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

Part III Investment Banking, 15 Counterparty Risk in the Trading Book

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: 08 July 2020

Subject(s):
Credit risk — Derivatives — Financial regulation

(p. 299) 15  Counterparty Risk in the Trading Book

A.  Introduction

15.01  The effect of the rules set out in this chapter is that certain exposures whose value can fluctuate over time should be treated as having a greater degree of risk than their actual mark-to-market value. In order to explain why this is, consider a bank which owns 100 of shares in A, but also has a derivative in place with X under which it is entitled to be paid the value of 100 shares in A. Clearly both positions give rise to the same risk as to the future price of A, and both will be valued by reference to the value of the shares in A. There is, however, a difference between the two. For the physical position, fluctuations in the value of the shares will result simply in gains or losses to the holder. Fluctuations in the value of the derivative, however, bring in an extra factor. This is that if the value of the shares (p. 300) in A increases, the bank’s credit exposure to X will increase. The rules set out in this chapter seek to capture this extra level of risk by treating the value of the derivative as being slightly higher than its mark-to-market value, thereby requiring a slightly higher level of capital to be held against it. This is the counterparty credit risk requirement (CCR).

15.02  What is being calculated here is either the valuation amount (for standardized banks) or the exposure at default (EAD) (for internal ratings-based (IRB) banks)—ie the value of the credit exposure. The rules apply to three different types of transaction: derivatives; securities financing; and long settlement transaction.

15.03  ‘Derivatives’ is not a term with a defined meaning, and this is entirely deliberate. However, at a first approximation, it includes at least:

  1. (1)  an interest rate contract, being:

    1. (a)  a single-currency interest rate swap;

    2. (b)  a basis-swap;

    3. (c)  a forward rate agreement;

    4. (d)  an interest rate future;

    5. (e)  a purchased interest rate option; and

    6. (f)  other contracts of similar nature.

  2. (2)  a foreign currency contract or contract concerning gold, being:

    1. (a)  a cross-currency interest rate swap;

    2. (b)  a forward foreign currency contract;

    3. (c)  a currency future;

    4. (d)  a currency option purchased;

    5. (e)  other contracts of a similar nature; and

    6. (f)  a contract concerning gold of a nature similar to (2)(a)–(e).

  3. (3)  a contract of a nature similar to those in 1(a)–(e) and 2(a)–(d) concerning other reference items or indices, including any financial asset.

15.04  The term ‘securities financing transaction’ is a portmanteau term which includes a number of different transaction types. The most significant of these are repo (short for repurchase) transactions, in which one party sells securities to another on the basis that it will buy them back again at a future date at a specified price, and stock lending transactions, in which one party transfers title to securities to another on the basis that it will be able to call for the return of those securities on the specified date. It may be helpful to note at this point that although repo and stock lending transactions look almost identical to lawyers, in economic terms they are very different. A repo is economically equivalent to a borrowing of a specific amount of money secured on the stock transferred, with an interest rate charged on that money in terms of the differential between the purchase price and the repurchase price. This effectively gives the participants a long or short interest rate position on the repo price. A stock loan is economically equivalent to a borrowing of stock against collateral, and the lender is in principle rewarded with a (p. 301) lending fee rather than an interest rate. Amongst other advantages, this makes stock loans easier to administer for lending institutions. The best way to envisage the difference between the two is that in a repo, the amount of the repayment is fixed and the amount of securities collateral therefore varies. In a stock loan, the value of the repayment varies, so the amount of securities used does not.

15.05  Securities financing transactions also include margin lending transactions, where an institution lends money to a customer in order for that customer to purchase, sell, hold, or otherwise trade in securities. The aim of this provision was to catch ‘prime brokerage’ and similar arrangements, where banks provided leverage to securities investors by extending credit based exclusively on securities portfolios. A distinction is drawn here between securities financing transactions (ie loan transactions entered into for the express purpose of financing securities business) and lending transactions which just happen to be collateralized by securities. The distinction is, however, not exactly a bright line.

15.06  The rules also apply to long settlement transactions. In some respects this is an anti-avoidance measure. If I agree to sell securities on terms that the buyer need only pay for them in twelve months’ time, what I have created is in fact a securities financing contract, but in legal and regulatory terms it is a contract for sale. The rule which is applied is therefore that any contract for the sale of a security, a commodity, or a foreign currency amount against cash, other financial instruments, or commodities, or vice versa, may be treated as a long settlement transaction. This will be the case where the settlement or delivery date is contractually specified as more than the lower of the market standard for the particular transaction and five business days after the date on which the firm enters into the transaction. Thus any sale of securities which has a settlement period of more than five days will be a long settlement transaction.

B.  Counterparty Credit Risk Calculation Methods

15.07  The basis of the valuation mechanism prescribed for derivative and other variable transactions is first to establish the amount due under the transaction, and then to apply a regulatory ‘haircut’ to that amount to arrive at the appropriate exposure value or EAD. This calculation is performed using one of three approaches:

  • •  the current exposure method (CEM), also known as the mark-to-market (MTM) method;

  • •  the standardized method (SM); or

  • •  the internal model method (IMM).

The CEM was introduced by Basle I and has the longest pedigree—a very large number of firms still use it to evaluate their CCR exposures. The SM and IMM were both introduced by Basel II.

(p. 302) One or other of these methods must be applied to derivatives and securities financing transactions within a single entity—they cannot be mixed for different derivative types. There is one exception to this rule which arises from the fact that the rules require a firm to use the mark-to-market approach for any non-linear exposure for which it cannot calculate a model value.1 Methods may be mixed across different group members in accordance with the consolidation rules. A firm must calculate the exposure value of a long settlement transaction in accordance with either one of these methods or, if it is permitted to use it, the master netting agreement internal models approach. If a firm enters into a transaction which is structurally a long settlement transaction in order to execute a trade which is structurally a derivative or securities financing trade, it may opt to simply treat that trade as a derivative or securities financing trade and apply to it whatever calculation method it applies to such transactions. The use of the long settlement approach is therefore, in effect, optional for derivatives trading firms with established processes.

15.08  Derivative exposures are generally effected under master documents produced by the International Swaps and Derivatives Association (ISDA) or other master documents which provide for broad netting. Exposures with any given counterparty are assessed using netting sets (see later). Where there is more than one netting set, there is more than one exposure (see later).

15.09  Exposures to certain central counterparties attract a CCR of zero under Basel II, and a CCR of 2 per cent under Basel III. This is a concession designed to encourage central counterparty use, and applies only where the central counterparty’s credit risk exposure with all participants in its arrangements are fully collateralized on a daily basis. However, this concession is not available for exposures arising from collateral held by the central counterparty for the participant to mitigate losses in the event of the default of other participants in the central counterparty’s arrangements—broadly, default fund contributions.

15.10  The inclusion of long settlement transactions within this framework focuses attention on the fact that although a firm is most unlikely to engage in derivatives or financing transactions by mistake or through inadvertence, it may well, for a variety of good reasons, find itself party to long settlement transactions. There is therefore a concessionary rule which has the effect that a firm may, if it wishes, weight long settlement transactions using the mark-to-market method and the standardized approach to credit weighting regardless of any other factor.

15.11  Finally, the collateral rules apply where credit protection is purchased against an asset held in the banking book. Where such protection is held, the credit exposure (p. 303) on the relevant credit derivative is, for the purpose of this section, set to zero. This is because the counterparty risk concerned is already accounted for in the collateral rules. Equally, where a firm has sold protection out of the banking book and as a result is treated as having the asset protected on its books, the CCR for the sold derivative is also set to zero.

C.  The Current Exposure or Mark-to-Market Method

15.12  The basis of the mark-to-market method is that the firm concerned must undertake a two-stage calculation. First, it must determine the current replacement cost of all contracts with positive values at market prices. This gives a mark-to-market value for each contract. Second, the firm must determine the ‘add-on’ to be applied to each contract, known as the potential future exposure (PFE). The exposure value for this purpose is then calculated as the sum of the mark-to-market value and the PFE.

15.13  Since the market value of a particular contract will be the value of the net obligation arising under the contract, the netting of the different legs of the derivative contract is already embedded in the valuation. Consequently, the adjusted value of the mark-to-market value of these contracts will constitute the capital requirement to be applied to these positions.

PFE calculation

15.14  The PFE is calculated by multiplying the nominal principal amount of the underlying contract by the percentage set out in Table 15.1.

Table 15.1  PFEs for Different Types of Contracts

Residual Maturity

Interest Rate Contracts

Contracts Concerning Foreign Currency Rates and Gold

Contracts Concerning Equities

Contracts Concerning Precious Metals except Gold

Contracts Concerning Commodities other than Precious Metals

One year or less

0%

1%

6%

7%

10%

Over one year but under five years

0.5%

5%

8%

7%

12%

Over five years

1.5%

7.5%

10%

8%

15%

Any contract not falling within one of the columns in Table 15.1 is treated as falling within the highest column. Where a contract requires more than one payment of principle, a separate multiplier should be applied to each payment to be made as if it were the only payment, and the PFEs for each payment are then (p. 304) totalled. For resetting contracts, maturity should be treated as the period until the next reset date.2

15.15  Once again, we note that the percentages shown in Table 1.5 are onerous for commodity dealings. For firms which apply the commodity extended maturity ladder (a regime created for commodity specialists—see para 13.41) there is a concessionary regime which applies the percentages shown in Table 15.2 to commodity derivative exposures.

Table 15.2  Commodity Derivative PFEs

Residual Maturity

Precious Metals (except gold)

Base Metals

Agricultural Products (softs)

Other, Including Energy Products

One year or less

2%

2.5%

3%

10%

Over one year but under five years

5%

4%

5%

6%

Over five years

7.5%

8%

9%

10%

15.16  PFEs should in theory be applied to all contracts, whether or not they have a positive mark-to-market value. Thus a fixed-floating interest rate derivative which currently has a negative value for the reporting bank will nonetheless attract a PFE. However, this only applies where there is some possibility of a positive amount being paid to the institution at some point. For a contract such as a written option, where the bank receives a payment up front and the only remaining issue is as to how much (if anything) it will have to pay out to the counterparty, no PFE should apply.

15.17  When a PFE is applied to a position where the mark-to-market exposure of the contract is negative, the exposure value of the contract is not deducted. However, where the exposure relates to collateral provided to cover a contract with a negative mark-to-market value—for example, where the reporting bank has provided a counterparty with £100 of collateral to cover the bank’s obligations under a contract which has a negative mark-to-market for it of £20—it is permitted to deduct the £20 from the £100 before calculating the PFE.

15.18  The determination of the principal amount of the contract is not always straightforward—some derivatives may provide that the principal will change on the occurrence of particular events. In the case of such contracts firms must reflect the effect of such provisions in their calculation.

(p. 305) Netting within the mark-to-market method

15.19  Where a number of obligations arise under a contract which provides for netting by novation (ie where the terms of the contract have the effect that no matter how many individual transactions are entered into under the contract, only a single net amount will ever be payable), both the mark-to-market value and the PFE should be calculated on the net balance payable under the novation. However, it is relatively unusual to encounter full netting by novation arrangements in derivatives, and the vast majority of derivatives transactions are done under close-out netting provisions. Under a close-out netting arrangement, multiple transactions done under the relevant agreement continue to be separately settled unless an event of default occurs, whereupon the netting provisions take effect and the obligations under the contract are reduced to a single net amount. Where a contract contains close-out netting provisions, the mark-to-market amount is calculated on a net basis in the same way as for novation netting. However, the PFE is increased by a factor which represents the increased level of risk which regulators consider to attach to close-out netting over novation netting. The relevant calculation is:

Part III Investment Banking, 15 Counterparty Risk in the Trading Book

where:

  1. (1)  PCEred = the reduced figure for potential future credit exposure for all contracts with a given counterparty included in a legally valid bilateral netting agreement;

  2. (2)  PCEgross = the sum of the figures for potential future credit exposure for all contracts with a given counterparty which are included in a legally valid bilateral netting agreement and are calculated by multiplying their notional principal amounts by the PFE percentages;3 and

  3. (3)  NGR = ‘net-to-gross ratio’: the quotient of the net replacement cost for all contracts included in a legally valid bilateral netting agreement with a given counterparty (numerator) and the gross replacement cost for all contracts included in a legally valid bilateral netting agreement with that counterparty (denominator).

It should be noted that the effect of this formula is that positions that net to zero will usually have a significant positive value, and that this value may be significant for large netted positions.

(p. 306) D.  The Standardized Method

15.20  The basis of the standardized method is that derivatives are disaggregated into one or more ‘payment legs’. A transaction which involves mutual payments—for example a fixed-floating interest rate swap—has two payment legs. Unless these obligations are denominated in the same currency and are payable on a net basis, they are treated as gross obligations even if the contractual documentation permits netting of payments. This includes the notional principal of the transaction. Where a payment leg gives rise to interest or FX risk, the position must be included in the calculation of the appropriate interest rate risk or foreign exchange risk calculation (however, payments with a remaining maturity of less than one year are disregarded for interest rate risk purposes).

15.21  In general, the risk position for any transaction relating to financial instruments with a linear risk profile is the effective notional value of the position—ie the value of the underlying financial instruments established as current market price multiplied by quantity. Where the obligation is to make payments of a specified calculated amount (such as a fixed-floating swap), the exposure is the notional value of the outstanding gross payments multiplied by modified duration. Modified duration is a measure of the weighted average term to maturity of a security, and for this purpose is calculated as the delta of the value of the position divided by the delta of the interest level. In this calculation, risk positions are assigned ‘signs’ according to whether the position is positive or negative for the reporting institution.

15.22  Collateral must be reflected in the calculation of the position, with collateral posted treated as an obligation immediately due to the counterparty, and collateral received treated as a claim immediately due from the counterparty. The effect of this is to net the collateral against the claim due. Where the collateral is the ‘wrong way’—for example where an institution has provided collateral to a counterparty, but at the time of the transaction has money due to that counterparty rather than receivable from it, the collateral simply increases the credit risk exposure of the institution to the counterparty.

15.23  The standardized method may only be used for financial derivative instruments and long settlement transactions—it does not apply to securities financing arrangements. Also, it may only be used for derivatives with a ‘linear’ risk profile. A linear risk profile is a term which describes a transaction where the amount due under the transaction varies directly with another factor. An example might be an obligation to deliver a specified number of specified securities, where the value of the derivative is the number of securities multiplied by the market price of those securities, as the amount due will vary directly (p. 307) with the market price. A non-linear risk profile would arise under a transaction where the obligation was to pay a variable amount on the occurrence of a specified uncertain event, since for such a transaction the return does not vary directly with any single underlying factor. In general, non-linear risks must be dealt with by applying the mark-to-market method. This must be applied even where the firm has a CAD 1 model or a value at risk (VaR) model if that model is not capable of estimating the delta or modified duration of the position. Each such exposure must be treated as a separate exposure—ie no two non-linear positions treated under the mark-to-market method may be netted against each other.

E.  Credit Risk Exposure Calculation

15.24  Once the reporting firm has established its individual risk positions, it must calculate its net exposure. In principle, the exposure calculation is done instrument-by-instrument—thus, long and short notional positions in any identical instrument may be set-off against each other. Positions in each instrument constitute a ‘hedging set’ of positions—thus long and short positions in any particular instrument are treated as hedging each other.

15.25  Underlying financial instruments other than debt instruments must be assigned by a firm to the same respective hedging sets only if they are identical or ‘similar’ instruments.

  • •  For equities, similar instruments are those of the same issuer. An equity index is treated as a separate issuer.

  • •  For precious metals, similar instruments are those of the same metal. A precious metal index is treated as a separate precious metal.

  • •  For electric power, similar instruments are those delivery rights and obligations that refer to the same peak or off-peak load time interval within any 24-hour interval.

  • •  Actual and synthetic positions in debt instruments of a certain issuer, or from reference debt instruments of the same issuer that are emulated by payment legs, or that underlie a credit default swap may be included in the same hedging set.

  • •  For certain low-risk instruments, positions in different instruments may be grouped together into a ‘hedging set’. Instruments may be included in a hedging set if: (a) they are in the same currency; (b) they satisfy the criteria for an interest rate specific risk PRR of less than 1.6 per cent; and (c) they fall within the same grouping in Table 15.3.

Table 15.3  Hedging Sets for Low-Risk Instruments

Government Referenced Interest Rates

Non-government Referenced Interest Rates

Maturity

<= 1 year

<= 1 year

Maturity

>1– <= 5 years

>1– <= 5 years

Maturity

5 years

>5 years

  • •  There is one hedging set for each issuer of a reference debt instrument that underlies a credit default swap.

(p. 308) 15.26  It is important to remember that the exposure which is being calculated here is the exposure to one particular counterparty. It is therefore a precondition for the inclusion of any exposure in a hedging set that it be covered by a netting agreement which permits the exposure to be set-off against other exposures to the same counterparty. Thus, if a reporting institution has an exposure to a particular counterparty which arises under a stand-alone agreement which is not covered by the general netting arrangements between it and the counterparty, it may not include that position within any hedging set, and must treat it as giving rise to a separate stand-alone exposure.

15.27  Once the net positions in the various different notional underlying instruments and exposure classes have been calculated, those positions must be turned into a credit requirement through the use of a CCR multiplier. The weightings to be used are those set out in Table 15.4.(p. 309)

Table 15.4  CCR Multipliers

Hedging Set Categories

CCR Multiplier

Interest rates

0.2%

Interest rates for risk positions from a reference debt instrument that underlies a credit default swap and to which a capital charge of 1.60% or less applies under the interest rate risk approach

0.3%

Interest rates for risk positions from a debt instrument or reference debt instrument to which a capital charge of more than 1.60% applies under the IRR approach

0.6%

Exchange rates

2.5%

Electric power

4.0%

Gold

5.0%

Equity

7.0%

Precious metals (except gold)

8.5%

Other commodities

10.0%

Underlying instruments of financial derivative instruments that are not in any of the above categories

10.0%

15.28  Once all this has been done, the firm must finally calculate its actual exposure value net of collateral. This is done by calculating:

  1. (1)  the current market value of the total portfolio of transactions included in the netting set, less the current market value of all of the collateral held in respect of those transactions;

  2. (2)  the sum of the risk positions (after collateral) of each hedging set within the netting set, weighted using the appropriate CCR multiplier.

The higher of these two figures is the actual exposure value. The effect of this is that each netting set exposure has a floor equal to market value less collateral (ie the exposure can never be less than the net mark-to-market value of the exposure to the counterparty), but that where the risk-adjusted calculation gives a higher figure, it is that higher figure which is used. This would generally be expected to be the case.

15.29  Finally, the total exposure calculated in this way is multiplied by a scaling factor of 1.4 (β‎ or beta). This scaling factor was introduced at the Basel level in order to ensure a degree of robustness within the system, and its calculation is well beyond the scope of this work. However, its effect is that if the product of the calculation described is £10m, then the credit exposure which will arise from the CCR standardized method will be taken to be £14m, and this will be the EAD of the position.

F.  The CCR Internal Model Method

15.30  A CCR internal model is, as it sounds, a model which is used by an institution to calculate counterparty credit risk. The output of a CCR model is the CRR for the exposures modelled. There are few generally applicable rules which can be specified in respect of such models—they are developed by banks and approved and reviewed independently by regulators. Such models may well not catch all of the types of derivatives which fall within the CRR regime—where this is the case the firm may use one of the other CRR methods for such exposure. However, a firm which has introduced a CRR model for part of its CRR exposures is expected to roll out that model to their other derivative exposures (except immaterial exposures) within a reasonable period of time, and in particular is expected to apply that model to new types of derivative transaction which it enters into. However, in general, regulators will try very hard to avoid a firm ‘cherry-picking’ its exposures—using a CRR model where it gains an advantage from doing so but using the standardized or mark-to-market methods where this would produce a lower risk charge.

15.31  CCR models, like any other CCR calculation, work at the level of the netting set. They must compute the exposure value for the netting set at each future date, and should catch movement in collateral values. The output of a model should be an estimate of effective positive exposure (EPE). EPE is defined as the weighted (p. 310) average effective exposure level (EE) at a series of times during the first year of exposure based on a variety of estimates of different market risk factors. The principle for the calculation of effective EE is simply that it can go up but not down—thus, if the output of the firm’s model suggests that EE over the next 12 months is as shown in Table 15.5, then effective EE will be as shown in Table 15.6.

Table 15.5  Illustrative Actual EE

t0

t1

t2

t3

t4

t5

£10m

£20m

£20m

£30m

£20m

£10m

Table 15.6  Illustrative Effective EE

£10m

£20m

£20m

£30m

£30m

£30m4

15.32  It is a principle that the output of the CCR model—no matter how high it may be—must be multiplied by α‎ (alpha). α‎, like β‎, is 1.4, but whereas β‎ is set by regulatory fiat at 1.4, α‎ is in principle either 1.4 or such higher number as the regulator may choose to require. However, if a firm can demonstrate that the output of its model is always at least 1.2 times EPE, then the regulator may give the firm permission to disapply the modifier.

15.33  A CRR model may take into account collateral by reducing the exposure to the relevant counterparty. However, if this is done it is important that the collateral not be reflected in any other calculation, since this would result in double-counting.

15.34  The requirements for the establishment, verification, and operation of a CCR model are very similar to those which apply to the development and operation of a credit model. The firm must have a control unit that is responsible for the design and implementation of its CCR management system, including the initial and ongoing validation of the model, and this unit must control input data integrity and produce and analyse reports on the output of the firm’s risk measurement model, including an evaluation of the relationship between measures of risk exposure and credit and trading limits. The unit must be independent from the parts of the business responsible for originating, renewing, or trading exposures and free from undue influence; it must also be adequately staffed and it must report directly to the senior management of the firm.

15.35  CCR models are also subject to the use test, in that a CCR model may not be used for regulatory reporting if it is not closely integrated into the actual credit risk management process of the firm. It is not permitted to use a model for regulatory (p. 311) reporting unless the results of the same model are also used by the firm itself for that purpose. CCR models must be appropriately stress tested.

15.36  Firms must be able to demonstrate to their regulator that their models have sufficient flexibility to capture general and specific wrong-way risk. General wrong-way risk is the risk that the probability of default of counterparties is correlated with another factor which is used in the calculation. An example would be an interest rate derivative with a highly leveraged counterparty—the effect of this could be that increases in interest rates could reduce the exposure to the counterparty on the specific transaction, but increase the likelihood of the counterparty actually defaulting. Specific wrong-way risk arises where the risk is embedded in a particular transaction—taking a synthetic exposure to the credit of X and accepting collateral in the form of other credit claims on X would be an example. There is no specific rule relating to the way in which this is done, but regulators will require firms to show that they have considered and incorporated into their model the existence of wrong-way risk.

15.37  As with all of the model provisions of Basel, the rules specify that the model must be validated and operated by reference to data which are validated independently of the business line, cover at least three years, and reflect a full business cycle.

G.  Contractual Netting within the CCR Regime

15.38  The netting requirements imposed in calculating CCR risk are slightly different from those applied in the calculation of on-balance sheet netting, since they must contemplate close-out and novation netting as well as simple set-off. However, the basic principle remains the same—netting may not be recognized unless it is legally robust and supported by appropriate legal opinions.

15.39  The CCR netting regime applies to simple bilateral agreements relating to individual products or groups of products (generally referred to as master agreements). There are three groups of products for this purpose:

  1. (1)  financial derivatives;

  2. (2)  repo and securities lending; and

  3. (3)  margin lending.

Within these groups, netting may be recognized across different agreements. However, in order to net across these groups—in other words, to net exposures arising under a repo agreement against exposures arising under a derivative transaction—two criteria must be met. First, the firm must have in place with the counterparty ‘contractual cross-product netting agreements’—ie written bilateral agreements which create a single legal obligation covering all included bilateral master agreements and transactions belonging to different product categories. Multilateral arrangements do not fall within this classification. Secondly, the (p. 312) firm must have a recognized CCR internal model. Firms which operate the CCR mark-to-market or standardized methods may not net across these classes.

15.40  The criteria which must be satisfied before netting can be recognized are as follows:

  1. (1)  the firm must have a contractual netting agreement with its counterparty which creates a single legal obligation, covering all included transactions, such that, in the event of a counterparty’s failure to perform owing to default, bankruptcy, liquidation, or any other similar circumstance, the firm would have a claim to receive or an obligation to pay only the net sum of the positive and negative mark-to-market values of included individual transactions;

  2. (2)  the firm must be in a position to provide its regulator, if requested, with written and reasoned legal opinions to the effect that, in the event of a legal challenge, the relevant courts and administrative authorities would, in the cases described under (1), find that the firm’s claims and obligations would be limited to the net sum, as described in (1), under:

    1. (a)  the law of the jurisdiction in which the counterparty is incorporated and, if a foreign branch of an undertaking is involved, also under the law of the jurisdiction in which the branch is located; or

    2. (b)  the law that governs the individual transactions included; or

    3. (c)  the law that governs any contract or agreement necessary to effect the contractual netting;

  3. (3)  the firm must have procedures in place to ensure that the legal validity of its contractual netting is kept under review in the light of possible changes in the relevant laws;

  4. (4)  the firm must maintain all required documentation in its files;

  5. (5)  the effects of netting must be factored into the firm’s measurement of each counterparty’s aggregate credit risk exposure and the firm must manage its CCR on such a basis; and

  6. (6)  the firm must aggregate credit risk to each counterparty to arrive at a single legal exposure across transactions; this aggregation must be factored into credit limit purposes and internal capital purposes.

15.41  There is an interesting quirk in the position of EU national supervisors as regards netting. Ordinarily, the judgement as to whether a particular legal opinion constitutes a ‘clean’ legal opinion is a matter for the home regulator alone. However, under the EU CRR,5 where a regulator in one of the jurisdictions required to be covered by (2) is not satisfied that the laws of its jurisdiction are sufficiently robust to permit netting, no other regulator may recognize netting which involves that jurisdiction. Thus, regulators cannot designate their jurisdictions as ‘good’ jurisdictions, but they can designate them as ‘bad’ jurisdictions, and if they do the latter then their decision is unquestionable by other regulators.

(p. 313) 15.42  A firm must not recognize for netting purposes any contract which contains a ‘walkaway’ clause—ie a clause which permits a non-defaulting counterparty to make limited payments only, or no payments at all, to the estate of the defaulter, even if the defaulter is a net creditor. This is true whether or not it is possible to obtain a clean netting opinion on the contract in the relevant jurisdictions.

15.43  In order for a cross-product netting agreement to be recognized for netting purposes, the netting effected under the agreement and the legal opinions relating to the agreement must include all of the agreements within the cross-product netting agreement. Thus a cross-product netting agreement may only be recognized if:

  1. (a)  the netting it effects captures all (and not some only) of the master agreements which it covers;

  2. (b)  the legal opinions address the validity of the entire agreement across all of the relevant products; and

  3. (c)  the bilateral agreements included under the cross-product master continue to comply on a stand-alone basis with the requirements for recognition.

H.  CCR Models and Securities Financing Transactions

15.44  Where a firm has a CCR model which covers securities financing transactions, it should use that model to calculate its exposure. If it has a master netting agreement internal model, it may use that model. If it has both, it may choose which approach to apply. If it has neither, it may use the master netting agreement approach contained in the collateral rules (if it is permitted to do so—a firm which uses the financial collateral simple method will not be permitted to do this). If all of this fails, then it will be obliged to treat the exposure as a collateralized receivable and apply the relevant collateral method (simple or comprehensive, as appropriate) to recognize the securities financed as collateral. However, it should be noted that the financial collateral simple method is not available in respect of securities financing transactions in the trading book.

I.  Credit Derivatives

15.45  No CCR is attributed to a credit derivative entered into in the banking book, since these are already dealt with under the existing banking book rules. For credit derivatives in the trading book (including total return swaps), a PFCE (potential future credit exposure) figure must be calculated by multiplying the nominal amount of the instrument by 5 per cent where the reference obligation would be a qualifying debt security, or 10 per cent otherwise. Where the notional exposure arising from the swap represents a long position in the underlying, 0 per cent is used.

(p. 314) 15.46  For a first-to-default transaction, the appropriate percentage for the PFCE will be determined by the lowest credit quality of the underlying obligations in the basket. If there are non-qualifying items in the basket, the percentage applicable to the non-qualifying reference obligation should be used. For second and subsequent to default transactions, underlying assets should continue to be allocated according to credit quality—ie for a second-to-default transaction, the applicable percentage figure is the percentage applicable to the second lowest credit quality.

15.47  Where a credit derivative included in the trading book forms part of an internal hedge and the credit protection is recognized for the purposes of the calculation of the credit risk capital component, there is deemed to be no counterparty risk arising from the position in the credit derivative.

J.  Collateral in the Trading Book

15.48  Credit exposures can clearly be mitigated by collateral whether they are in the banking or the trading book. The treatment of collateral in the trading book is in principle the same as in the banking book, and firms are required to apply in the trading book the same collateral treatment which they apply to their banking book positions. However, the collateral rules which are applied to trading book exposures are slightly more generous than those which apply in the banking book. In particular, in the context of repo transactions, securities or commodities lending, or borrowing transactions, all instruments and commodities eligible to be held in the trading book may be recognized as eligible collateral. For long settlement transactions and financial derivatives in the trading book, commodities eligible to be held in the trading book may be recognized as eligible collateral. The volatility adjustment which is applied in such cases is the adjustment which applies to non-main-index equities.

15.49  A particular problem arises in this context as regards master netting agreements where such agreements cover both trading book and banking book transactions. Where such a master agreement covers repo or securities lending in the trading book, all of the transactions which the agreement covers must comply with the stricter eligibility rules which apply in the banking book. What this means in practice is that the relaxation in collateral eligibility previously described is not available for agreements which are subject to a master netting agreement which also covers banking book transactions.

K.  Double Default in the Trading Book

15.50  The double default calculation provided for in the IRB approach is based on the formula involving an IRB risk weighting multiplied by a function of the (p. 315) protection provider, and this approach applies in the trading book as well as the banking book. There are, however, some ways in which this approach is applied differently in the trading book. First, in the trading book, value adjustments made to take account of the credit quality of the counterparty may be included in the calculation of total exposure—this is an exception to the usual rule that valuation adjustments taken in respect of credit quality may not be included in exposure calculations.6 Secondly, if the trading book approach recognizes the credit risk of the counterparty in full, then the expected loss for the counterparty risk exposure (and therefore the risk-weighted amount under the IRB approach) must be zero.

15.51  There is a wrinkle as regards credit exposures to clearing houses. In general, exposures to clearing houses have a weighting of 0 per cent. However, this only applies to exposures ‘resulting from’ transactions cleared by the clearing house. A clearing house member may make payments to the clearing house for reasons which are not directly related to particular transactions (for example, where a clearing house enters into money market transactions with banks as part of its proprietary treasury operations). These payments are not covered by the 0 per cent weighting and are therefore treated as exposures to a regulated financial institution.

15.52  For the purposes of counterparty credit risk, a firm may net exposures arising from items in the trading book against exposures arising from items in the non-trading book. Where this is done, the net balance must be allocated to whichever book had the greater gross balance—thus if there is a large positive exposure in the trading book and a smaller negative exposure in the banking book, the resulting net positive balance must be dealt with under the trading book rules. However, this calculation must be performed carefully, since some rules may not tally—for example, if the net balance falls within the banking book, the calculation will have to be reperformed, since some of the collateral recognized in the trading book may not be eligible under the banking book rules.

L.  Rules Common to Banking and Trading Books

Unsettled transactions

15.53  Unsettled transactions may arise in respect of both the trading and the non-trading book, and this rule applies in respect of both. No transaction is instantaneously settled, and all transactions are unsettled for some period of time, even if that period is measured in minutes. The effect of this rule is therefore to define the point at which a capital charge is required to be taken on the basis that the period for which the transaction has remained unsettled has become excessive.

(p. 316) 15.54  The rule applies to securities, currency, and commodities transactions but does not apply to repo or securities lending exposures. It begins with the establishment of the due settlement date. This will be determined by reference to practice in the particular market concerned. Once this date has passed, the firm must calculate its potential settlement exposure. This is defined as the amount which the firm could lose if the trade were not to settle, and is calculated as the difference between the agreed settlement price of the transaction and the current market value. Thus, if a firm has agreed to buy 100 securities at £1 each, if the price of the securities rises to £2 after the settlement date then the potential settlement exposure is £100 (£1 × 100 = £100, less £2 × 100 = £200). Note that if the price of the securities were to fall rather than rise, there would be no potential settlement exposure, since the firm would not be exposed to any risk of loss.

15.55  The capital requirement for an unsettled transaction is calculated by multiplying the potential settlement exposure by a factor. The factor is set out in Table 15.7.

Table 15.7  Risk Factors for Unsettled Transactions

Number of Working Days after Due Settlement Date

Factor

5–15

8%

16–30

50%

31–45

75%

46 or more

100%

Note that this rule may be disapplied in cases of a system-wide failure of a settlement or clearing system until the situation is rectified.

Free deliveries

15.56  In modern securities, currency, and commodities markets, the ordinary mechanism for the settlement of transactions is delivery against payment (DvP), by which delivery of securities and the payment of the price occur simultaneously. However, there may be circumstances in which, for a variety of reasons, one side is prepared to permit one half of a transaction to be performed before the other half. For the reporting bank, this could happen in one of two ways—it could either pay for assets before receiving them, or deliver assets before receiving payment for them. Both of these circumstances are caught by the free delivery rule. The rule on free deliveries applies in respect of the trading book and the non-trading book. It applies only where at least one day has elapsed between the payment and the non-delivery.

15.57  Free delivery treatment varies according to whether the free delivery is in the banking or the trading book. In the banking book, the treatment resembles that for unsettled transactions, in that it begins with a determination of the positive exposure of the firm. Thus, if the firm has paid for but not received 100 securities (p. 317) valued at £1, if the value of the securities increases to £2 then the firm’s exposure is £100. For a banking book exposure, from the payment date to a day four days thereafter, the firm may treat itself as having a counterparty exposure to the transaction counterparty calculated in accordance with its normal means of calculating banking book counterparty exposures. However, when the transaction becomes more than five days old, the transaction—plus any positive exposure arising on the unsettled transaction—must be deducted from capital.

15.58  In the trading book, the position is slightly different, in that up to the first capital payment leg there is no capital charge, but thereafter the position is treated as for the banking book. An IRB firm may, however, elect not to perform this calculation, but simply to allocate standardized weighting to all such exposures in the trading book. Finally, a blanket 100 per cent weighting may be applied to all free delivery exposures where the total firm exposure arising from such exposures is not material.

M.  Basel III Foundation and CCR

Counterparty credit risk

15.59  A central feature of the Basel III view of the world is that the existing risk models failed to capture the risks inherent in the assets which banks owned. This shortfall was particularly apparent in the trading book. As a result, the Basel III foundation document prescribed a series of changes to the approach to credit risk in the trading and banking books, aimed specifically at counterparty credit risk, credit valuation adjustments, and wrong-way risk.

General wrong-way risk

15.60  This is to be captured through the introduction of a ‘worst of’ approach to EPE (see para 15.31) for banks who have permission to use internal models to calculate exposures. In effect, banks will be obliged to perform two EPE calculations: one based on three years’ historical market experience, and the other based on a ‘stress calibration’. This calculation must be done on a whole portfolio basis, and not counterparty by counterparty. A ‘stress calibration’ is performed by assessing the current portfolio against a three-year historical period which included a period of stress to the bank’s counterparties generally. An oversimplified summary of this approach is that until we suffer an even bigger market crisis than the one just gone, banks will be required to use the data from that market crisis in calculating the likely diminution in value in their portfolios. Regulators will ascertain whether the stress involved is severe enough by requiring banks to assess representative portfolios against the two (historical and stressed) data sets and noting the difference between them. If the difference is small, regulators will be likely to conclude that the level of stress used to calculate the stressed EPE is insufficient.

(p. 318) Collateralized counterparties and margin period of risk

15.61  The adjustments made by the Basel III framework provisions to over-the-counter (OTC) derivatives revolve around margin issues. It is now well known that where a market is in crisis, positions which are apparently risk-free because they are collateralized may become sources of risk. This is because where the value of such a position moves significantly, risk only remains low if the counterparty is able immediately to produce the significant amounts of new collateral necessary to ensure that the position remains fully collateralized. The relevant documentation will almost always provide that each counterparty should provide new margin daily equal to the amount of the movement (daily remargining and mark-to-market valuation) but in rapid and illiquid markets this may not alone be sufficient. The approach which Basel II takes to this is to require the bank to assume that no new collateral is provided for a set period, to model the increase in exposure which would result, and to take that increase as the exposure value (effective EPE). The basic set periods are five business days for netting sets consisting only of repo-style transactions, and ten business days for all other netting sets. However, Basel III provides that in some cases a higher period is imposed:

  • •  If a netting set includes more than 5,000 trades at any point during a quarter, a period of twenty business days is imposed for the margin period of risk for the following quarter.

  • •  If a netting set contains one or more trades involving either illiquid collateral or illiquid OTC derivatives (ie derivatives that cannot be easily replaced), a supervisory floor of twenty business days is imposed for the margin period of risk. For these purposes, ‘illiquid collateral’ means collateral that would not be readily available in stressed market conditions, and ‘OTC derivatives that cannot be easily replaced’ is to be assessed in the same way. A derivative is likely to qualify as illiquid for this purpose if it is not marked daily or is valued using models with inputs that are not observed in the market. In addition, concentration of trades and collateral on a particular counterparty should be taken into account in assessing how easily an entire position could be replaced.

15.62  It is also a known phenomenon that where banks are being called for significant amounts of margin, margin call disputes increase rapidly. Consequently an anti-avoidance rule is introduced such that if more than two margin call disputes have arisen in any particular netting set over the previous two quarters that have lasted longer than the basic set periods, then for the next two quarters the period used must be at least double the set period for the next two quarters.

15.63  Some margin agreements may not provide for daily remargining, but for remargining every n days. In such cases the period to be used is the set period plus n days minus one day. The minimum holding period for various products is summarized in Table 15.8.(p. 319)

Table 15.8  Minimum Holding Periods

Transaction Type

Minimum Holding Period

Condition

Repo-style transaction

5 business days

daily remargining

Other capital market transactions

10 business days

daily remargining

Secured lending

20 business days

daily revaluation

Downgrade triggers

15.64  Derivative documentation frequently provides that where a counterparty is downgraded, it will immediately provide further collateral to its counterparty to compensate for the increased credit risk which that counterparty is now running. However, counterparties which have recently been downgraded are not always ideally placed to provide further collateral immediately. Consequently, a provision is introduced which prevents banks from assuming in their risk models that any such collateral will be provided.

Collateral management

15.65  It is also required that a bank may only apply the IMM to modelling collateralized transactions if it has an adequately staffed and resourced collateral management unit whose responsibilities include:

  • •  calculating and making margin calls;

  • •  managing margin call disputes;

  • •  reporting levels of independent amounts, initial margins, and variation margins accurately on a daily basis;

  • •  controlling the integrity of the data used to make margin calls, and ensuring that it is consistent and reconciled regularly with all relevant sources of data within the bank;

  • •  tracking the extent of reuse of collateral (both cash and non-cash) and the rights that the bank gives away to its respective counterparties for the collateral that it posts;

  • •  tracking concentration to individual collateral asset classes accepted by the banks;

  • •  ensuring the accurate reflection of legal terms in collateral and netting agreements into exposure measurements.

The collateral management unit must report on a regular basis to senior management.

Securitization and resecuritization collateral

15.66  The general principle adopted by Basel III is that resecuritization interests are not eligible for financial collateral regardless of their credit rating. This prohibition (p. 320) applies whether the bank is using the supervisory haircuts method, their own estimates of haircuts method, the repo VaR method, or the IMM.

Securitization interests remain eligible collateral, and the standard supervisory haircut formula is amended as shown in Table 15.9.

Table 15.9  Standardized Haircuts for Securitization Collateral

Issue Rating for Debt Securities

Residual Maturity

Sovereigns

Other Issuers

Securitization Exposures

<1 year

0.5

1

2

AAA to AA-/A-1

>1 year <5 years

2

4

8

>5 years

4

8

16

A+ to BBB-/

<1 year

1

2

4

A-2/A-3/P-3 and

>1 year <5 years

3

6

12

unrated bank securities

>5 years

6

12

24

BB+ to BB-

All

15

Not eligible

Not eligible

N.  Basel III Final and CCR (SA-CCR)

15.67  The Standardized Approach to the CCR (SA-CCR) is a rewriting of the Basel CRR regime. CCR risk is currently measured through a variety of methods (current exposure method and standardized method), and both of these will be replaced by SA-CCR. Banks which use IMM to calculate exposures will continue to do so.

15.68  SA-CCR is required for credit risk capital, as well as exposures to central counterparties (CCPs) and the leverage ratio. It is particularly important for derivatives as it provides for improved netting benefit and recognition of margin for both cleared and bilateral trades.

15.69  The basis of the SA-CCR is a calculation involving two variables—replacement cost (RC) and potential future exposure (PFE). In essence, these two are added together and then multiplied by a regulatory uplift.

The formula is:

Part III Investment Banking, 15 Counterparty Risk in the Trading Book

The alpha value is the same 1.4 multiplier used in the IMM.

RC is intended to be the cost which would be incurred if the counterparty defaulted and was closed out of all transactions immediately.

(p. 321) PFE takes account of the risk that exposure may increase before the transaction is terminated. For unmargined transactions, the risk is assumed to exist for one year. For margined transactions, the expected exposure period is five, ten, or twenty days.

Replacement cost

15.70  Replacement cost (RC) is calculated at the netting set level, so positive and negative MTMs of trades are offset only when they are within the same netting set. This has the effect of giving a positive regulatory benefit to transactions where the reporting entity is out-of-the-money (ie it owes money to the counterparty). Ordinarily this would be disregarded in calculating capital requirements, but within a netting set it can be recognized.

15.71  The RC formula differs as between margined and non-margined transactions. These terms have a special meaning here. For this purpose an ‘unmargined’ transaction is a transaction in which variation margin is not exchanged over its life. An ‘unmargined’ transaction may well be (and probably is) collateralized.

15.72  For unmargined transactions, the RC formula is:

Part III Investment Banking, 15 Counterparty Risk in the Trading Book

Where:

  1. (1)  V is the sum of the mark-to-market valuations of the transactions within the netting set.

  2. (2)  C is the collateral value held by the reporting entity after taking account of haircuts for non-cash collateral. Thu,s if the institution holds a collateral balance of 100 of a type which attracts a 10 per cent haircut, C will be 90. However, if the position is that the institution has received 120 from the counterparty but provided 20 of the same type of collateral to the counterparty, C will be 108 (120 reduced by a 10 per cent haircut) minus 22 (20 increased by a 10 per cent haircut), being 86. The calculation is floored at zero, since the amount of collateral held may be greater than the MTM value of the positions.

15.73  The effect of this is that RC is the difference between the MTM value of the relevant positions and the haircutted value of the collateral held by the reporting institution, and will be zero where the institution holds collateral whose haircutted value exceeds the MTM value of the relevant positions.

15.74  For margined transactions the formula is

Part III Investment Banking, 15 Counterparty Risk in the Trading Book

This is the same calculation as before with an extra element. The new element, TH + MA – NICA, represents the largest exposure that would not trigger a call for increased variation margin (VM).

(p. 322) 15.75  The first two parts of this new element are relatively straightforward. It is normal practice in OTC derivatives to provide that although the parties are in theory required to exchange margin on a daily basis, the continual exchange of small amounts of margin is unlikely to be economically efficient. Consequently it is usual to provide both that no margin call should be made unless there has been a minimum level of change in the MTM value of the derivative (a ‘transfer threshold’ or TH) and that any margin call that is made should be for a specified minimum amount (a ‘minimum transfer amount’ or MTA). Thus TH represents the first of these—it is the positive threshold before the counterparty must provide further variation margin to the reporting institution—and MTA is the minimum amount of variation margin that may be called. For arrangements where the counterparty can call any amount of collateral that it requires without limitation (common in cleared transactions, although unusual with OTC transactions), both of these are set to zero.

15.76  NICA requires a little more explanation. Margin arrangements generally come in two components: a variable amount, which fluctuates, according to the mark-to-market value of the collateralized transactions, and a separate amount, generally referred to as initial margin or independent amount, which is established at the beginning of the transaction and does not fluctuate. In order to avoid differences of terminology, Basel has invented a new name for such collateral balances and calls them ‘independent collateral amounts’ (ICA). However, once the ICA related to a particular position or netting set has been established, it must be reduced to reflect any unsegregated collateral provided by the reporting institution to the counterparty. Thus collateral provided by the counterparty is reduced by the value of unsegregated collateral provided by the reporting entity to the counterparty to produce the net independent collateral amount, or NICA. Note that segregated collateral provided by the reporting entity is not applied in this way, since such collateral is not exposed to the credit risk of the collateral taker.

15.77  The effect of this is that RC is either the difference between the MTM value of the relevant positions and the haircutted value of the collateral held by the reporting institution or the amount by which that value could move without triggering a margin call less initial margin received. The first of these will be zero where the institution holds collateral whose haircutted value exceeds the MTM value of the relevant positions; the second will be zero where the transfer threshold plus the minimum transfer amount is less than the net initial margin balance received from the counterparty.

Potential future exposure

15.78  PFE is calculated on a netting set basis—PFEs are calculated for each asset class within a netting set and then aggregated into a total PFE. A netting set for this purpose is the set of transactions within which partial or full offsetting is recognized.

(p. 323) There are four broad asset classes for this purpose:

  1. (1)  Interest rate derivatives. A hedging set consists of all derivatives that reference interest rates of the same currency such as USD, EUR, JPY, etc. Hedging sets are further divided into maturity categories. Interest rate basis swaps must be in separate hedging sets for each pair of risk factors, so LIBOR 3m vs 6m form a single hedging set, one that is distinct from LIBOR 1m vs 3m and FedFunds vs LIBOR 3m. Long and short positions in the same hedging set are permitted to fully offset each other within maturity categories; across maturity categories, partial offset is recognized.

  2. (2)  Foreign exchange derivatives. A hedging set consists of derivatives that reference the same foreign exchange currency pair such as USD/Yen, Euro/Yen, or USD/Euro. Long and short positions in the same currency pair are permitted to perfectly offset, but no offset may be recognized across currency pairs.

  3. (3)  Credit derivatives and equity derivatives. A single hedging set is employed for each asset class. Full offset is recognized for derivatives referencing the same entity (name or index), while partial offset is recognized between derivatives referencing different entities.

  4. (4)  Commodity derivatives. Four hedging sets are employed for different classes of commodities (one for each of energy, metals, agricultural, and other commodities). Within the same hedging set, full offset is recognized between derivatives referencing the same commodity and partial offset is recognized between derivatives referencing different commodities. No offset is recognized between different hedging sets.

15.79  Allocating transactions between these asset classes is not always entirely straightforward. However, the principle enunciated by the Basel Committee on Banking Supervision (BCBS) is that most derivative transactions have one primary risk driver, defined by its reference underlying instrument (for example an interest rate curve for an interest rate swap or a reference entity for a credit default swap), and will fall into one asset class. Complex or hybrid derivatives may be allocated to more than one asset class.

15.80  If multiple margin agreements apply to a single netting set, that netting set must be subdivided into two or more netting sets, one per applicable margin agreement. If a single margin agreement applies across multiple netting sets, the RC calculation is done once for the entire margin agreement, and the result is then applied proportionately within the individual netting sets where the margin agreement applies. It may be noted that, since in this situation collateral is likely to have been called for based on the net position, whereas within the SA-CCR approach there is no offsetting of positive and negative MTMs within different netting sets, an overall surplus of collateral may result in an actual shortfall in some netting sets under the SA-CCR calculation.

(p. 324) 15.81  In each class, there are separate hedging sets for basis transactions and volatility transactions.

  1. (1)  A basis transaction is a transaction where both legs are denominated in the same currency where the cash flows depend on different risk factors from the same asset class (a basis transaction with two floating legs in different currencies should be treated as a simple foreign exchange contract). Examples would be three-month LIBOR vs six-month LIBOR or Brent Crude versus Henry Hub gas. Each basis is its own separate hedging set.

  2. (2)  A volatility transaction is one in which the reference asset depends on the volatility (historical or implied) of a risk factor. Common examples of volatility transactions include variance and volatility swaps and options on volatility indices. Volatility transactions form hedging sets according to the rules of their respective asset classes. For example, all equity volatility transactions form a single volatility hedging set. For a hedging set consisting of volatility transactions, the supervisory factor is multiplied by 5.

15.82  The PFE formula is:

Part III Investment Banking, 15 Counterparty Risk in the Trading Book

This really does not describe at all well what is actually going on. The basic PFE requirement is to add a factor into the risk charge to take account of the fact that the value of an exposure may change faster than a reporting institution could close it out, thereby resulting in a greater loss than is visible on the point-in-time numbers. However, banks know this perfectly well, and in practice take excess collateral precisely to protect themselves against this risk. Consequently, the basis of the PFE calculation is a set of prescribed add-ons, which are applied class-by-class, and the multiplier. The multiplier then reduces this amount by a percentage which is proportional to the amount of overcollateralization. ‘Proportional’ in this context is important—the add-on does not reduce one-for-one, and the multiplier is floored at 5 per cent. This means that that no matter how much overcollateralization is present for a particular asset class in a particular netting set, the applicable PFE will always be at least 5 per cent of the add-on requirement.

15.83  This has the effect of flooring the EAD calculation at 5 per cent of PFE multiplied by the 1.4 ‘alpha’ value.

Add-on

15.84  In general, the process by which an add-on is calculated is as follows:

  1. (1)  An adjusted notional amount based on actual notional or price is calculated at the trade level. For interest rate and credit derivatives, this amount incorporates a supervisory measure of duration.

  2. (p. 325) (2)  A maturity factor (MF) reflecting the time horizon appropriate for the type of transaction is calculated at the trade level and is applied to the adjusted notional. There are two different calculations for the MF, one for margined and one for unmargined transactions.

  3. (3)  A supervisory delta adjustment is made to this trade-level adjusted notional amount based on the position (long or short) and whether the trade is an option, collateralized debt obligation (CDO) tranche, or neither, resulting in an effective notional amount.

  4. (4)  A supervisory factor (SF) is applied to each effective notional amount to reflect volatility.

  5. (5)  The trades within each asset class are separated into hedging sets and an aggregation method is applied to aggregate all the trade-level inputs at the hedging-set level and finally at the asset-class level. For credit, equity, and commodity derivatives, this involves the application of a supervisory correlation parameter to capture important basis risks and diversification.

Adjusted notional

15.85  For interest rate and credit derivatives, the trade-level adjusted notional is the product of the trade notional amount, converted to the domestic currency, and a supervisory duration (SD) factor. The effect of the SD factor is to increase the notional proportionately to the duration, so a three-month swap with a £10m notional amount will have a lower notional value for this purpose than the same swap with a two-year term. For foreign exchange derivatives, the adjusted notional is the value of the foreign currency leg of the transaction converted to the domestic currency (if both legs are in currencies other than the domestic currency, both are converted and the larger of the two values is used). For equity and commodity derivatives, the adjusted notional is the market price of one underlying unit multiplied by the number of units referenced.

Supervisory delta adjustments

15.86  These parameters are also defined at the trade level and are applied to the adjusted notional amounts to reflect the direction of the transaction and its non-linearity. For derivatives that are not options or CDO tranches, the value of this parameter is +1 for long (MTM increases when the value of the primary risk factor increases) or –1 for short (MTM decreases when the value of the primary risk factor increases).

Supervisory correlation parameters

15.87  These parameters only apply to the PFE add-on calculation for equity, credit, and commodity derivatives. For these asset classes, the supervisory correlation parameters are derived from a single-factor model and specify the weight between systematic and idiosyncratic components. This weight determines the degree of (p. 326) offset between individual trades, recognizing that imperfect hedges provide some, but not perfect, offset.

Add-on for interest rate derivatives

15.88  The add-on for interest rate derivatives captures the risk of interest rate derivatives of different maturities being imperfectly correlated. To address this risk, the SA-CCR divides interest rate derivatives into maturity categories (also referred to as ‘buckets’) based on the end date of the transactions. The three relevant maturity categories are: less than one year; between one and five years; and more than five years. The SA-CCR allows full recognition of offsetting positions within a maturity category. Across maturity categories, the SA-CCR recognizes partial offset.

15.89  The add-on for interest rate derivatives is the sum of the add-ons for each hedging set of interest rates derivatives transacted with a counterparty in a netting set. The add-on for a hedging set of interest rate derivatives is calculated in two steps.

15.90  For a basis transaction hedging set, the supervisory factor applicable to its relevant asset class must be multiplied by 0.5.

Footnotes:

1  Netting may not be recognized in calculating such an exposure, which must be treated as if it constituted a separate netting set containing only itself.

2  Subject to a restriction that a rate of 0 per cent may not be applied to any exposure arising under a resetting contract with a term of over one year, even if the next reset is within one year and a rate of 0 per cent would otherwise be applicable under the table. In such a case a 0 per cent weighting is used instead.

3  Perfectly matching contracts included in the netting agreement may be taken into account as a single contract with a notional principal equivalent to the net receipts.

4  Note that a firm should be capable of establishing EE daily, and should be able to establish sufficient data points in the forthcoming year to adequately reflect the time structure of future cash flows.

5  Capital Requirements Regulation (Regulation No 575/2013/EU), Art 296(2).

6  Note that valuation adjustments used in this way may not also be counted towards upper Tier 2.