Response to discussion on the EU implementation of MKR and CCR revised standards
Go back
The qualitative assessment is in our view part of step 1. It will determine the risk category, or on rare occasion the risk categories, to which a product should be mapped.
Moreover, as stated in our response to Question 4, cases where a qualitative assessment would not suffice, and uncertainty remains as to which categories an instrument should belong to, should be the exception rather than the rule.
If the slicing approach was to be ruled out, an alternative to the simple addition of the full notional amount exist. This alternative approach increases the overall CCR exposure amount but in a more risk sensitive, commmensurate and less punitive way. It is depicted below based on option 2 of paragraph 72:
For the first risk category, the effective notional ‘N’ of the transaction is used to calculate the SA CCR add on;
For the second risk category, the add on shall be calculated using an adjusted notional amount N*=N·a2/a1.
The best solution would be to use the FRTB SBA sensitivities. If not possible, the formula should be slightly amended as follows:
When assuming that (R+) is log-normal (instead of R in the initial formula) we should introduce’ defined as follows:
We consider that (dR/R+) = ’ dW instead of (dR/R) = dW, we can then deduce that * R* dW = ’ * (R+) * dW, which means that in the formula should be replaced with σ^'= σ* R/(R+ λ)
[please see the formula in the attached document]
The EBA should define a flexible option.
For the sake of simplicity and stability, we ask for :
1) Use of a market convention if any;
2) In the absence of a market convention, use of the value of the λ internally used for a currency or, for example, if a normal model is used internally, a value of λ per currency agreed upon by the Risk department;
3) For the rare instances where λ is insufficiently large for a trade, use of a higher λ for that trade, for instance such that : Min{P+λ, K+λ}=1%.
Moreover, we would like to take the opportunity of this consultation to remind that in our view the fall back approach for option supervisory delta should be +/-0.5 rather than +/-1 as in the European Commission draft Regulation amending Regulation 575/2013 (cf. article 279a(1)(c)). A supervisory delta of +/-0.5 was the approach taken in the draft Basel standard (BCBS 254, paragraph 48] and justified as follow (cf. BCBS 254, footnote 10):
“The Basel Committee chose to set delta to 0.5 for non-linear instruments because (i) this is the mid-point of the spectrum of all possible deltas from 0 to 1 that strikes a good balance between trades being “free” (delta is 0) and overly effective hedges (delta is 1); (ii) this is the only choice of delta that respects the put-call parity (a combination of a bought call and a sold put of the same strike is equivalent to a forward). In this context, sold call (put) options are treated as bought put (call) options. Although sold options do not present counterparty credit risk on their own, supervisory delta adjustments are included in a hedging set because they affect the mark-to-market value of the entire hedging set.”
Examples in addition to those noted in Paragraph 102 include:
o Changes in GAAP requires instruments that was booked under amortized cost to be fair valued or vice versa ;
o Changes in the intent and ability to trade (due to liquidity and other market condition).
We also propose that regular buy and sell activity between a trading desk and another distinct banking book unit of the bank which acts as a client should be exempted from the rule on movement between regulatory books, provided transactions are conducted on an arm’s length basis.
Dramatic shift in the liquidity conditions of a large portion of financial instruments that may lead to re-consider the management intent on such instruments;
A modification in the accounting standards that implies the need to change the accounting valuation for certain instruments (implementation of IFRS 9 being a recent illustration);
A change in the business model of some activity affecting financial assets.
We propose that regulators allow banks discretion to re-assign instruments between regulatory books due to exceptional circumstances beyond banks’ control through an implicit approval process (“notification approach”) rather than having to seek approval from regulators for each re-assignment (“authorization approach”).
We are looking for the conclusions of the EBA discussion paper on the treatment of structural FX positions.
There may be a single or multiple such notional trading desk for each of FX and commodity positions.
It is our view that the banking book notional trading desks shall generally be kept separate from trading book desks, if only because an actual trading desk business strategy and risk management structure (cf. BCBS d352, Appendix A) is unlikely to apply to the banking book exposures.
Nevertheless, risks transfers from notional trading desks to trading desks should be taken into account so that hedging benefits could be recognized.
Hence our view can be summarised as follow:
Notional desks for commodities and FX exposures are only for the purpose of own funds requirements calculation.
Notional trading desks should include only non-trading book positions
The trading desk qualitative requirements shouldn’t apply to the notional desks. In particular it seems inappropriate to consider that non-trading book activities should have the same constraint as trading book activities relating to profitability, business plan, etc.
Nevertheless we would like to point that in our view the volatility and variance swaps should not be subject to an exotic risk add on (cf. BCBS d352 footnote 14) since it is possible to capture their volatility risk within the SBM capital charge.
Indeed, though not an optional product and hence theoretically not subject to a Vega risk charge, volatility and variance swaps are risk managed with sensitivities to implied volatility.
Hence it would make sense to have a caveat for those non-optional products and capitalise them within the SBM Vega risk charge rather than in the risk insensitive RRAO framework. It would be a similar treatment to the one of CMS (cf. BCBS FAQ d395).
Payoffs deemed exotic by the regulatory text are usually managed in a prudent manner by banks: some of their features are smoothed so that they can be managed as a static set of vanilla options: US or European exercise. If a bank can demonstrate and document that it systematically books and manages such payoffs as portfolios of vanilla options, theses payoffs should benefit from an exclusion of the scope of RRAO charge.
Examples:
European Exercice Digital payoffs prudently managed as a call (resp. put) spread of vanilla european exercise options which are not in the scope of the exotic payoff RRAO
Bermudean / Corridor options prudently managed without correlation
Also, we would like to clarify that in our view the exclusion of the draft Regulation amending Regulation 575/2013, article 325v(4)(c) also apply to back-to-back exactly matching transactions with premium payments spread.
- The notional be capped by the maximum loss if this last quantity is bounded,
- Appropriated weights be defined for leveraged products,
- The weights increase with the time to maturity, when the RRAO is generated by correlation risk, explained by materiality considerations.
This is illustrated by the following:
Asian options: Asian options should not be considered as bearing “gap risk”, or any other residual risk. Indeed they have a very smooth behavior (smoother than plain vanilla options), and do not present additional risk compared to those caught by the SBM. In this respect, even if they cannot be perfectly replicated as a finite linear combination of vanilla options, they should not be subject to the RRAO. This is particularly penalizing for the commodity business.
Vanilla options on commodities differential: Another case to consider are vanilla options on commodities differential (for example: options on the spread between two oil qualities), it must be clear that they can be seen as options on one underlying, and that they are excluded from the RRAO.
CMS options: the RRAO measure is over penalizing, and if no appropriate weight is defined they should be excluded. This consideration may be generalized to all leveraged products.
Binary / digital options: there is a problem with the RRAO computation, and if it is not solved that would appeal for exclusion. The problem is that with respect to the RRAO, the appropriate notional of a (vanilla) barrier option is the “size” of the digit, not the full notional of the option. This would make homogeneous capital charges for a pure digit, and for the digit embedded in a barrier option. Hence, for “vanilla digits” (meaning: payoffs having no other exotic features that the digit), we propose to retain the size of the digits and not the notional of the options:
o Pure digit paying €100 if Stock A > Strike: notional = €100.
o Barrier call option, strike K, barrier K’>K, for N stocks A: notional = Nx(K’-K).
o Call spread strategy with strikes K and K’, for N stocks A: notional = Nx(K’-K).
Contractual features: In some instances contractual features other than the product type may be considered to determine a product eligibility to the residual risk add-on since those contractual feature may have a strong influence on the level of residual risk. For instance, the remaining maturity may be considered for some products to decide on their eligibility to the residual risk add-on. It is the case of spread options (options on curve slope) which may be liable for a residual risk add-on only if of long maturity.
Bermudan options: By construction, a Bermudan option value is framed by the corresponding swaptions (inferior bound) and cap/floor (superior bound). Both swaption and cap/floor are vanilla products which are entirely covered in SBM.
Nevertheless the correlation risk is very limited on vanilla Bermudan options with short callable periods and should generate no or fewer RRAO. A maturity threshold could be 10 years, because indeed the market liquidity is very important for bermudan options with maturity less than 10 years in line with small correlation risk on those bermudas.
This is illustrated hereunder where sensitivities in basis points by maturity are given for a 1% shock on the Bermuda market correlation, and where it is shown also how the RRAO should increase as a time to maturity function up to the 10 years threshold.
[Please see the graph in the attached document]
Recommendation for Bermudan options:
RRAO notional multiplicator should linearly increases from 0.0% to a 0.1% cap as Tenor (callable period) increases from 0Y to 10Y.
Hedging strategy: The determination of the notional to be used might be unclear for certain instruments. Other residual risks may also depend on factors such as the hedging strategy (a pure back-to-back would not bear market risk), or the residual maturity of the instrument (where risk might be significantly different for short on long term positions), which is not taken into account in the RRAO add-on for other.
Example: spread options (options on curve slope) are not fully replicable by a portfolio of simple options as they are sensitive to the correlation between both handles of the slope c (Cega). They therefore are subject to additional capital (RRAO) in SA. Two issues can be identified for those derivatives:
o It is unclear how to measure the notional of the trade. (Is it the “face amount of the trade” or “the sum of all its leveraged legs”?). The later choice is subject to the way the deal is structured for an equivalent amount of Cega and unfortunately sensitive to the number of instruments used to replicate the transactions.
More precisely: according to the hedging strategy the RRAO capital charge would be very different:
0 for perfect B2B,
Par x 0,1% if unhedged,
2 x Par x 0,1% for a hedge with the same deal shifted by a couple of days,
(n+1) x Par x 0,1% if the hedge is structured by n trades with different maturities.
o For short to medium dated option, the 0,1% RRAO capital charge for Cega is far too high.
Example: a spread option 2Y/10Y de 500M€ maturity 4 months and a half, one can compute 500k€ of RRAO as of 31/12/2017, with a Cega sensitivity almost zero (<10€…). The same spread option with a 10Y maturity would have a 79K€ Cega making the 500k€ of RRAO more in line (a 10% Cega shock super conservative, 0.1% implies a 39% shock for a 3Y option…).
Result: for a CIB with a classical FI product mix, under 5Y RRAO for spread option can represent up to 3/4th of the total which looks abnormal.
Recommendations and proposal for Spread Options:
o Leverage should not be the base of the notional computation, only the par face value of the option should be taken into account;
o RRAO notional multiplicator should linearly increases from 0,0% to a 0,1% cap as expiries increases from 0Y to 10Y.
It should be complemented by the fact that if an instrument qualifies for two residual risks (for example correlation and digital risk), it must be counted only once in the computation.
Also, we would like to take the opportunity of this consultation to reiterate our view on the draft Regulation amending Regulation 575/2013, article 325be(4): We consider that credit institutions should have the flexibility not to account for the maturity of the position when computing the effective liquidity horizon (LH) of risk factors of that position, provided that such option is applied consistently across all risk factors of all positions. Indeed applying the draft Regulation amending Regulation 575/2013, article 325be(4) as is (i.e. capping the liquidity horizon of each instrument to the instrument maturity) will be operationally extremely challenging, bring inconsistency with the constant risk assumption prevailing in the expected shortfall (ES) formula and result in broken hedges.
As an example, a Q&A could confirm that agencies and supras should be classified as sovereign.
What matters to the banking industry most is consistency across various regulations. We would therefore suggest that the list of liquid currency pairs and liquid interest rate currencies are aligned with that of other regulations.
Actually, as mentioned in our response to question 42 we support the alignment between regulations wherever possible and this is a good example of such desirable alignment.
Valuation adjustments for which a separate regulatory capital treatment has been specified as part of the rule (e.g. CVA);
Valuation which are deducted from CET1 (e.g PVAs);
Valuation adjustments that are updated at a less than daily frequency in the measure of P&L.
This list of exclusions is generally fit for purpose. However, there may be specific daily value adjustments that are not market risk related: They should not be included in the Hypothetical P&L (cf. EBA Discussion paper EBA/DP/2017/04, article 177) and neither the Risk Theoretical P&L since RTPL should be in line with Hypothetical P&L (HPL) (cf. EBA Discussion paper EBA/DP/2017/04, article 226).
For instance some risk factors may be subject to a daily IPV (independent price verification) value adjustment. IPV is not a market risk adjustment and it cannot be captured in the risk model hence it should be excluded to both HPL and RTPL.
Hence we support the EBA view expressed in the Discussion paper EBA/DP/2017/04, article 186 stating that, on a case by case basis, daily adjustments may be excluded from HPL, and hence RTPL, conditional to a supervisory authorisation.
“186.Finally, only the valuation adjustments related to market risk that are updated daily and are not in the above list of systematic exclusions would be included in the hypothetical P&L (at the ‘top of the house’ only or also at desk level, depending on the level at which these VAs are computed) […].”.
Hence, most daily valuations that should not be included in the Hypothetical P&L (see example in our response to question 47) will be clearly identified and excluded.
Nevertheless, we remain attached to the flexibility of a case by case exclusion on condition of supervisory authorities’ approval as expressed later on in the Discussion paper EBA/DP/2017/04, article 186.
Indeed, NII is an accrual concept / carry effect related to cash flows and is unrelated to market risk.
However there is no shared definition of NII and the significance of NII may vary from one bank to another.
Hence it is our view that option 2 is indeed the best one as it provides flexibility and maintains consistency between RTPL and HPL.
We think that it is worth clarifying that the ES model mentioned in paragraphs 222 and 223 of the Discussion paper EBA/DP/2017/04 is the ES current period full set.
Paragraph 223 of the Discussion paper EBA/DP/2017/04, also raised the question of the modelling choice for NMRF calculation. As explained in the next section, we would have no alternative choice but to use sensitivity approach to calculate NMRF charge for a vast majority of NMRF. Given the conservative nature of the NMRF, we believe that in the specific case of NMRF, PLAT objective should be to test only the risk factor coverage and not the pricing function accuracy.
Definition of the observation period: This question has a very significant impact. Option (c) where the observation is the stress period (we voluntarily exclude the word “include”) is preferable: it has the merit of simplicity and also the one of defining the length of the stress period, which is a driver of the output.
In the event of scarce data, a fall-back approach could be Option (b).
Option (d) is the worst of all proposed options and would lead to extreme difficulties and operational complexities.
Types of data acceptable for the observations:
o Option (b) and (c) are preferred (i.e. the data should correspond to the same data type as for MRF actors) and gauge risk factor data could be used to proxy NMRF with lesser observations;
o Option (a) appears too restrictive and would imply disregarding observation that are representative of the market, but do not check all the conditions for “Verifiable prices” (e.g. broker quotes were commitment can’t be proved). The link between the price of transactions and risk factors is not direct (i.e. it is complex in some cases);
o Note that external data, like consensus, would also be a relevant source of data in case no internal data is available.
Definition of the Liquidity Horizons LH(j) for a NMRF: For the sake of simplicity, Option (b) is preferable.
Calibration of the confidence level CLsigma parameter: A high confidence level such as 90% seems appropriate.
Calibration of parameter CES equiv: Though more studies based on real data distributions are necessary to answer this point, Option (c) seems to be the best one.
Calibration of kj: Option (e) where the parameter is set to 1 is highly preferable to limit the operational burden. Any other option is, in our view, not feasible in practice due to computational complexities.
If, however, the EBA was to insist to have calculated, the best alternative would be Option (d) complemented with Option (a) for those risk factors that would require a calculation.
Calibration of parameter CLsigma, CES equiv, ????????t, calibration to achieve the target calibration ‘at least as high as an expected shortfall’:
o Generally speaking, the calibration method should not be too prescriptive and options where institutions have the flexibility to propose their calibration method and document it, are preferred;
o Regarding the calibration of kappa, option (e) is preferable and simplifies the implementation.
Those discussions could result in significant evolutions of the SES framework, in particular calculating the SES charge at a curve or surface level (as opposed to the elemental level represented by each specific point of the curve or surface).
Consequently the presented methodology applies only to elemental / scalar risk factors, which may be not the target framework.
Additionally, being at the elemental risk factor level has the following drawbacks:
There are several thousands of elemental risk factors:
o Operational burden to calibrate for each risk factor (in this respect, we intended to make homogeneous groups of risk factors and to calibrate a shock applicable for the whole group). The concept of “gauge data” in paragraph 254 (Discussion paper EBA/DP/2017/04) requires further description.
o CPU burden to compute SES for each elemental RF (seems unfeasible).
This is aggravated by the idea to compute the NMRF impact on the range of extreme shocks (cf. paragraphs 238 & 247 of the Discussion paper EBA/DP/2017/04), and not only at the extreme shocks.
And above that, shocking one elemental risk factor and letting the other unchanged generally makes generally no sense and besides creates numerical issues / arbitrage situations (term structures, volatility surfaces…):
o Many of non modellable risk factors will be bases between the long term (non modellable) part of a curve and the short or medium term (modellable) part, and also the (non modellable) aisles of smiles in volatility surfaces vs the (modellable) center of the surface
o In this context, the scenario “this particular point of the curve/surface experiments a large shock while the rest of the curve/surface is unchanged” has little economic meaning: this theoretical deformation induces arbitrages / inconsistencies and would not occur in the real world.
o A full valuation P&L computation will consequently exhibit numerical errors, and (if not #ERR) the result integrates a desarbitrage algorithm: as such it is not interpretable as a P&L resulting from the shock, missing the objective.
[please see the graph in the attached document]
Consequently, The concept of [stressing single risk factors and computing the induced P&L with full valuation] need to be reviewed.
Note that another topic in discussion with the Basel Market Risk Group (MRG) is the aggregation across risk factors using a formula that takes into account the low correlation between risk factors (i.e. as opposed to a simple sum). This point is quite independent of the calculation itself, but should also require a RTS in the future.
Also, the EBA has taken the approach to directly calculate the P&L impact of an instantaneous extreme shock observed over the risk factor liquidity horizon (ranging from 10 to 120 business days). Those shocks may be so extremes (while other risk factors are kept unchanged) that they will often fail pricers. Besides, we believe that the intention was to capitalise non-modellable risk factors at least as conservatively as modellable risk factors would be (leaving aside the fact that netting and diversification benefits are forfeited). Hence we consider that the same approach should be taken for NMRF as in the stressed expected shortfall (cf. draft Regulation amending Regulation 575/2013, article 325bd), i.e. use 10 day stressed returns and scale the P&L impacts with a quadratic formula. This approach should be complemented with a max loss provision whereby if the resulting capital charge attached to an instrument NMRF exceeds the instrument maximum drop of value, the capital charge shall be cap by it.
Nevertheless computational burden is a major concern for both approaches.
Point a: Frequency of review of the extreme scenarios of future shocks. A full monthly review seems not necessary. We remind that stressed scenarios are, by definition, stressed and instable by nature.
Point f: By construction, there is very little chance that the NMRF framework is less severe than the ES.
Equity correlations;
Equity volatility smile, as well as some idiosyncratic equity parameters like dividends and repos;
Interest rates correlations, as well as long term part of the curves (notably emerging), and smile aisles.
Also, the answer would ultimately depend on each firm’s choice to proxy or not the NMRF as per footnote 40 of the Basel standard BCBS d352 (i.e. residual basis has least likely the same distribution than the outright factor itself).
In the framework proposed by the EBA, we identify mainly two areas where a statistical assumption on the NMRF distribution would be helpful:
Calibration of CS, when a direct estimation on empirical data is not possible or deemed not sufficiently robust;
Computation of the kappa adjustment, when the loss profile is locally convex around CS.
Regarding the first area, the EBA provides in annex 3 several values for theoretical SGT distributions. Such distributions requires the calibration of five parameters, one of which (lambda) drives the distribution asymmetry. Figure 6 in the Annex gives evidence that the asymmetry has a direct impact on the estimation of the CES parameter (scaling factor from standard deviation to ES) with up to 33% increases between positive asymmetry (lambda=0.30) and no asymmetry (lambda=0).
Indeed, an asymmetrical distribution usually implies that one tail is fatter than the other. For illustration purpose, we have plotted in the following graph a skewed normal distribution function (‘Snorm’ in red). Whereas the kurtosis is equal to 3 for the whole distribution, the kurtosis restricted to the left tail only and right tail only equal 3.8 and 2.2 respectively. Similarly, the ratio ES>97.5 / σ is above 5 while ES<2.5 / (-σ) is below 1. In other words, the fatter one tail the thinner the other.
[please see the graph in the attached document]
Although we recognize asymmetry exists at the RF level, it would be quite a punitive and unrealistic assumption that, for each NMRF, the ES of the loss function would systematically coincide with the fatter tail. Most likely, the SES framework will cover a large number of risk factors and a sound base scenario should remain agnostic to which tail realizes the ES.
Owing to the above considerations, we strongly recommend the EBA to dismiss any statistical asymmetry at the RF level and just consider symmetrical distributions in its effort to capture potential “fat-tailedness”. This should lead to review and actually lower the proposed range on question 6) for CS value.
As detailed in previous answers, we reiterate our preference for option E, kappa=1.
Should this option not be retained, option D should be considered (kappa adjustment only on material risk factors, and when not covered by an add-on), and in this case we strongly advise to consider an operational simplification mixing options B and C (quadratic non linear losses with some Student distribution), that can additionally be well proxied by a closed from / second order Taylor Young development:
SES=f(CS)+Γ/2 CS²(ϕ-1) with ϕ=(ES(x^2))/(ES(x)²)
With f(x) the loss corresponding to the return x, ∆ and Γ the first and second order derivatives on CS.
Note that although this type of simplified approach could be considered, it would still mean additional model risk (distributional assumptions for ϕ) and complexity (estimation of the local convexity) compared to our proposal to set kappa at 1.
The above formula also shows that the non-linearity adjustment comes into play regardless of the sign of Γ. In particular, when the argmax FS is not an endpoint of CSSRFR, Γ is negative and the adjustment should reflect how much lower the ES of losses is compared to the max loss f(FS). Since there is in general no particular reason why Γ should be positive or negative, kappa=1 (ie no adjustment) is certainly the fairest assumption.
A limited number of RFs;
Outrights RFs but not for basis issued from footnote 40.
Even if, Option 2 is the preferred method, the current size of prescribed weights seems overly penalizing.
The recalibrated version of the current standardised approach should apply both to banks with small trading books and to subsidiaries of larger banks, including subsidiaries of G-SIB and D-SIB, as long as they meet the criteria on a standalone basis for the calculation of market risk own funds requirements. It should be clarified that RWAs calculated with the simplified approach at the entity level may be considered at the consolidated level of a group.
The recalibrated version of the current standardised approach should be simple enough that it will not overburden banks or bank subsidiaries with limited trading activities. Further simplification of the recalibrated version of the current standardised approach should be sought while a sufficient level of risk sensitivity retained.
The proposed recalibrated version of the current standardised approach shall be benchmarked with the full standardised approach (SbM) in order to improve its design and calibration. It shall not be the case that the recalibrated version of the current standardised approach lead to fewer own funds requirements than the standardised approach (SbM). Conversely, it shall not lead to overly inflated own funds requirements: there should still be some risk sensitiveness and commensurateness.
French banks favour a new simplified standardised approach (R-SbM) instead of a scalar factor applying to risk classes.
As currently drafted, such RTS only applies to institutions with no IRBA approval to estimate internal PD/LGD. We consider such RTS should also be applicable to IRBA-validated institutions for those issuers in DRC scope which are not covered by internal credit methodologies.
Technical standards on assessment methodology & Default Risk Charge: In paragraph 289 of the Discussion Paper, the EBA recommends that DRC guidelines (cf. draft Regulation amending Regulation 575/2013, article 325bn(2)) are directly addressed in the revised RTS on assessment methodology. We would like to go one step further and suggest that the above-mentioned DRC RTS is tackled together with the guidelines in the revised RTS on assessment methodology.
Indeed, the revised RTS on assessment methodology should then clarify that flexibility offered to non-IRBA institutions to use alternative approaches (external ratings simplified approaches) should also apply to IRBA-validated institutions for issuers with no internal PD/LGD. In case such flexibility is not granted, IRBA-validated institutions will potentially have to rate internally thousands of issuers, for which we do not have commercial relationship and hence do not rate internally (ex. equity indices constituents), meaning in most cases collecting comprehensive information (e.g. various balance sheet ratios) on each and single issuer which is unmanageable.
Question 5. What are your views about the qualitative approach used as a starting point under step 2?
The qualitative approach of step 2 is required to avoid computing too many unwarranted sensitivities under this approach.The qualitative assessment is in our view part of step 1. It will determine the risk category, or on rare occasion the risk categories, to which a product should be mapped.
Moreover, as stated in our response to Question 4, cases where a qualitative assessment would not suffice, and uncertainty remains as to which categories an instrument should belong to, should be the exception rather than the rule.
Question 6. Which would be the most appropriate option for the quantitative approach? Would you recommend another option?
Even if the banking industry considers that most instruments risk category may be determined in step 1, in the unlikely event that going to step 2 is required, we support option n°3 that authorize using internal and market sensitivities.Question 7. What values would be reasonable for the threshold(s) (X, Y, and their equivalents for Options 3 and 4) that determine the number of material risk drivers? Please provide rationales for proposed levels.
First of all, we would remind that we support maintaining the slicing of products approach (see paragraph 79). Indeed on many instances being sensitive to risk drivers belonging to several risk categories does not mean being more risky.If the slicing approach was to be ruled out, an alternative to the simple addition of the full notional amount exist. This alternative approach increases the overall CCR exposure amount but in a more risk sensitive, commmensurate and less punitive way. It is depicted below based on option 2 of paragraph 72:
For the first risk category, the effective notional ‘N’ of the transaction is used to calculate the SA CCR add on;
For the second risk category, the add on shall be calculated using an adjusted notional amount N*=N·a2/a1.
Question 8. Do you have any views on the appropriateness of devising a fallback approach? Can you identify any cases where reverting to the fallback approach is necessary?
Whenever derivative instruments are out of FRTB SA scope (for instance when belonging to the banking book) and not captured under step 1, recourses to internal policies shall be allowed to determine the risk categories provided that banks internal knowledge will provide fairest and more accurate outcomes than the conservative step 3.Question 9. Do you have any views on the appropriateness of a cap on the number of risk categories to which a single derivative transaction can be allocated? If yes, what value would you recommend for that cap (three or four)?
Derivative transactions with more than two risk categories are an extremely rare situation. Indeed, even for the sub-set of instruments that would have otherwise gone in step 2, a large majority of them are expected to be mapped only to one or two risk categories only.Question 10. Do you have any further comment or consideration on the mandate under discussion?
No comment.Question 11. Do you have any views on the most appropriate approach to compute supervisory delta in a negative interest rates environment? Please elaborate.
We ask for a log-normal approach to compute supervisory delta in a negative interest rates environment. The metric should remain simple and easy to implement for credit institutions.The best solution would be to use the FRTB SBA sensitivities. If not possible, the formula should be slightly amended as follows:
When assuming that (R+) is log-normal (instead of R in the initial formula) we should introduce’ defined as follows:
We consider that (dR/R+) = ’ dW instead of (dR/R) = dW, we can then deduce that * R* dW = ’ * (R+) * dW, which means that in the formula should be replaced with σ^'= σ* R/(R+ λ)
[please see the formula in the attached document]
Question 12. Which one of the two options do you think is more appropriate from an EU perspective (i.e. maximum harmonisation)? Are you aware of any issue these two options could raise?
The more appropriate option is the second option presented in paragraph 91 (i.e. “Banks could be required via EBA RTS to reflect the market convention for the λ parameter […] for the relevant jurisdiction.”).The EBA should define a flexible option.
For the sake of simplicity and stability, we ask for :
1) Use of a market convention if any;
2) In the absence of a market convention, use of the value of the λ internally used for a currency or, for example, if a normal model is used internally, a value of λ per currency agreed upon by the Risk department;
3) For the rare instances where λ is insufficiently large for a trade, use of a higher λ for that trade, for instance such that : Min{P+λ, K+λ}=1%.
Moreover, we would like to take the opportunity of this consultation to remind that in our view the fall back approach for option supervisory delta should be +/-0.5 rather than +/-1 as in the European Commission draft Regulation amending Regulation 575/2013 (cf. article 279a(1)(c)). A supervisory delta of +/-0.5 was the approach taken in the draft Basel standard (BCBS 254, paragraph 48] and justified as follow (cf. BCBS 254, footnote 10):
“The Basel Committee chose to set delta to 0.5 for non-linear instruments because (i) this is the mid-point of the spectrum of all possible deltas from 0 to 1 that strikes a good balance between trades being “free” (delta is 0) and overly effective hedges (delta is 1); (ii) this is the only choice of delta that respects the put-call parity (a combination of a bought call and a sold put of the same strike is equivalent to a forward). In this context, sold call (put) options are treated as bought put (call) options. Although sold options do not present counterparty credit risk on their own, supervisory delta adjustments are included in a hedging set because they affect the mark-to-market value of the entire hedging set.”
Question 13. Do you agree that the definition of a long position in the primary risk driver and a short position in the primary risk driver in Article 279a(2) of the CRR2 proposal is sufficiently clear for banks to determine whether they hold a long or a short position?
Yes we agree.Question 14. Do you agree that changes in instruments’ circumstances that imply a shift between the presumptive lists should be accepted as ‘exceptional circumstances’? Please provide examples.
The option to move from the banking book to the trading book should be preserved, controlled by an implicit approval process without requiring the full reclassification approval from the supervisory authority and associated capital surcharge. For example: Could we advise the supervisory authority of the move related to a change in instruments' circumstance and provide associated documentation. In the absence of further requests we should assume the change approved. Examples in addition to those noted in Paragraph 102 include:
o Changes in GAAP requires instruments that was booked under amortized cost to be fair valued or vice versa ;
o Changes in the intent and ability to trade (due to liquidity and other market condition).
We also propose that regular buy and sell activity between a trading desk and another distinct banking book unit of the bank which acts as a client should be exempted from the rule on movement between regulatory books, provided transactions are conducted on an arm’s length basis.
Question 15. Do you agree that CTP positions that become illiquid must remain in the TB?
We support that correlation trading portfolio (CTP) positions that become illiquid remain in the trading book. Liquidity should not be a discriminating factor since trading intent of CTP products remain. In fact the Basel standard does not seem to allow that it could be otherwise. More generally, and as mentioned in paragraph 27 of the Basel standard (BCBS d352 ), illiquidity should not alone be a reason for re-classification, for any instrument.Question 16. Please provide examples of cases where exceptional circumstances might warrant the approval of reclassification.
Following circumstances may warrant the approval of reclassification: Dramatic shift in the liquidity conditions of a large portion of financial instruments that may lead to re-consider the management intent on such instruments;
A modification in the accounting standards that implies the need to change the accounting valuation for certain instruments (implementation of IFRS 9 being a recent illustration);
A change in the business model of some activity affecting financial assets.
We propose that regulators allow banks discretion to re-assign instruments between regulatory books due to exceptional circumstances beyond banks’ control through an implicit approval process (“notification approach”) rather than having to seek approval from regulators for each re-assignment (“authorization approach”).
Question 17. Do institutions have any particular issue in identifying non-trading book FX and commodity positions subject to market risk? What kinds of transactions do those positions correspond to and how material are they with respect to current RWAs for market risks?
Non-trading book positions subject to FX risk have been covered by the EBA discussion paper (EBA/DP/2017/01) on the treatment of structural FX under Article 352(2) of the Regulation (EU) 575/2017.We are looking for the conclusions of the EBA discussion paper on the treatment of structural FX positions.
Question 18. What issues would institutions face to value those positions in order to calculate the own funds requirement for market risks using the FRTB standards? Currently, do you revalue all components for the purposes of computing the own funds requirement for market risks? If not, which ones? Currently, how frequently are those positions valued?
Currently, the computation is based on the accounting positions which are usually known/validated on a quarterly (structural position) basis, and sometime more often.Question 19. For the non-trading book positions subject to the market risk charge that are not accounted for at fair value (or in the case of FX, are non-monetary), do stakeholders have the capacity to mark these positions to market and how frequently can this be done? Do stakeholders have the capacity to “mark to market” the FX component of the non-monetary item subject to FX risk on a frequent basis (for example daily)?
The FX component of the non-trading book items is in accounts. These items are valued, on a monthly or quarterly basis.Question 20. Does IFRS 13, i.e. Fair Value Measurement, have an impact on the frequency of non-trading book revaluations? If yes, please explain how.
We ask for a deletion of any direct link between accounting purposes and regulatory purposes. For operational concerns, accounting data cannot be revaluated on a daily frequency.Question 21. Are there other factors (for example impairments or write-downs) that can affect the valuation of non-trading book FX positions?
No comment.Question 22. Do stakeholders have a view on what minimum number of notional trading desks should be allowed? What would be the negative consequences of applying some restrictions to the number of notional trading desks allowed (for example only one notional desk for FX positions and only one for commodities)?
The Basel standard (BCBS d352, article 46) requires that banking book FX and commodity positions are capitalised in the market risk framework as part of notional trading desks (plural). There is no condition on the number of such notional trading desks.There may be a single or multiple such notional trading desk for each of FX and commodity positions.
Question 23. Do you consider that trading book positions should not be included in notional trading desks? Would you agree that, for trading desks that include trading and non-trading book instruments, all the trading desk requirements should apply? Do you consider that for notional trading desks all the trading desk requirements should apply? If this is not the case, which qualitative requirements of Article 104b(2) of the CRR2 proposal could not practically apply to notional trading desks?
A notional trading desk is by nature notional (i.e. theoretical). It is just there to group some exposures together for own funds requirements calculation. Hence the strict requirements of the draft Regulation amending Regulation 575/2013 (Article 104b(2)) cannot apply.It is our view that the banking book notional trading desks shall generally be kept separate from trading book desks, if only because an actual trading desk business strategy and risk management structure (cf. BCBS d352, Appendix A) is unlikely to apply to the banking book exposures.
Nevertheless, risks transfers from notional trading desks to trading desks should be taken into account so that hedging benefits could be recognized.
Hence our view can be summarised as follow:
Notional desks for commodities and FX exposures are only for the purpose of own funds requirements calculation.
Notional trading desks should include only non-trading book positions
The trading desk qualitative requirements shouldn’t apply to the notional desks. In particular it seems inappropriate to consider that non-trading book activities should have the same constraint as trading book activities relating to profitability, business plan, etc.
Question 24. Do you see a reason why backtesting requirements should not apply to notional trading desks?
It does not seem straightforward that a notional trading desk be capitalised in the internal model approach since the banking book FX positions may not be re-valued on a daily basis. Hence back-testing is irrelevant.Question 25. Do you see a reason why P and L attribution requirements should not apply to notional trading desks?
It does not seem straightforward that a notional trading desk be capitalised in the internal model approach since the banking book FX positions may not be re-valued on a daily basis. Hence P&L attribution is irrelevant.Question 26. Do you agree with the proposed general definitions of instruments referencing an exotic underlying and instruments bearing other residual risks? Do you think that these definitions are clear? If not, how would you specify what is an ‘exotic underlying’ and what are ‘instruments that reference exotic underlyings’? Please provide your views, including rationale and examples.
The proposed definitions of exotic underlying is clear when complemented with both the Basel standard (cf. BCBS n°d352) examples listed in paragraph 58 and with the exceptions listed in paragraph 58(h).Nevertheless we would like to point that in our view the volatility and variance swaps should not be subject to an exotic risk add on (cf. BCBS d352 footnote 14) since it is possible to capture their volatility risk within the SBM capital charge.
Indeed, though not an optional product and hence theoretically not subject to a Vega risk charge, volatility and variance swaps are risk managed with sensitivities to implied volatility.
Hence it would make sense to have a caveat for those non-optional products and capitalise them within the SBM Vega risk charge rather than in the risk insensitive RRAO framework. It would be a similar treatment to the one of CMS (cf. BCBS FAQ d395).
Payoffs deemed exotic by the regulatory text are usually managed in a prudent manner by banks: some of their features are smoothed so that they can be managed as a static set of vanilla options: US or European exercise. If a bank can demonstrate and document that it systematically books and manages such payoffs as portfolios of vanilla options, theses payoffs should benefit from an exclusion of the scope of RRAO charge.
Examples:
European Exercice Digital payoffs prudently managed as a call (resp. put) spread of vanilla european exercise options which are not in the scope of the exotic payoff RRAO
Bermudean / Corridor options prudently managed without correlation
Question 27. Do you agree with complementing, for the sake of clarity, those definitions with a non-exhaustive list of instruments bearing other residual risk? Similarly, do you agree with retaining the possibility of excluding some instruments from the RRAO?
The banking industry would rather have a principle based approach than an exhaustive or even partial list of instruments subject to residual risk. The principle based approach should account for the magnitude of residual risk within the instrument which may depend on several contractual features as explained in our response to question 28.Also, we would like to clarify that in our view the exclusion of the draft Regulation amending Regulation 575/2013, article 325v(4)(c) also apply to back-to-back exactly matching transactions with premium payments spread.
Question 28. More specifically, do you consider that there are particular instruments (or underlyings) which, while meeting the definitions above (in line with point (d) of paragraph 58 of the FRTB), should be excluded from the RRAO? Alternatively, on the contrary, do you consider that there are instruments (or underlyings) that are not captured by the definitions above and that should be subject to the RRAO? Please provide your views, including rationale and examples.
The assessment whether an instrument shall be liable to a residual or exotic risk add on depends not only on the instrument meeting the definition of exotic or residual risk (cf. BCBS d352, article 58(e)&(g)) ,but as well on the significance of this residual or exotic risk. The latter is apparent by the exclusion of some products from an exotic risk add on charge (cf. BCBS d352, article 58(h)) and the EBA discussion on which instruments bearing behavioural risk shall be subject to a residual risk add on (cf. Discussion paper EBA/DP/2017/04, article 142(k)). In this context we would suggest that :- The notional be capped by the maximum loss if this last quantity is bounded,
- Appropriated weights be defined for leveraged products,
- The weights increase with the time to maturity, when the RRAO is generated by correlation risk, explained by materiality considerations.
This is illustrated by the following:
Asian options: Asian options should not be considered as bearing “gap risk”, or any other residual risk. Indeed they have a very smooth behavior (smoother than plain vanilla options), and do not present additional risk compared to those caught by the SBM. In this respect, even if they cannot be perfectly replicated as a finite linear combination of vanilla options, they should not be subject to the RRAO. This is particularly penalizing for the commodity business.
Vanilla options on commodities differential: Another case to consider are vanilla options on commodities differential (for example: options on the spread between two oil qualities), it must be clear that they can be seen as options on one underlying, and that they are excluded from the RRAO.
CMS options: the RRAO measure is over penalizing, and if no appropriate weight is defined they should be excluded. This consideration may be generalized to all leveraged products.
Binary / digital options: there is a problem with the RRAO computation, and if it is not solved that would appeal for exclusion. The problem is that with respect to the RRAO, the appropriate notional of a (vanilla) barrier option is the “size” of the digit, not the full notional of the option. This would make homogeneous capital charges for a pure digit, and for the digit embedded in a barrier option. Hence, for “vanilla digits” (meaning: payoffs having no other exotic features that the digit), we propose to retain the size of the digits and not the notional of the options:
o Pure digit paying €100 if Stock A > Strike: notional = €100.
o Barrier call option, strike K, barrier K’>K, for N stocks A: notional = Nx(K’-K).
o Call spread strategy with strikes K and K’, for N stocks A: notional = Nx(K’-K).
Contractual features: In some instances contractual features other than the product type may be considered to determine a product eligibility to the residual risk add-on since those contractual feature may have a strong influence on the level of residual risk. For instance, the remaining maturity may be considered for some products to decide on their eligibility to the residual risk add-on. It is the case of spread options (options on curve slope) which may be liable for a residual risk add-on only if of long maturity.
Bermudan options: By construction, a Bermudan option value is framed by the corresponding swaptions (inferior bound) and cap/floor (superior bound). Both swaption and cap/floor are vanilla products which are entirely covered in SBM.
Nevertheless the correlation risk is very limited on vanilla Bermudan options with short callable periods and should generate no or fewer RRAO. A maturity threshold could be 10 years, because indeed the market liquidity is very important for bermudan options with maturity less than 10 years in line with small correlation risk on those bermudas.
This is illustrated hereunder where sensitivities in basis points by maturity are given for a 1% shock on the Bermuda market correlation, and where it is shown also how the RRAO should increase as a time to maturity function up to the 10 years threshold.
[Please see the graph in the attached document]
Recommendation for Bermudan options:
RRAO notional multiplicator should linearly increases from 0.0% to a 0.1% cap as Tenor (callable period) increases from 0Y to 10Y.
Hedging strategy: The determination of the notional to be used might be unclear for certain instruments. Other residual risks may also depend on factors such as the hedging strategy (a pure back-to-back would not bear market risk), or the residual maturity of the instrument (where risk might be significantly different for short on long term positions), which is not taken into account in the RRAO add-on for other.
Example: spread options (options on curve slope) are not fully replicable by a portfolio of simple options as they are sensitive to the correlation between both handles of the slope c (Cega). They therefore are subject to additional capital (RRAO) in SA. Two issues can be identified for those derivatives:
o It is unclear how to measure the notional of the trade. (Is it the “face amount of the trade” or “the sum of all its leveraged legs”?). The later choice is subject to the way the deal is structured for an equivalent amount of Cega and unfortunately sensitive to the number of instruments used to replicate the transactions.
More precisely: according to the hedging strategy the RRAO capital charge would be very different:
0 for perfect B2B,
Par x 0,1% if unhedged,
2 x Par x 0,1% for a hedge with the same deal shifted by a couple of days,
(n+1) x Par x 0,1% if the hedge is structured by n trades with different maturities.
o For short to medium dated option, the 0,1% RRAO capital charge for Cega is far too high.
Example: a spread option 2Y/10Y de 500M€ maturity 4 months and a half, one can compute 500k€ of RRAO as of 31/12/2017, with a Cega sensitivity almost zero (<10€…). The same spread option with a 10Y maturity would have a 79K€ Cega making the 500k€ of RRAO more in line (a 10% Cega shock super conservative, 0.1% implies a 39% shock for a 3Y option…).
Result: for a CIB with a classical FI product mix, under 5Y RRAO for spread option can represent up to 3/4th of the total which looks abnormal.
Recommendations and proposal for Spread Options:
o Leverage should not be the base of the notional computation, only the par face value of the option should be taken into account;
o RRAO notional multiplicator should linearly increases from 0,0% to a 0,1% cap as expiries increases from 0Y to 10Y.
Question 29. Although the proposed list of options does not aim at being exhaustive, since there is a general definition, do you find that any important option type meeting the criteria in point (i) of point (e) of paragraph 58 of the FRTB is missing? Conversely, do you think that any of the options in the list does not meet general criteria?
No comment.Question 30. Do you think there are any instruments, not meeting the general definitions above, whose risk would however be poorly captured within the standardised approach and should therefore be included in the list of instruments subject to the RRAO?
No comment.Question 31. What are your views on the proposed treatment for behavioural risks? Do you have any proposal for a more objective/prescriptive approach to identifying instruments with behavioural risks?
The proposed treatment is quite clear. Nevertheless some examples could be useful.Question 32. What are your views on the role that the list in point (h) of paragraph 58 of the FRTB should play?
Point (h) is particularly crucial to avoid any ambiguities and clarifies the need of an assessment of an exotic factor materiality to decide on the instrument subjection to an exotic risk add on (see answer to question 28).It should be complemented by the fact that if an instrument qualifies for two residual risks (for example correlation and digital risk), it must be counted only once in the computation.
Question 33. Are there any cases in which instruments could meet the definitions of both ‘instrument referencing an exotic underlying’ and ‘instrument bearing other residual risks’?
We haven’t identified such instruments at the moment but we cannot exclude that some may exist in the future. With the “no double counting” clarification mentioned in the answer to the previous question, they will be weighted by the greater weight.Question 34. What is your view on the outlined approach? Please provide background and reasoning for your position.
No comment.Question 35. Do you have in mind risk factors for which additional guidance is needed? If yes, which ones?
No comment.Question 36. Do you have in mind any risk factor categories or subcategories to add to those listed in Table 2 of Article 325be of the CRR2 proposal?
At this stage we do not anticipate the need to define new categories or sub-categories, though we may leave this possibility opened. This said, often, to avoid broken hedges, there is a trend toward choosing the larger, or one of the largest, liquidity horizon within an asset class which may render the definition of sub-categories pointless.Also, we would like to take the opportunity of this consultation to reiterate our view on the draft Regulation amending Regulation 575/2013, article 325be(4): We consider that credit institutions should have the flexibility not to account for the maturity of the position when computing the effective liquidity horizon (LH) of risk factors of that position, provided that such option is applied consistently across all risk factors of all positions. Indeed applying the draft Regulation amending Regulation 575/2013, article 325be(4) as is (i.e. capping the liquidity horizon of each instrument to the instrument maturity) will be operationally extremely challenging, bring inconsistency with the constant risk assumption prevailing in the expected shortfall (ES) formula and result in broken hedges.
Question 37. Would you think that QAs could be sufficient to provide additional guidance (instead of RTS)?
We believe that Q&A would be sufficient to provide additional guidance while keeping enough flexibility.As an example, a Q&A could confirm that agencies and supras should be classified as sovereign.
Question 38. What is your view on the definition and level of the threshold used for assigning currencies to the most liquid category?
No comment.Question 39. If you agree with the threshold outlined, would you agree that the list of selected currencies should be updated on a triennial basis following the publication of the BIS OTC derivative statistics?
No comment.Question 40. If you do not agree with the threshold outlined, please provide reasoning for establishing another selection criterion.
No comment.Question 41. What is your view on the definition and level of the threshold used for currency pairs to be considered most liquid?
No comment.Question 42. If you agree with the threshold outlined, would you agree that the list of selected currencies should be updated on a triennial basis following the publication of the BIS OTC derivative statistics?
Updating too frequently the list of currency pairs may lead to some increased own funds requirements volatility. On the other hand not updating the list may result in a regulation out of sync with the prevailing market conditions.What matters to the banking industry most is consistency across various regulations. We would therefore suggest that the list of liquid currency pairs and liquid interest rate currencies are aligned with that of other regulations.
Question 43. If you do not agree with the threshold outlined, please provide reasoning for establishing other selection criteria.
No comment.Question 44. Do you consider that triangulation of currency pairs should be allowed? Is triangulation used in practice to hedge less liquid FX positions?
We strongly support the introduction of the triangulation of currency pairs as it reflects the regular risk management practice.Question 45. What is your view on the definition and level of the threshold for defining small and large capitalisations for equity price and volatility?
No comment.Question 46. Do you see any problems in using the ITS published by ESMA to specify the equities that can be considered as large capitalisations?
No, we do not see problem in using the ITS published by ESMA to specify the equities that can be considered as large capitalizations.Actually, as mentioned in our response to question 42 we support the alignment between regulations wherever possible and this is a good example of such desirable alignment.
Question 47. Do you agree with the list of criteria for systematic exclusions from hypothetical P and L?
The criteria for systematic exclusion from hypothetical are clearly defined and can be summed up as such: Valuation adjustments for which a separate regulatory capital treatment has been specified as part of the rule (e.g. CVA);
Valuation which are deducted from CET1 (e.g PVAs);
Valuation adjustments that are updated at a less than daily frequency in the measure of P&L.
This list of exclusions is generally fit for purpose. However, there may be specific daily value adjustments that are not market risk related: They should not be included in the Hypothetical P&L (cf. EBA Discussion paper EBA/DP/2017/04, article 177) and neither the Risk Theoretical P&L since RTPL should be in line with Hypothetical P&L (HPL) (cf. EBA Discussion paper EBA/DP/2017/04, article 226).
For instance some risk factors may be subject to a daily IPV (independent price verification) value adjustment. IPV is not a market risk adjustment and it cannot be captured in the risk model hence it should be excluded to both HPL and RTPL.
Hence we support the EBA view expressed in the Discussion paper EBA/DP/2017/04, article 186 stating that, on a case by case basis, daily adjustments may be excluded from HPL, and hence RTPL, conditional to a supervisory authorisation.
Question 48. Do you have numerous valuation adjustments not computed at desk levels? For those VAs, would it be possible to calculate them at desk level? If not, explain why.
A majority of value adjustments are made at entity level only to reflect diversification and netting benefits.Question 49. Do you agree with the criteria defined for the inclusion of a valuation adjustment in the hypothetical P and L? If not, please give arguments. Do you agree with the proposal to provide only criteria for inclusion in or exclusion from the hypothetical P and L, in order to allow some flexibility, or do you think that we should have non-exhaustive lists supplemented by criteria?
We agree with the criteria defined for the inclusion of a valuation in the Hypothetical P&L. Nevertheless we propose a small amendment to paragraph 186 as such:“186.Finally, only the valuation adjustments related to market risk that are updated daily and are not in the above list of systematic exclusions would be included in the hypothetical P&L (at the ‘top of the house’ only or also at desk level, depending on the level at which these VAs are computed) […].”.
Hence, most daily valuations that should not be included in the Hypothetical P&L (see example in our response to question 47) will be clearly identified and excluded.
Nevertheless, we remain attached to the flexibility of a case by case exclusion on condition of supervisory authorities’ approval as expressed later on in the Discussion paper EBA/DP/2017/04, article 186.
Question 50. Do you agree with developing additional guidance on specific valuation adjustments: related to market risk versus not related to market list, possible daily frequency update in the P and L versus not daily, ‘top of the house’ versus desk-level computation?
As the criteria proposed for exclusion and inclusion in the Hypothetical P&L are sufficient, we don’t think that additional guidance is necessary.Question 51. Did you have overshootings that are mainly caused by valuation adjustments included in the hypothetical P and L? If yes, which valuation adjustments were mainly causing overshootings? Did you identify types of desks which were more frequently affected by such overshootings? Are these desks likely to breach the backtesting thresholds because of these overshootings (how frequently do the overshootings occur)?
No comment.Question 52. Do you agree with the list of criteria for systematic exclusions from the actual P and L?
No comment.Question 53. Do you agree with the criteria defined for the inclusion of a valuation adjustment in the actual P and L? If not, please provide arguments.
No comment.Question 54. Did you have overshootings that are mainly caused by valuation adjustments included in the actual P and L? If yes, which valuation adjustments were mainly causing overshootings? Did you identify types of desks which were more frequently impacted by such overshootings? Are these desks likely to breach the backtesting thresholds because of these overshootings (how frequently do the overshootings occur)?
No comment.Question 55. According to you, is the net interest income part of the time effect?
Yes the net interest income is part of the time effect.Question 56. Do you agree with the proposed definition for net interest income? If not, what would be your proposal?
As we favour proposal 2 and the use of a generic term for the “P&L due to passage of time”, we believe that there is no need to define net interest income.Question 57. Would you like further indications of the elements to take into account in the time effect? Which elements would you include in the time effect?
In line with our response to question 56, we support the use of a generic definition of P&L due to passage of time with no need to define the time effect and to mention it in the regulation.Question 58. Regarding the different proposals, do you agree with EBA that Proposal 2 would achieve the best outcome? If not, what would be your suggestion?
In principle, option 3 would make sense, keeping time effect after removal of the Net Interest Income in the P&L.Indeed, NII is an accrual concept / carry effect related to cash flows and is unrelated to market risk.
However there is no shared definition of NII and the significance of NII may vary from one bank to another.
Hence it is our view that option 2 is indeed the best one as it provides flexibility and maintains consistency between RTPL and HPL.
Question 59. Do you agree with the principle of including in or excluding from the risk-theoretical P and L the same valuation adjustments as for the hypothetical P and L?
We agree that consistency shall be preserved between HPL and RTPL and this is true as well of value adjustments.We think that it is worth clarifying that the ES model mentioned in paragraphs 222 and 223 of the Discussion paper EBA/DP/2017/04 is the ES current period full set.
Paragraph 223 of the Discussion paper EBA/DP/2017/04, also raised the question of the modelling choice for NMRF calculation. As explained in the next section, we would have no alternative choice but to use sensitivity approach to calculate NMRF charge for a vast majority of NMRF. Given the conservative nature of the NMRF, we believe that in the specific case of NMRF, PLAT objective should be to test only the risk factor coverage and not the pricing function accuracy.
Question 60. What are your preferred options for points 1-8 above? How would you justify these preferences?
Please find hereafter our comments on points 1 to 8: Definition of the observation period: This question has a very significant impact. Option (c) where the observation is the stress period (we voluntarily exclude the word “include”) is preferable: it has the merit of simplicity and also the one of defining the length of the stress period, which is a driver of the output.
In the event of scarce data, a fall-back approach could be Option (b).
Option (d) is the worst of all proposed options and would lead to extreme difficulties and operational complexities.
Types of data acceptable for the observations:
o Option (b) and (c) are preferred (i.e. the data should correspond to the same data type as for MRF actors) and gauge risk factor data could be used to proxy NMRF with lesser observations;
o Option (a) appears too restrictive and would imply disregarding observation that are representative of the market, but do not check all the conditions for “Verifiable prices” (e.g. broker quotes were commitment can’t be proved). The link between the price of transactions and risk factors is not direct (i.e. it is complex in some cases);
o Note that external data, like consensus, would also be a relevant source of data in case no internal data is available.
Definition of the Liquidity Horizons LH(j) for a NMRF: For the sake of simplicity, Option (b) is preferable.
Calibration of the confidence level CLsigma parameter: A high confidence level such as 90% seems appropriate.
Calibration of parameter CES equiv: Though more studies based on real data distributions are necessary to answer this point, Option (c) seems to be the best one.
Calibration of kj: Option (e) where the parameter is set to 1 is highly preferable to limit the operational burden. Any other option is, in our view, not feasible in practice due to computational complexities.
If, however, the EBA was to insist to have calculated, the best alternative would be Option (d) complemented with Option (a) for those risk factors that would require a calculation.
Calibration of parameter CLsigma, CES equiv, ????????t, calibration to achieve the target calibration ‘at least as high as an expected shortfall’:
o Generally speaking, the calibration method should not be too prescriptive and options where institutions have the flexibility to propose their calibration method and document it, are preferred;
o Regarding the calibration of kappa, option (e) is preferable and simplifies the implementation.
Question 61. Do you have any observations or concerns about the overall methodology proposed for point (a) of the mandate?
Regrettably, the EBA strictly sticks to the draft mandate which was established before some discussions have taken place between the Basel Market Risk Group (MRG) and the industry (i.e. ISDA working group).Those discussions could result in significant evolutions of the SES framework, in particular calculating the SES charge at a curve or surface level (as opposed to the elemental level represented by each specific point of the curve or surface).
Consequently the presented methodology applies only to elemental / scalar risk factors, which may be not the target framework.
Additionally, being at the elemental risk factor level has the following drawbacks:
There are several thousands of elemental risk factors:
o Operational burden to calibrate for each risk factor (in this respect, we intended to make homogeneous groups of risk factors and to calibrate a shock applicable for the whole group). The concept of “gauge data” in paragraph 254 (Discussion paper EBA/DP/2017/04) requires further description.
o CPU burden to compute SES for each elemental RF (seems unfeasible).
This is aggravated by the idea to compute the NMRF impact on the range of extreme shocks (cf. paragraphs 238 & 247 of the Discussion paper EBA/DP/2017/04), and not only at the extreme shocks.
And above that, shocking one elemental risk factor and letting the other unchanged generally makes generally no sense and besides creates numerical issues / arbitrage situations (term structures, volatility surfaces…):
o Many of non modellable risk factors will be bases between the long term (non modellable) part of a curve and the short or medium term (modellable) part, and also the (non modellable) aisles of smiles in volatility surfaces vs the (modellable) center of the surface
o In this context, the scenario “this particular point of the curve/surface experiments a large shock while the rest of the curve/surface is unchanged” has little economic meaning: this theoretical deformation induces arbitrages / inconsistencies and would not occur in the real world.
o A full valuation P&L computation will consequently exhibit numerical errors, and (if not #ERR) the result integrates a desarbitrage algorithm: as such it is not interpretable as a P&L resulting from the shock, missing the objective.
[please see the graph in the attached document]
Consequently, The concept of [stressing single risk factors and computing the induced P&L with full valuation] need to be reviewed.
Note that another topic in discussion with the Basel Market Risk Group (MRG) is the aggregation across risk factors using a formula that takes into account the low correlation between risk factors (i.e. as opposed to a simple sum). This point is quite independent of the calculation itself, but should also require a RTS in the future.
Also, the EBA has taken the approach to directly calculate the P&L impact of an instantaneous extreme shock observed over the risk factor liquidity horizon (ranging from 10 to 120 business days). Those shocks may be so extremes (while other risk factors are kept unchanged) that they will often fail pricers. Besides, we believe that the intention was to capitalise non-modellable risk factors at least as conservatively as modellable risk factors would be (leaving aside the fact that netting and diversification benefits are forfeited). Hence we consider that the same approach should be taken for NMRF as in the stressed expected shortfall (cf. draft Regulation amending Regulation 575/2013, article 325bd), i.e. use 10 day stressed returns and scale the P&L impacts with a quadratic formula. This approach should be complemented with a max loss provision whereby if the resulting capital charge attached to an instrument NMRF exceeds the instrument maximum drop of value, the capital charge shall be cap by it.
Question 62. Do you have an alternative proposal for the calculation of an extreme scenario of future shock or stress scenario risk measure?
No comment.Question 63. Do you have any comment on the ‘risk factor based approach’ versus the ‘direct loss based approach’? Is computational effort a concern?
The Risk Factor based approach is largely preferable because the outputs can be analyzed/challenged by someone having knowledge of the market behaviour.Nevertheless computational burden is a major concern for both approaches.
Question 64. Is there a case for allowing institutions to calculate a standalone expected shortfall directly?
No comment.Question 65. Do you have any views on points (a)-(g) above?
Please find below our main comments : Point a: Frequency of review of the extreme scenarios of future shocks. A full monthly review seems not necessary. We remind that stressed scenarios are, by definition, stressed and instable by nature.
Point f: By construction, there is very little chance that the NMRF framework is less severe than the ES.
Question 66. What are the most relevant NMRFs for your institution in broad terms?
Most relevant factors are: Equity correlations;
Equity volatility smile, as well as some idiosyncratic equity parameters like dividends and repos;
Interest rates correlations, as well as long term part of the curves (notably emerging), and smile aisles.
Question 67. What are the most relevant statistical distributions for NMRFs?
Provide a precise answer to this question is difficult given the current uncertainty as per the final eligibility criteria at Basel level and given the large variety of risk factors that may become non-modellable (i.e. what is fit for certain factors may be quite approximate for others).Also, the answer would ultimately depend on each firm’s choice to proxy or not the NMRF as per footnote 40 of the Basel standard BCBS d352 (i.e. residual basis has least likely the same distribution than the outright factor itself).
In the framework proposed by the EBA, we identify mainly two areas where a statistical assumption on the NMRF distribution would be helpful:
Calibration of CS, when a direct estimation on empirical data is not possible or deemed not sufficiently robust;
Computation of the kappa adjustment, when the loss profile is locally convex around CS.
Regarding the first area, the EBA provides in annex 3 several values for theoretical SGT distributions. Such distributions requires the calibration of five parameters, one of which (lambda) drives the distribution asymmetry. Figure 6 in the Annex gives evidence that the asymmetry has a direct impact on the estimation of the CES parameter (scaling factor from standard deviation to ES) with up to 33% increases between positive asymmetry (lambda=0.30) and no asymmetry (lambda=0).
Indeed, an asymmetrical distribution usually implies that one tail is fatter than the other. For illustration purpose, we have plotted in the following graph a skewed normal distribution function (‘Snorm’ in red). Whereas the kurtosis is equal to 3 for the whole distribution, the kurtosis restricted to the left tail only and right tail only equal 3.8 and 2.2 respectively. Similarly, the ratio ES>97.5 / σ is above 5 while ES<2.5 / (-σ) is below 1. In other words, the fatter one tail the thinner the other.
[please see the graph in the attached document]
Although we recognize asymmetry exists at the RF level, it would be quite a punitive and unrealistic assumption that, for each NMRF, the ES of the loss function would systematically coincide with the fatter tail. Most likely, the SES framework will cover a large number of risk factors and a sound base scenario should remain agnostic to which tail realizes the ES.
Owing to the above considerations, we strongly recommend the EBA to dismiss any statistical asymmetry at the RF level and just consider symmetrical distributions in its effort to capture potential “fat-tailedness”. This should lead to review and actually lower the proposed range on question 6) for CS value.
Question 68. What are the most relevant non-linear tail loss profiles that need to be considered?
This question is linked to the computation of the kappa adjustment.As detailed in previous answers, we reiterate our preference for option E, kappa=1.
Should this option not be retained, option D should be considered (kappa adjustment only on material risk factors, and when not covered by an add-on), and in this case we strongly advise to consider an operational simplification mixing options B and C (quadratic non linear losses with some Student distribution), that can additionally be well proxied by a closed from / second order Taylor Young development:
SES=f(CS)+Γ/2 CS²(ϕ-1) with ϕ=(ES(x^2))/(ES(x)²)
With f(x) the loss corresponding to the return x, ∆ and Γ the first and second order derivatives on CS.
Note that although this type of simplified approach could be considered, it would still mean additional model risk (distributional assumptions for ϕ) and complexity (estimation of the local convexity) compared to our proposal to set kappa at 1.
The above formula also shows that the non-linearity adjustment comes into play regardless of the sign of Γ. In particular, when the argmax FS is not an endpoint of CSSRFR, Γ is negative and the adjustment should reflect how much lower the ES of losses is compared to the max loss f(FS). Since there is in general no particular reason why Γ should be positive or negative, kappa=1 (ie no adjustment) is certainly the fairest assumption.
Question 69. What is the materiality of non-linear tail losses in practice?
No comment.Question 70. Do you deem Option 1 (the ‘maximum possible loss’) or Option 2 (the prescribed risk weights) more suitable as a fallback approach? What is the reason for your preference?
Option 2 (the prescribed risk weights) is more relevant but note that SBM shocks are only available for: A limited number of RFs;
Outrights RFs but not for basis issued from footnote 40.
Even if, Option 2 is the preferred method, the current size of prescribed weights seems overly penalizing.
Question 71. Do you deem the risk factor categories and respective shocks presented in the tables in Annex 2 appropriate for the (types of) NMRFs you expect? If not, what is your proposal to remedy the issues you see?
Please, see answer to question n°70.Question 72. Do you agree that, to the extent possible, new FRTB models in the EU should be approved according to updated, harmonised RTS on assessment methodology? Do you agree that, in the absence of such revised standards, relevant parts of the published RTS on assessment methodology, provided they are in line with the new requirements, should apply?
No comment.Question 73. Do you agree that a recalibrated version of the current standardised approach – for banks below the EUR 300 million threshold (as currently proposed in the CRR2 proposal) – is preferable in the EU to the implementation of the BCBS reduced SBM? Do you agree that the recalibration should be carried out simply at the risk class level by applying a scalar, such that the recalibrated approach is generally more conservative – but not systematically more conservative – than the FRTB SA?
A recalibrated version of the current standardised approach is necessary. It should be aligned with the conclusions of the Basel Committee consultative document BCBS d408 on a simplified standardised approach (R-SbM).The recalibrated version of the current standardised approach should apply both to banks with small trading books and to subsidiaries of larger banks, including subsidiaries of G-SIB and D-SIB, as long as they meet the criteria on a standalone basis for the calculation of market risk own funds requirements. It should be clarified that RWAs calculated with the simplified approach at the entity level may be considered at the consolidated level of a group.
The recalibrated version of the current standardised approach should be simple enough that it will not overburden banks or bank subsidiaries with limited trading activities. Further simplification of the recalibrated version of the current standardised approach should be sought while a sufficient level of risk sensitivity retained.
The proposed recalibrated version of the current standardised approach shall be benchmarked with the full standardised approach (SbM) in order to improve its design and calibration. It shall not be the case that the recalibrated version of the current standardised approach lead to fewer own funds requirements than the standardised approach (SbM). Conversely, it shall not lead to overly inflated own funds requirements: there should still be some risk sensitiveness and commensurateness.
French banks favour a new simplified standardised approach (R-SbM) instead of a scalar factor applying to risk classes.
Question 74. Do you have any comment on the items mentioned in this section or wish to raise additional implementation issues?
Technical standard on Default Risk Charge: We would like to express our deepest concern regarding the scope of application of the EBA RTS on DRC, as defined in the draft Regulation amending Regulation 575/2013, article 325bq(12). EBA shall develop draft regulatory technical standards to specify the requirements that have to be fulfilled by an institution's internal methodology or external sources for estimating default probabilities and loss given default in accordance articles 325bq(5)(e) and 325bq(6)(d) of the draft Regulation amending Regulation 575/2013.As currently drafted, such RTS only applies to institutions with no IRBA approval to estimate internal PD/LGD. We consider such RTS should also be applicable to IRBA-validated institutions for those issuers in DRC scope which are not covered by internal credit methodologies.
Technical standards on assessment methodology & Default Risk Charge: In paragraph 289 of the Discussion Paper, the EBA recommends that DRC guidelines (cf. draft Regulation amending Regulation 575/2013, article 325bn(2)) are directly addressed in the revised RTS on assessment methodology. We would like to go one step further and suggest that the above-mentioned DRC RTS is tackled together with the guidelines in the revised RTS on assessment methodology.
Indeed, the revised RTS on assessment methodology should then clarify that flexibility offered to non-IRBA institutions to use alternative approaches (external ratings simplified approaches) should also apply to IRBA-validated institutions for issuers with no internal PD/LGD. In case such flexibility is not granted, IRBA-validated institutions will potentially have to rate internally thousands of issuers, for which we do not have commercial relationship and hence do not rate internally (ex. equity indices constituents), meaning in most cases collecting comprehensive information (e.g. various balance sheet ratios) on each and single issuer which is unmanageable.