We generally agree with the market data types, alternative methods and sources of information listed in Articles 3.2 and 3.3. However, we suggest Articles 3.2 and 3.3 should be reworded as the current wording implies that the listed items have to be considered concurrently, regardless of relevance. Not all data sources have the same relevance or reliability, which will be dependent on market conditions, institution specific conditions and trade specific conditions. In addition, if a financial instrument is already priced at its natural bound, an institution need not consider prudent shifts to the valuation inputs. Finally, the wording of Articles 3.2 and 3.3 should leave the market data types, alternative methods and sources of information open. The markets are constantly evolving and having a definite list will not be advisable.
We propose Article 3.2 should be reworded as:
“The market data used to determine a prudent value shall consider available and reliable data sources, including the following if relevant:
a) Exchange prices in a liquid market;
b) Trades in the exact same or very similar instrument, either from the institution’s own records or, where available, trades from across the market;
c) Tradable quotes from brokers and other market participants;
d) Consensus service data;
e) Indicative broker quotes; and/or
f) Counterparty collateral valuations.”
We also propose Article 3.3 should be reworded as:
“For cases where an expert based approach is applied for the purpose of Articles 8 to 10, alternative methods and sources of information shall be considered, including the following if relevant:
a) The use of proxy data based on similar instruments for which sufficient data is available;
b) The application of prudent shifts to valuation inputs; and/or
c) The identification of natural bounds to the value of an instrument.”
We agree that the introduction of a threshold is useful for the implementation of the simplified approach.
We propose the threshold should be refined by excluding off-balance sheet fair value assets and liabilities from AVA calculations. This exclusion will reduce confusion around what comprises ‘off-balance sheet’ and allow ease of implementation of the simplified approach.
No, we do not see any practical issues.
We recommend that the EBA modify the simplified approach by considering the following:
i. Remove references to “net unrealised profit” from AVA calculations.
Firstly, the inclusion of net unrealised profit in AVA calculations unduly penalises profitable institutions and favours less profitable institutions. Secondly, institutions which hold the same positions would have different AVA numbers due to different accounting basis in determining the cost of investment (LIFO, FIFO or average cost). Lastly, some institutions may manipulate the net unrealised profit number by closing off existing deals which are profitable and by entering into new deals in the same financial instrument.
ii. Remove references to “off-balance sheet fair value assets and liabilities” from AVA calculations.
As mentioned in our response to Q2, this exclusion will allow ease of implementation of the simplified approach.
iii. Introduce multipliers to balance sheet valuations of fair value positions based on accounting fair value levelling for AVA calculations.
Level 1 positions may have a zero multiplier and level 2/3 positions may have a non-zero multiplier. Level 3 positions may also have a higher non-zero multiplier than Level 2 positions as Level 3 positions are generally traded in inactive markets and have unobservable pricing inputs which are significant to the position’s overall valuation. For institutions where the local GAAP does not require fair value levelling, a common multiplier (e.g. 0.1% of sum of absolute value of fair value assets and liabilities) may be introduced.
Please see our response to Q4.
We are of the view that a differentiated treatment may be considered by excluding Level 1 positions from AVA calculations. Such an approach utilises readily available information and keeps the simplified approach less complex to implement for institutions with smaller fair value positions.
We recognise that the intent of Article 7.4 is to impose punitive measures on institutions which have not implemented prudent valuation approaches considered acceptable to the regulators. However we strongly recommend that the EBA consider the following amendments to avoid unintended outcomes:
i. Remove references to “net unrealised profit” from AVA calculations
As detailed in our response to Q4, the inclusion of net unrealised profit in Article 7.4 is not sufficiently robust and will create undesirable consequences.
ii. Phase in an adequate implementation period
We urge the EBA to allow an adequate implementation period for institutions to implement the prudent valuation requirements after the issuance of the finalised RTS, as the actual system and process changes required would not be known until the issuance of the finalised RTS. The determination of AVA is a complex exercise. Setting an unrealistically short timeline would backfire and lead to the undesirable effect of institutions not able to build and implement adequately reliable systems and processes for AVA reporting.
We support the overall approach to calculate AVAs for market price uncertainty, close-out costs and unearned credit spreads, with the following specific comments:
i. Zero Market Price Uncertainty AVA (Article 8.2)
We are pleased that the EBA is taking a balanced view in Article 8.2, by allowing institutions to exclude a position from market price uncertainty AVA calculations where the position is proved to have zero market price uncertainty. We share the EBA’s view that the proposed tests of a liquid two-way market and materiality of valuation uncertainty from market data sources would reasonably demonstrate zero market price uncertainty.
ii. Granularity of Valuation Exposures (Articles 8.4b, 8.4c, 9.5b and 9.5c)
We are of the view that the requirements of the above articles to prove granularity of valuation exposures are too complex operationally and have limited benefits which would not justify the significant costs of implementation. We recommend that the EBA consider alternatives by allowing institutions to either formulate their own methods to prove granularity of valuation exposures or provide the prescribed proof of granularity of valuation exposures based on a sampling approach.
iii. Zero Close-out Costs AVA (Article 9.2)
Close-out costs AVA should be required only when market price uncertainty AVA has been calculated based on exit price. We suggest the wording of Article 9.2 should be amended as follows:
“When an institution has calculated a market price uncertainty AVA for a valuation exposure based on exit price, the close-out cost AVA may be assessed to have zero value.”
In addition, we suggest that the article incorporates an option for institutions to consider direct exit price of the portfolio, i.e. costs of selling the portfolio to another market participant, in addition to using bid offer spreads for valuation inputs in assessing close-out costs AVA. This will reflect the realities of the options available to institutions in the event that an institution decides to exit its portfolio.
iv. Expert Based Approaches (Article 8.5b and 9.6b)
We would like to highlight to the EBA that as the expert based approach is applied where insufficient data exists, the relevance and reliability of available but less comparable external market data information would be questionable. Regulators should acknowledge the limited usefulness of external market data information, given that the expert based approach is applied under circumstances of data insufficiency.
We support the overall approach to calculate AVA from Articles 11 to 16, with the following specific comments:
i. Model Risks AVA (Article 11)
We generally agree with the draft article.
ii. Concentrated Positions AVA (Article 12)
We generally agree with the draft article with the proposed minor amendments:
• Remove references to “typical daily trading volume of the institution”. An institution’s ability to close out the valuation exposure is not reflected by its typical daily trading volume. An institution may choose not to trade certain positions on a daily basis for various reasons other than its inability to trade in the market.
• Rephrase Article 12.2 to allow flexibility in determining the prudent exit period across asset classes. The current wording of Article 12.2 implies that volatility of the valuation input, volatility of bid offer spread and the impact of the hypothetical exit strategy on market prices should be considered concurrently in determining the prudent exit period, which need not be the case for all trades.
• Introduce different holding periods for trading book and banking book positions. We agree that trading book positions may have a ten day holding period which is consistent with Article 365 of Regulation (EU) 575/2013. However, we view that banking book positions should have a longer holding period of one year due to the different nature of banking book positions and to be consistent with current capital computation requirements for the banking book. The EBA needs to be mindful of the dangers of specifying a shorter time horizon than market participants intended when they enter into such transactions. A shorter time horizon would imply stress conditions and curtail institutions’ ability to provide financial solutions. This would consequently create a looping effect and increase hedging costs in the markets.
iii. Investing and Funding Costs AVA (Article 13)
We propose a slight amendment to Article 13.2 to emphasise that investing and funding costs AVA should be calculated based on expected contractual lifetime and not contractual maturity date of derivative trades: “Institutions shall estimate the AVA by including the expected funding costs and benefits over the expected contractual lifetime of each derivative trade which is not strongly collateralised”.
iv. Future Administrative Costs AVA (Article 14)
We suggest that the article incorporates an option for institutions to consider the costs of selling the portfolio to another market participant, in addition to the costs of managing the portfolio over the expected life of the valuation exposures for which a direct exit price is not applied for close-out costs AVA. This will reflect the realities of the options available to institutions in the event that an institution decides to exit its portfolio.
v. Early Termination AVA (Article 15)
We generally agree with the draft article.
vi. Operational Risks AVA (Article 16)
As we have iterated in our response to the discussion paper on prudent valuation, we view that this article is a duplication of operational risk capital. We view that the current operational risk capital computation rules and its related add-ons have covered any operational risk aspects (including those arising from balance sheet substantiation controls and legal disputes) related to fair value positions. In addition, it is not best practice that any “at risk” valuation amount identified through the balance sheet substantiation or legal dispute assessments should be reflected as a CET1 capital deduction, as such “at risk” valuation amount should be reflected directly in shareholders’ equity.
Given the evolving nature of capital rules, we view that the RTS should not specify an exhaustive list of AVAs with zero value. In addition, institutions should be allowed to prove to the regulator if AVA is determined to be zero for any AVA category.
We agree with the proposed aggregation method as it is simple and straightforward to apply.
Please see our response to Q10.
We see the rationale behind ongoing monitoring of valuation prudence, but we view that the proposed method in Article 20 is too prescriptive without considering other available methods to prove valuation prudence. The interpolation method proposed in Article 20.3 is also unlikely to shed light on whether the institution’s implemented prudent valuation framework is sufficiently prudent by not adequately considering market movements between quarterly reporting dates. Given that the proposed testing approach is likely to be applicable to liquid trades with little or no valuation uncertainty, we believe that the costs of implementation of Article 20 will far exceed its benefits.
We view that alternative validation tests (e.g. trade reviews and profit and loss attribution) which are well developed in the market are more effective ongoing monitoring tools.
We are supportive of the consultation and generally agree with the benefits raised with suggestions for minor amendments, except for the following which we view would have excessive costs of implementation which would not justify the minimal benefits:
i. Granularity of Valuation Buckets (Articles 8.4 and 9.5)