciprocal of 8 per cent) to derive a risk-weighted assets figure for the holding. The main requirement for the internal models approach is that the total risk capital charge across the model as a whole must not be less than that which would have applied if the PD/LGD approach had been used. The requirements which apply to the use of this model are broadly similar to those which apply for the use of any other model—the model must be robust, must have been effectively back-tested against data, must be integrated into the institution’s management information systems, and must be verified at appropriate intervals.

Basel III and equity

10.99  Basel III prohibits the use of the IRB approach for equity—consequently the standardized approach (see paras 8.45 ff) is required to be used for all equity holdings. However, equity investments in funds are required to be assessed according (p. 215) to the separate rules governing banks’ investments in funds.3 These generally require a look-through approach to be applied where possible: a mandate-based approach (ie a hypothetical look-through approach based on the assumption that the fund has invested to the greatest extent in the highest risk investment permitted by its mandate) where full transparency is not available, and a 1250 per cent risk weighting for all other fund equity exposures.(p. 216)

Footnotes:

1  Committee of European Bank Supervisors, Consultation Paper 10, Final version 11 July 2005.

2  The full criteria are set out in Annex 6 of the Basel II text.

3  ‘Capital Requirements for Banks’ Equity Investments in Funds’, BCBS, December 2013.