IMPACT OF THE NEW

CVA RISK CAPITAL CHARGE

Including final targeted revisions from July 2020

analytical. quantitative. tech. August 2019 / February 2021

Content

1. Introduction Page 3

2. Sample Calculations Page 4

3. Calculation Results Page 6

4. Conclusion Page 9
1. Introduction portfolio is:

_
In July 2020, the Basel Committee on Banking Supervisi- ​K​ CVA​= 2.33 ​√ h    ​  
on (BCBS) has published the final targeted revisions to the _____________________________________
​  ​​(    
​∑ ​​ ​0.5 ∙ ​wi ​​ ∙ ​M ​i​ ∙ ​EAD​ i​)​ ​​ 2​ + ​∑ ​​ ​0.75 ∙ ​wi ​2 ​​ ∙ ​​(​M ​i​ ∙ ​EAD​ i​)​ ​​ 2​ ​ ​
calculation of the new CVA risk capital charge (BCBS 507, √ i i
see [10]), with implementation date January 2023. The
motivation for the new framework (BCBS 424, see [7]) is to where:

▪ ensure that all important drivers of CVA risk, ▪ h is the risk horizon, i.e. 1 year
including CVA hedges, are covered in the Basel ▪ i is the considered counterparty
regulatory capital standard ▪ ​​w​ ​​​ is the counterparty risk weight based on rating class
i
▪ align the capital standard with the fair value (1 to 6)
measurement of CVA applied under various ▪ ​​Mi​  ​​​is the effective maturity of the set of transactions
accounting regimes with counterparty i
▪ ensure consistency with the proposed revisions to ▪ ​​EAD​  ​​​ is the regulatory exposure for counterparty i,
i
the market risk framework under the Basel determined according to one of the regulatory
Committee's Fundamental Review of the Trading Book methods (such as e.g. the Current Exposure Method)
(BCBS 352 and 457, see [8])

Future Basic Approach

This paper highlights the key differences between current
and future calculation approaches for regulatory CVA
risk capital charges. The Basel Committee proposes two Banks must use the BA-CVA, which closely relates to
new CVA risk capital frameworks1, to acknowledge banks’ the current standardized method, unless they receive
different capabilities regarding the computation of CVA approval from their relevant supervisory authority to use
sensitivities: the SA-CVA. Improvements to the current standardized
approach include a better recognition of eligible credit
▪ the basic approach (BA-CVA), based on a formula risk hedges. Exposure hedges will remain excluded from
similar to the current standardized method capital charge calculations in the BA-CVA. The formula
▪ the standardized approach (SA-CVA), which is for the basic approach in its reduced form (without CDS
based on CVA sensitivities or CDS index hedges) is (see [7]):

​​​​DS​ BA-CVA ​​∙​ K​ reduced​= ​
Comparing BCBS 507 with BCBS 424, the main revisions _______________________
are a multiplicative factor of 0.65 for BA-CVA, a reduc-
√ρ ∙ ​∑​ ​ ​​SCVA​ c​)​​​ 2​ + ​(1 - ​ρ​​ 2​)​ ∙ ​∑​ ​ ​​SCVA​ 2c​ ​ ​​
DS​ BA-CVA​​ ∙ ​  ​​(   
c c
tion of the SA-CVA multiplier from 1.25 to 1 as well as the
reduction of several SA-CVA risk weights (in particular where:
for interest rates, FX and volatilities). The committee also
adjusted the scope by exempting certain SFTs and client ▪ ρ = 0.5
cleared derivatives and introduced a new handling of ▪ c is the considered counterparty
credit and equity indices. ▪ ​​SCVA​ c​​​is the CVA capital requirement that counter-
party c would receive if considered on a stand alone
In section 2, we define different sample portfolios as basis, calculated as: ​
​RW​  ​​
well as a market data environment which we then use to SCVA​  c​= _ ​ α ​ ∙ ​∑​  NS∊c​​ M
c
​ ​ NS​​ ∙ ​EAD​  NS​​ ∙ ​DF​  NS​​​
calculate example CVA risk capital requirements for the ▪ ​​RW​ c​​​is the counterparty risk weight based on credit
different approaches. In section 3, we present and discuss quality (grade) and sector (similar to SA-CVA)
our calculation results. We conclude by summarizing and ▪ α is the so-called conversion factor which is
by recommending courses of action. defined for CCR capital in the IMM and in the SA-CCR;
it is currently set to 1.4

Current Standardized Approach 1 The internal model approach IMA-CVA, which was discussed in a consultative docu-

ment (BCBS 325, see [4]) and in the CVA QIS (see [5]), has been eliminated later on; the

elimination was announced in a consultation paper regarding credit risk RWA (BCBS

The current standardized approach for the calculation of 362, see [6]) and confirmed in BCBS 424 (see [7]). A further approach may be followed
the CVA risk capital charge has been implemented as part by banks with less than 100 billion EUR notional in uncleared bilateral derivatives.
of the Capital Requirements Regulation (CRR, see [13], These may choose to set the CVA risk capital equal to their capital requirement for
article 384); the corresponding formula for an unhedged counterparty credit risk (BCBS 424, see [7]).

Page 3
▪ ​​​MNS
​  ​​​is the effective maturity of the netting set NS The multiplier ​​m​ CVA​​​has default value 1 (see [10]) but may
▪ EAD​  NS​​​is the regulatory exposure at default of the be increased by the bank’s supervisory authority, e.g. to
netting set NS, determined according to the SA-CCR capture wrong way risk, and the ​​γ​  bc​​​ are regulatory corre-
(see [2]) lation parameters (see [7]). The so-called bucketed capital
▪ ​​DF​  NS​​​is the supervisory discount factor2 of the netting charges ​​Kb​  ​​​ are defined as:
set NS, which is a function of the effective maturity ____________________________________
​​​K ​b​= ​  ​(​  ∑​ ​ ​WS​ 2k ​ ​+ ​  ∑​ ​ ​  ∑​  ​ ​ρ ​kl​ ∙ ​WS​ k​ ∙ ​WS​ l​)​ + R ∙ ​  ∑​ ​​​​(​WS​ Hdg
k )
▪ ​​DS​  BA-CVA​​​= 0.65 is the discount scalar introduced in [10]
√     
k∊b k∊b l∊b,l≠k k∊b
​  ​ ​​​ 2​ ​​​

Differences between the current standardized approach

and the basic approach are further outlined in section 3, These are based on weighted sensitivities ​​WS​ k​​​ as well
where the resulting quantitative impact is discussed. as intra-bucket correlation parameter ρ​​ ​  kl​​​, and R is the
hedging disallowance factor, set to 0.01, that prevents the
possibility of perfect hedging of CVA risk.
Future Standardized Approach (SA-CVA)
Furthermore, ​​Sb​  ​​​has been introduced in BCBS 507 (see
[10]) as the sum of weighted sensitivities in bucket b,
The new standardized approach for CVA risk (SA-CVA) is capped at ​​K​b ​​​and floored at ​​-K​b  ​:
based on CVA sensitivities, following the idea and princi-
{ ( k∊​b​ )}
ples of the standardized approach for market risk (SA-TB) ​​Sb​ ​​ =​ max ​​ -​Kb​  ​; min ​ ​  ∑​ ​​ ​W S​k  ​​ ; ​Kb​  ​ ​ ​​
defined as part of the FRTB (see [8]). In contrast to SA-TB,
banks need supervisory approval to use SA-CVA and the Compared to previous formulations of SA-CVA, this
following requirements must be fulfilled (see [7]): allows for a better offsetting of weighted sensitivities with
opposite signs.
1 The calculation of regulatory CVA and CVA sensitivi-
ties from predefined market risk factors must be based
on the simulation of exposure paths (discounted with the 2. Sample Calculations
risk-free interest rate).

2 Default probabilities must be market-implied and the To carry out sample calculations we define synthetic
credit spreads of illiquid counterparties must be approxi- portfolios and consider actual market data3 in order to
mated with a suitable method. obtain a realistic assessment of potential CVA risk capital
charges. Our calculations of SA-CVA and BA-CVA capital
3 The bank must have a dedicated CVA desk or similar charges are based on the risk weights that have been
function responsible for risk management and hedging of introduced in the latest BCBS paper (BCBS 507, see [10]).
CVA.

Banks qualifying for SA-CVA need to follow general prin- Sample portfolios and market data
ciples to calculate regulatory CVA. The exposure scenari-
os must be calculated from the same models, calibration,
market, and transaction data as used for accounting CVA Our sample portfolios consist of interest rate swaps (EUR)
(and/or front-office CVA). Model calibration must be and cross currency swaps (USD/EUR) and cover typical
carried out with respect to market-implied parameters maturities and moneyness levels. They include the follo-
wherever possible. The recognition of collateral requires wing spot starting trades:
the capturing of all relevant contractual features and a
margin period of risk with a supervisory floor of at least A. Vanilla swap
10 business days. The requirements on documentation,
independent validation, and processes show similarities ▪ Currency: EUR
to those applicable to internal model banks (compare ▪ Notional: 100m
IMM-CCR). ▪ Maturities: 1y, 4y and 10y
▪ Pay: annual fixed rate, fixed AT PAR4
Generally, CVA risk assumes that the bank itself is default ▪ Rec: bi-annual floating rate, fixed at Euribor 6M
risk-free. In particular, it disregards the debt valuation
adjustment (DVA).

The formula for regulatory capital based on SA-CVA is:

2 For banks using the IMM to calculate EAD, the supervisory discount factor is set to 1.
__________________
​​​K​ CVA​= ​m​ CVA​∙ ​  ​∑
  
​ ​ ​​K​ 2b ​ ​+ ​∑​ ​ ​​  ∑​ ​ ​γ​ bc​​ ∙ ​Sb​ ​ ∙ ​Sc​ ​
√ b b c≠b
3 This market data reflects a snapshot taken as of the end of October 2017.

4 We assume trades at par to have a market value slightly positive, ergo NGR is set to 1.

Page 4
B. Cross currency swap Simulation approach

▪ Currency: EUR/USD For BA-CVA and SA-CVA, we implement a framework

▪ Notional5: 100m EUR vs USD amount corresponding that complies with the standards and definitions laid out
via spot FX in the finalized BCBS framework (BCBS 424, see [7]) and
▪ Maturities: 1y, 4y and 10y with the final targeted revisions (BCBS 507, see [10]). For
▪ Pay: bi-annual floating rate, fixed at Euribor 6M CVA sensitivity calculations, we use a hybrid Hull-White
▪ Rec: bi-annual floating rate, fixed at USDLibor 6M model with one driving Brownian motion per IR curve
and one driving Brownian motion per FX rate. Exposures,
C. Same swap legs as Trade A but with fixed rate = variation margin and initial margin (IM) are simulated
AT PAR · 1.2 and maturity 10y within the same Monte Carlo framework. The pathwise
simulation of IM is based on the Nadaraya-Watson kernel
D. Same swap legs as Trade A but with fixed rate = regression method [1]. As required by BCBS 424, sensi-
AT PAR · 0.8 and maturity 10y tivities are based on 1 bp tenor shifts for IR and credit
spread delta and on relative 1% shifts for FX delta, as well
All calculations are carried out using the regulatory as relative parallel 1% shifts for IR and FX volatilities for
formulas that apply in the non-hedged case. Both the vega sensitivities. All calculated sensitivities are used
uncollateralized and the collateralized case (with and wit- as an input to our aggregation tool, which computes the
hout initial margin) are considered. The CSA is assumed CVA risk capital charges. An important additional input
to have a minimum transfer amount (MTA) of 500k EUR6, for the calculation of capital charges is given by each
a bilateral threshold of zero and a margin period of risk counterparty’s risk weight. Our selection of investment
(MPOR) of 10 business days. Initial margin is assumed to grade financial and corporate9 counterparties leads to
cover the 99% confidence interval of the clean PnL over the following counterparty risk weights for the current
the MPOR, in line with BCBS 3177 (see [3]). standardized approach and the future basic approach,
and credit spread delta risk weights for the future stan-
For the sample calculations, we consider two different dardized approach:
types of counterparties:

1 A financial counterparty with an investment grade Approach Financial Corporate

rating. Interbank portfolios typically benefit from signifi-
cant netting effects, and they are generally collateralized, Current
with the collateral mitigating exposures at default. Standardized 1.0% 1.0%
Approach
2 A corporate counterparty with an investment grade
rating. Portfolios with corporates are generally smaller, BA-CVA 5.0% 3.0%
with no large netting effects and no collateral in place.
We point out that the exemption of certain corporate SA-CVA 5.0% 3.0%
counterparties with respect to CVA risk, which is currently
valid under European regulation, is not recognized in this Table1: Risk weights for the current and future
paper. approaches

The following market data as of 30 October 2017 is used

for the calculations:

▪ Interest rate (IR) curves for both currencies

(discounting via EONIA and USD FEDFUNDS and 5 The notionals of Trade B are assumed to be exchanged at the start and maturity of
projection of forward rates via Euribor and USDLibor) the swap.
▪ IR volatilities based on co-terminal ATM swaptions up 6 According to BCBS 317 (see [3], 2.3), “all margin transfers between parties may be
to 30 years subject to a de-minimis minimum transfer amount not to exceed €500,000”.
▪ EUR/USD FX spot, FX basis curve, FX volatilities up to 7 The threshold of the initial margin is set to zero (BCBS 317 [3] allows max. €50m) and
fifteen years it is assumed that Trade B is not subject to the FX notional exemption (which may be
▪ A flat CDS spread of 40 bps8 for both the financial and applied to cross currency swaps according to BCBS 317 [3]).
the corporate counterparty 8 For reasons of generality and comparability, we use an artificial CDS spread instead
▪ A generic recovery rate of 40% for PD and CVA of actual market data.
computations 9 The corporate is assumed to belong to one of the following sectors: basic materials,
▪ Realistic correlations for IR and FX, based on a 4Y energy, industrials, agriculture, manufacturing, mining and quarrying, consumer
history of zero rates and FX spot rates goods and services, transportation and storage, administrative and support service

activities.

Page 5
3. Calculation Results Trades No CSA CSA

Trade A 1y 11,364 11,364

Current CVA risk capital charge Trade A 4y 42,236 42,236

We first compute current CVA risk capital charges, using Trade A 10y 275,035 275,035
the methodology described in article 384 of the CRR (see
[13]), ignoring the corporate exemption. We calculate A {1y,4y,10y} netted 257,697 257,697
EADs according to the Current Exposure Method (CEM)
as described in article 274 of the CRR10 and we recognize Trade B 1y 113,635 113,635
netting according to article 298 of the CRR. Since we as-
sume the same credit quality “3” for both counterparties, Trade B 4y 422,357 422,357
we obtain identical risk weights of 1% and, thus, identical
CVA risk charges. Trade B 10y 1,375,175 1,375,175

Note that only Trade D has positive market value, resul- B {1y,4y,10y} netted 1,803,880 1,803,880
ting in a difference between collateralized and uncollate-
ralized replacement costs. Trade C 10y 110,014 110,014

Trade D 10y 584,132 366,713

10 In the CRR, the CEM is referred to as the Mark-to-Market method. IM is not men-

tioned in this method, hence we do not distinguish between collateralization with or Table 2: Calculation results for the CVA risk capital
without IM for the current CVA risk capital charge. charge under the current standardized approach

EAD CEM EAD SA-CCR

Trade No CSA CSA No CSA CSA CSA + IM

Trade A 1y 500,000 500,000 682,811 904,836 640,221

Trade A 4y 500,000 500,000 2,537,639 1,461,200 330,465

Trade A 10y 1,500,000 1,500,000 5,508,517 2,352,517 654,945

A {1y,4y,10y} netted 2,500,000 2,500,000 8,728,955 3,318,565 1,103,119

Trade B 1y 5,000,000 5,000,000 5,599,966 2,379,966 1,952,070

Trade B 4y 5,000,000 5,000,000 5,600,083 2,380,000 955,881

Trade B 10y 7,500,000 7,500,000 5,599,890 2,379,890 513,034

B {1y,4y,10y} netted 17,500,000 17,500,000 16,799,897 5,739,897 2,700,504

Trade C 10y 600,000 600,000 4,452,031 1,522,948 349,993

Trade D 10y 3,185,768 2,000,000 7,868,646 4,012,646 1,310,253

Table 3: EADs for CEM and SA-CCR for different trades and maturities

Page 6
Future BA-CVA charge Future SA-CVA charge

Our calculations for the future basic approach (BA-CVA) Results for the future standardized approach are presen-
are based on EAD figures derived from the new standard- ted in Table 4. The most relevant risk factors are credit
ized approach for CCR capital (SA-CCR) presented in BCBS spreads, especially for the collateralized portfolios. For
279 (see [2]), which was scheduled to become effective by those portfolios with cross currency swaps, the FX Vega
January 1, 2017.11 cannot be neglected either. Both aspects are prominent-
ly visible in Trade B with a maturity of 10y, due to the
The SA-CCR provides a more risk sensitive approach than large exposure originating from the notional exchange
the CEM, it recognizes netting and margin agreements at maturity. For instance, in the Financial case without
in a better way, and it incorporates the IMM multiplier α CSA, the bucketed capital charge ​​K​ b​​​is around 530k EUR
to account for model inaccuracies. Table 3 demonstrates for credit spread delta, 80k EUR for IR delta and 95k EUR
that for interest rate swaps without CSA, the SA-CCR EAD for IR vega, while the total charge for FX is around 25k
is significantly higher than the EAD implied by the CEM. EUR. Since we assume identical credit spreads (and, thus,
It also shows that while the CEM recognizes CSAs only to identical CVA sensitivities) for both counterparties, the
a small degree, the SA-CCR acknowledges CSAs in a more differences between Financial and Corporate are solely
accurate way.12 due to different risk weights for credit spread delta.

Let us note that the SA-CCR requires a mapping of every The new formula
trade to the risk category representing the trade’s main
risk driver. Following the EBA FINAL draft RTS (see [12]),
{ (k∈b )}
​​ S​b  ​​ =​ max ​​ -​Kb​  ​; min ​ ​  ∑​  ​​ ​W S​k ​​ ; ​Kb​ ​ ​ ​​
we map the cross currency swap B to risk category FX.
​​
introduced in BCBS 507 (see [10]) reduces the SA-CVA
The resulting CVA risk charges for BA-CVA, which will be charge for the cross currency swap B because it allows a
compared to the other approaches further on, are presen- partial offsetting of the positive IR EUR delta with the ne-
ted in Table 4. gative IR USD delta. The largest effect of this change oc-
curs at maturity 1y, where it reduces the capital charge by
In addition to different EADs, the general differences bet- approximately 25% (regardless of the collateralization).
ween the current standardized approach and the future
basic approach are the following13:

▪ risk weights for different counterparty sectors from Comparison

the current range of 0.7% to 10% change to a new
range of 0.5% to 12% Table 4 compares the results for the current standardized
▪ there is a reduction by the conversion factor α =1.4, an method, for BA-CVA and for SA-CVA.
additional multiplication by 0.65 and the factor 2.33
does not appear explicitly any more The general observation is that the BA-CVA is unfavorable
in the uncollateralized case. In a collateralized setting, the
SA-CVA is most favorable.
Overall, for the investment grade financial counterparty
(with an old risk weight of one percent), the scaling factor For the Financial, CVA risk charges for BA-CVA show a
between the BA-CVA charge and the current standardized significant increase compared to the current standardized
approach charge (assuming unchanged EADs) is14: approach. Exceptions are the collateralized cross-curren-
cy swaps and the IM collateralized portfolios, for which
the larger collateralization benefit recognized by the SA-
___________
​RW​ BA-CVA
​  ​  ∙ ​scaling​​  BA-CVA​  5% ∙ 1.4-1 ∙ 0.65​  ​​  ​
​​ fin = ​​​ _______
 ​ 
​RW​​  curr-SA​  ∙ ​scaling​​  curr-SA​ 1% ∙ 2.33
 ​  
​ ≈ 1.00​ CCR EAD is visible.

11 We point out that the SA-CCR has still not been completely adopted by many

and for the corporate counterparty it is: countries, including the European Union (see [9]). SA-CCR is included in the CRR2 (see

[14]), which entered into force in June 2019. The rules need to be applied by banks two

years later.

= ​​ _
​RW​ BA-CVA
​  ​  ∙ ​scaling​​  BA-CVA​
___________ 3% ∙ 1.4 ∙ 0.65
-1 
​​ corp
​ ​​  curr-SA​  ∙ ​scaling​​  curr-SA​
RW
 ​  1% ∙ 2.33
 ​   
≈ 0.60​​  _​ ​  ​​ 12 For Trade A 1y, the CSA increases the SA-CCR EAD, which is due to the CSA’s mini-
  mum transfer amount (MTA). For longer maturities, such effects are dominated by the

PFE terms.

In particular, the differences between the financial and 13 The scaling factor of 1.4 included in the consultative paper (see [4]) to account for
the corporate counterparty are triggered by the different additional risk from increasing exposures was dropped in the finalized BCBS frame-
risk weights (5% vs. 3%) for the same rating class, so the work (see [7]) and instead, a factor of 0.65 has been introduced in BCBS 507 (see [10]).
ratio of financial to corporate is 5/3 = 1.67 for all trades. 14 We write curr-SA for the current standardized approach.

Page 7
Table 4: Comparison of the current standardized method, the BA-CVA and the SA-CVA
for different collateralization scenarios and different counterparty scenarios; Trade A
1y is abbreviated as A01, netting set A {1y,4y,10y} is abbreviated as AN, and similarly
for the other trades / netting sets. The spark lines indicate which method yields the
highest (red) or lowest (green) capital charge for the trade or netting set, whereas the
heat map relates all results.

Page 8
Without CSA, the basic approach significantly increases the Basel Committee has acknowledged simulation based
the capital charge for both counterparties – except for approaches to CVA, which already prevail at the accoun-
the 10y cross currency swap15 – compared to the current ting level, at the level of regulatory capital requirements.
standardized approach.
These results are also supported by the EBA impact
The ratio of Financial to Corporate for BA-CVA is flat at study16 (see [11]), which determined an increase of the
1.67, while for SA-CVA it is within the range of 1.1 to 1.6. CVA risk charge by more than 200% on average for smal-
The variation originates from varying credit sensitivities ler banks. With the removal of the corporate exemption,
between the trades which remain the dominant part of the average impact on all banks was estimated to be
the SA-CVA capital charge. 558%. The impact study also shows that, while the increa-
se in capital requirements is almost certain for small and
It is plausible that the future SA-CVA capital charge is large banks, medium sized banks have the chance to get
favorable for collateralized trades; the SA-CVA compu- through the reform with stable or even reduced capital
tations build upon real CVA sensitivities, fully acknow- requirements.
ledging the exposure mitigating effect of collateral. For
trades with maturities larger than 4y and without CSA, This chance must not be missed. Banks should prepare
SA-CVA charges are generally higher than the current now, conducting individual impact and gap analyses and
capital charge, whereas for trades with CSA this order is getting their CVA sensitivity calculations and CVA hed-
reversed. ging activities running in time, to avoid falling into the
BA-CVA capital trap.
The differences among the new approaches derived in
the unhedged case are likely to increase further in the With its extensive experience and know-how, d-fine has
case where CVA hedges are present, due to the following long been a leading consulting company for the financial
reasons (see [10]): industry. Our knowledge on financial markets, risk con-
trol and reporting, paired with deep regulatory insight,
▪ SA-CVA allows credit spread as well as exposure enables us to readily offer solutions for SA-CCR, the new
hedges, while BA-CVA only allows credit spread CVA risk charge and CVA management.
hedges
▪ the hedging benefit for BA-CVA is capped at 75% of
the unhedged BA-CVA, while SA-CVA in principle
allows hedging up to 90% of the unhedged SA-CVA ​​

4. Conclusion

Our sample calculations show the potential impact of the

two approaches that have been presented in the finali-
zed BCBS framework (BCBS 424, see [7]) and have been
revised in BCBS 507 (see [10]). The most important results
are:

▪ Compared to the current standardized approach, BA-

CVA will increase the CVA risk capital charge for many
uncollateralized portfolios
▪ Collateralized portfolios (and portfolios with exposure
hedges) will significantly benefit from SA-CVA

The new CVA risk regulation framework can be a turn- 15 Let us note that Trade B 10y without CSA is an exception since here the SA-CCR EAD
ing point for many medium-sized banks, for which the is smaller than the CEM EAD (see Table 3). This is because for FX contracts, the CEM
capital savings entailed by the SA-CVA for collateralized percentage applied to the notional changes from 5% to 7.5% for maturities exceeding
portfolios may outweigh the costs associated with the five years, which has no equivalent counterpart at the SA-CCR level.
introduction and maintenance of a Monte Carlo based 16 Since the impact study, regulatory SA-CVA parameters have been updated and
CVA sensitivity computation framework as well as an BA-CVA has been reduced by a factor of 0.65, which still leaves a significant overall
active CVA desk. For these banks it is the first time that increase.

Page 9
Appendix
References

[1] Andersen, Leif B.G. and Pykhtin, Michael and Sokol, Alexander, Credit Exposure

in the Presence of Initial Margin. July 2016

[2] Basel Committee on Banking Supervision. The standardised approach for measu

ring counterparty credit risk exposures (BCBS 279). March 2014

[3] Basel Committee on Banking Supervision. Margin requirements for non-centrally

cleared derivatives (BCBS 317). March 2015

[4] Basel Committee on Banking Supervision. Consultative Document: Review of the

Credit Valuation Adjustment Risk Framework (BCBS 325). July 2015

[5] Basel Committee on Banking Supervision. Instructions: CVA QIS. February 2016

[6] Basel Committee on Banking Supervision. Reducing variation in credit

risk-weighted assets - constraints on the use of internal model approaches (BCBS

362). March 2016

[7] Basel Committee on Banking Supervision. Basel III: Finalising post-crisis reforms

(BCBS 424). December 2017

[8] Basel Committee on Banking Supervision. Minimum capital requirements for

market risk (BCBS 352, 457). January 2016, revised January 2019

[9] Basel Committee on Banking Supervision. Sixteenth progress report on adoption

of the Basel regulatory framework (BCBS 464). May 2019.

[10] Basel Committee on Banking Supervision. Targeted revisions to the credit

valuation adjustment risk framework (BCBS 507). July 2020

[11] European Banking Authority: Basel III reforms: Impact study and key

recommendations. August 2019

[12] European Banking Authority. EBA FINAL draft Regulatory Technical Standards

on mapping of derivative transactions to risk categories, on supervisory delta

formula for interest rate options and on determination of long or short positions

in the Standardised Approach for Counterparty Credit Risk under Article 277(5)

and Article 279a(3)(a) and (b), respectively, of Regulation (EU) No 575/2013 (revised

Capital Requirements Regulation – CRR2). December 2019

[13] Regulation (EU) No 575/2013 of the European Parliament and of the Council of 26

June 2013 on Prudential Requirements for Credit Institutions and Investment

Firms and amending Regulation (EU) No 648/2012. Official Journal of the

European Union (L176). June 2013

[14] Regulation (EU) 2019/876 of the European Parliament and of the Council of 20

May 2019 amending Regulation (EU) No 575/2013 as regards the leverage ratio, the

net stable funding ratio, requirements for own funds and eligible liabilities,

counterparty credit risk, market risk, exposures to central counterparties, expo

sures to collective investment undertakings, large exposures, reporting and disclo

sure requirements, and Regulation (EU) No 648/2012

Authors

Dr Ruth Joachimi Dr Holger Plank

Manager, d-fine GmbH, Frankfurt Partner, d-fine AG, Zürich
[email protected] [email protected]

Dr Ken Lichtner Nadja Schuster

Senior Consultant, d-fine GmbH, Berlin Partner, d-fine GmbH, Frankfurt
[email protected] [email protected]

Page 10
Berlin
d-fine GmbH
Friedrichstraße 68
10117 Berlin
Germany
[email protected]

Dusseldorf
d-fine GmbH
Dreischeibenhaus 1
40211 Dusseldorf
Germany
[email protected]

Frankfurt
d-fine GmbH
An der Hauptwache 7
60313 Frankfurt
Germany
[email protected]

London
d-fine Ltd
6-7 Queen Street
London, EC4N 1SP
United Kingdom
[email protected]

Munich
d-fine GmbH
Bavariafilmplatz 8
82031 Grünwald
Germany
[email protected]

Vienna
d-fine Austria GmbH
Riemergasse 14 Top 12
1010 Vienna
Austria
[email protected]

Zurich
d-fine AG
Brandschenkestrasse 150
8002 Zurich
Switzerland
[email protected]