We broadly agree with the EBA’s methodology. Furthermore, the industry very much appreciates the helpful changes that have been made since the methodology was introduced in the EBA’s discussion paper. We particularly commend the EBA for reflecting the industry feedback and suggesting that mapping should be at the risk factor, instead of the ‘most material’ risk factor level.
Further issues of note
Cap on liquidity horizons
One area that requires further thought is the cap on liquidity horizons as defined in article 325bd (4). We are aware that this issue is not part of the EBA’s mandate and therefore not directly consulted upon. However, we want to stress this concern in our response. We believe that capping the LH for a risk factor to the maturity of the related instrument is not appropriate unconditionally and should be an option, and not a requirement.
From a theoretical standpoint, the maturity cap applies that no position shall be renewed beyond the expiry date, which does not reflect the nature of trading and client behavior whereby positions are often rolled over at maturity of the contract. From the coherence of the framework perspective, it is inconsistent with the general constant risk assumption prevailing in the ES formula itself, in which risk factors (from 0d to 10d, from 10d to 20d, from 20d to 40d, etc.) are deemed constant throughout the life of the positions in spite of the portfolio ageing. Added to that, it renders the allowance provided in the FRTB to increase the LH of a broad risk factor category at the desk level to avoid breaking of hedges ineffective. Finally, this capping requirement will create added operational and computational burdens to implement. We recommend that banks should be allowed to assign all instruments to the regulatory LH; capping at instrument maturity should be an option.
For example, a long small capitalisation equity position may be delta hedged by selling futures contracts. Futures are often kept till a few days to expiry before being rolled. When the future’s expiry falls to or below 10 business days, its liquidity horizon is capped at 10-day, while the equity stocks position remains at 20-day. The resulting broken hedge materializes in a long directional position in the 20-day partial expected shortfall measure [CRR2 Art 325bc(1)].We therefore recommend that the mapping of risk factors to Liquidity Horizons is aligned with the following steps. This is prescribed as in the Basel framework (in fact materially using the same wording) and clarifies the order in which the Liquidity Horizon capping should be considered. This allows banks to potentially extend Liquidity Horizons at the desk level and hence restores the original spirit of the framework, effectively introducing option not to cap the LH where not appropriate (subject to usual documentation and governance standards):
1. Banks must map each risk factor on to one of the existing risk factor categories using consistent and clearly documented procedures (see (1) of article 33.12 of the Basel paper);
2. Banks must cap Liquidity Horizon at the maturity of the related instrument; and
3. On a desk-by-desk basis, the Liquidity Horizon can be increased relative to the values as assigned via above steps 1 and 2 (i.e. the liquidity horizon specified below can be treated as a floor). Where the Liquidity Horizon is increased, the increased horizon must be 20, 40, 60 or 120 days and the rationale must be documented and be subject to supervisory approval.
As noted in the consultation paper, the treatment of bank’s interest rate risk in its domestic currency has been included in the level 1 text and therefore not part of the draft RTS. Nonetheless, we would like to bring an issue related to domestic currency for interest rate risk to your attention. Specifically, as it concerns non-EUR European currency interest rate risk.
As the Level 1 rules state under the IMA, interest rate risk in a bank’s domestic currency is considered to belong to the most liquid (10 day) bucket. The standardised approach has similar favourable treatment. On the basis that domestic currency is the same as reporting currency, then in our view this approach will inadvertently penalise banks operating with a significant presence in several countries and (home) currencies and in doing so create a barrier for banks from other jurisdictions to have significant participation (in the case of Europe) in non-Euro European interest rate markets.
For example, a bank whose domestic currency is DKK would be able to put DKK interest rate risk in the 10 day liquidity horizon bucket under IMA, while a non-DKK domestic currency bank - even those with significant presences in the DKK market - would have to consider DKK risk in the 20 day bucket, even though the risk is the same. We believe the rules create a competitive distortion in parts of the EU market, that may directly lead to a reduction in liquidity in these markets.
We recommend that changes are made to CRR to permit the relevant national competent authority to classify a local currency as “domestic” for specific banks domiciled elsewhere. Criteria for this determination could include access to the local central bank as well as a large presence in the local market.
While we understand domestic currency was not part of the EBAs level 2 mandates, we felt it important to raise what we consider a relevant issue for the European implementation of the FRTB. We therefore bring this issue to your attention – given the EU markets outside the Eurozone - in the hope you may take this issue forward in dialogues with EU policy makers and supervisors. We note that in practise this could be solved via a clarification of the definition of the word “domestic” (in contrast to reporting currency) either in the finally published EBA RTS standard or alternatively directly in the level 1 text.
Completeness of the list in Annex II
Having reviewed the list of currency pairs in Annex II, we would like to draw your attention to an omission from the list. All first-order crosses based on DKK/EUR, including USD/DKK, JPY/DKK and GBP/DKK have been included apart from CHF/DKK, which we believe should (for completeness) be added to annex II.
Proprietary indices and similar products such as funds and baskets, based on multi asset class components, reflect risks embedded in more than one category or more than one subcategory.
For multi-underlying products such as indices, funds or baskets, where a single risk factor is used to model it, we recommend adopting a liquidity horizon determination aligned with the RTS Article 2 proposal for handling underlying products belonging to several broad risk categories.
In what follows we assume that the multi-underlying product is a basket (long only positions) of bonds, equities and commodities. The broad risk category to which the multi-underlying product may be assigned is either Credit spread, Equity or Commodity.
1. A broad risk category shall be determined. This will affect both the partial expected shortfall measure [Art 325bc(1)] to which the risk factor is assigned and the liquidity horizon it may be attributed to.
Weights are summed for each broad risk category. The broad risk category of highest sum is assigned to the basket risk factor.
2. The attributed liquidity horizon should be the closest supervisory liquidity horizon to the underlying instruments’ weighted average liquidity horizon.
For example, a basket made of 20% of high-yield corporate, 40% of large capitalisation equities and 40% of small capitalisation equities.
- 20% of the basket belong to the broad risk category Credit spread, 80% to the broad risk category Equity. The assigned broad risk category should therefore be Equity.
- The weighted average liquidity horizon is 20%·60+40%·10+40%·20 = 24 days. The risk factor shall be attributed a liquidity horizon of 20 days (24 being closer to 20 than to 40)
For the basket volatility risk factor, we propose to keep the same broad risk factor category used for prices (step 1) and use the weighted average of supervisory liquidity horizon in step 2.
While we appreciate EBA’s effort in relation to homogenous indices (as well as similar products such as funds and baskets), requiring that the LHs should be determined using the weighted average LH of the constituents is problematic in practise. It results in operational burden while it can also lead to situations where small changes in a component’s ratings can result in significant overall change for the entire index.
As an example, even one constituent of an IG index being downgraded will result in the weighted average LH to become slightly greater than 20 days, and at that point, the LH of the entire index will have to be switched to the next highest LH, i.e. 40 days. This appears counterintuitive as an index is usually more liquid than its components. This approach can also lead to cliff effects when e.g. a single constituent of a “large cap index” falls temporarily into small cap category.
Separately, the approach may result in significant computational burden for risk factors that differ from spot/credit spread risk of homogeneous indices (e.g. volatility, dividend or repo), particularly in cases where the components’ weights may not be directly observable (e.g. for volatility index, the weights depend of the implied correlations between the volatilities at component level).
We recommend that, for homogeneous indices:
• When the composition of the index is standardised, the liquidity horizon should be determined consistently with the nature of the entire index. For example, a corporate IG index should be mapped to a liquidity horizon of 40 days and a corporate HY index to a liquidity horizon of 60 days;
• Otherwise, the weighted average liquidity horizon may be used and assigned to the closest corresponding liquidity horizon (out of 10, 20, 40, 60 and 120 days). For example, if the weighted average liquidity horizon is 26 days, the liquidity horizon of the index would be 20 days and if the weighted average liquidity horizon is 33 days, it would be 40 days.
This approach would accurately reflect that indices are economically more liquid than their components and keep as much comparability as possible between banks that opt for a look-through approach for index in ES and banks that opt for a non-look-through approach.
Furthermore, to allow for operational simplicity in the approach, we recommend that the update frequency of liquidity horizons should be aligned with the frequency of the other processes that impact directly the scope of the risk factors in the internal model (RFET and eligibility tests), i.e. on a quarterly basis.
Yes, index like products such as funds and baskets. For further feedback, please see our response to question 2 above.
The article 2.3 does not formally cover rates indices (e.g. IR swap-based indices).
The industry prefers the option two for identification of large capitalisations, as it simplifies the process and aligns the FRTB framework with other regulatory standards in the EU. However, we believe that the large cap definition be extended to include all indices in Annex 1 of 2016/1646 and not just those “quoted in the EU”.