Raiffeisen Bank International AG / Raiffeisen Zentralbank Österreich AG
No, we do not agree. As of alternative methods we propose to use approaches based on the one defined in the IFRS 13 Fair Value Measurement, namely:
- Market approach, including the usage of proxies for the valuation;
- Cost approach;
- Income approach.
Each of the methods stated can be enhanced to reflect the prudential consideration (for example, the entity can use market approach with prudential shifts of the valuation inputs).
Consequently alternative sources of information may include:
- Market indexes, interest rates and yield curves, benchmark curves, market estimation of default probabilities (CDs curves), implied volatilities;
- Historical volatilities;
- Statement about credit worthiness of issuer / counterpart, including internal/external rating and financial statements of issuer / counterpart;
- Probabilities of default attached to internal / external ratings;
- Statement about creditworthiness of issuer / counterpart industry sector / country, their internal / external ratings and probabilities of default.
Yes, we believe that institutions which do not have significant exposures measured at Fair Value should apply the simplified approach. From the other side we found more reasonable to define a threshold as a percentage of total assets OR minimum absolute value. (For example: the simplified approach cannot be applied of the sum of absolute values of on- and off-balance sheet fair value positions is more than 25% of total assets or more than EUR 15 bn.)
We do not see any issues here.
No, we do not agree. As we have stated in the section Proposals we understand valuation adjustments as adjustments needed to cover the risk connected with non-precise or inaccurate valuation of positions caused by product complexity or illiquidity. Therefore we do not see any connections of abovementioned factors with unrealized profit or market value of underlying positions. According to the proposed approach the position with unrealized loss will not produce any valuation adjustment which is not logical on our opinion. We would notice that position with significant unrealized losses quite often represents illiquid positions, subject to market uncertainty. To summarize, the proposed simplified approach on our view will lead to distorted estimation of valuation uncertainty in the institution’s balance sheet.
As an alternative approach we would propose to use the following (similar to the standard approach for market price uncertainty in our proposals) algorithm: notional netted value of financial instrument multiplied by:
- 1 basis points in case of non-derivative instrument;
- 5 basis point in case of derivative instrument
If and only if the financial instrument is attributed to the Fair Value Level 2 or Level 3 according to IFRS 12 Fair Value management Hierarchy.
Although the proposed approach is very simple, it is covering the most essential valuation uncertainty risks – products and model complexity (via differentiation between non-derivative and derivative positions and Fair value levels) as well as indirectly liquidity, via Fair value hierarchy.
On our opinion the liquidity is one of the essential inputs to the valuation uncertainty. However for the simplified approach we would recommend to keep the algorithm rather simple, and count the liquidity of position in the indirect way via Fair Value Hierarchy. According to the IFRS 13 Fair Value management Level 1 of Fair Value position means positions valued with observable inputs without any adjustment, which implies high liquidity of the valuation inputs and less valuation uncertainty. Please look at our comment to the previous question.
No, we do not agree. We assume that either the wording of Article 6 paragraph 4 is unclear, or, if to understand the statement literally, the big room for the manipulation is available. We would propose to consider as an example the core-approach relevant bank with only non-derivatives positions (bonds), measured at fair value; an assumption also is that there is no profit out of these positions, but only losses. Following the wording the mentioned paragraph:
- 100% of the net unrealised profit on the related financial instruments – is 0, since there is no profit;
- Either: 10% of the notional value of the related financial instruments in the case of derivatives, or 25% of the market value reduced by the amount determined in (a) of the related financial instruments in the case of non-derivatives – is also 0, since the institution may claim to measure AVA based on the first option, although the exposure in derivatives is 0.
Assuming that in the second statement not either.. or is meant, but end, we find proposed volumes of AVAs too big and the dependency of AVAs from unrealized profit as not proved. As an alternative approach we would propose the standard approach for AVA measurement for each valuation adjustment separately. Please refer to section Proposals for more details.
No, we do not agree. We would state that proposed approach under the condition of sufficient market data is rather artificial and can be rarely applied in the practice, since the necessity of the valuation adjustments is increasing with a decreasing of market and position liquidity. As we mentioned in section Proposals about Prudent valuation treatment and Valuation adjustments calculation we would propose institution to decide between standard, advanced and core approach for the AVAs calculation and would set the clear dependency between respective AVAs and market liquidity of valuation input and concentration of position. Please refer to the respective paragraphs in section Proposals.
We do not agree in general, since we do not see any clear guideline on AVAs calculation in the mentioned articles. Please refer to section Proposals regarding our view of respective AVA calculation.
Moreover we do not agree with a separation of Concentration positions AVA into a different category, as we see the concentration risk as an integral part of close out costs.
We do not find appropriate to mention any specific cases when 0 AVA can be assessed, since usually it is financial institution specific and market specific.
We do not agree with a proposed aggregation approach, as well as we do not agree to the proposed methodology of respective AVA calculation. On the valuation adjustment level risks related market price uncertainty / unearned credit spread and close out costs have different origination, thus respective AVAs can be summed up. Our view on the aggregation within the AVA category is prescribed in the respective part of section Proposals.
We agree that valuation adjustments related to model risk, operational risk, investment and funding costs, future administrative costs and early termination have an independent nature and are therefore additive. We do not agree with a separation of concentrated position AVA in a separate valuation adjustment, and consequently do not agree with summing up respective AVA value. Please refer to paragraph Close out costs in section Proposals.
We agree that the calculation of prudential valuation adjustments should be back tested on a regular basis. We would also notice that not only AVAs mentioned in Articles 8 to 10 are relevant for back testing check, but also model risk and other.
We would generalize the requirement: for the institution applying core approach prudent values of At Fair Value positions should be properly documented and back tested using actual exit prices for the institution. In case of significant back testing violations results should be reported to management and adjustment in the calculation should be done if necessary.
We agree with stated costs and benefits in general; however we think it is important to notice also the following:
- Indirect costs – ineffective management: the new increasing complexity of capital requirements calculation may lead to less effective bank management. Prudent valuation regime following the good intention of the regulator to protect taxpayers against hidden valuation distortion in bank’s balance sheet should be therefore introduced smoothly, gradually and in the consistent way. We would underline once more that unfortunately in the current version of Technical standards we did not find a clear guideline about the calculation of AVAs.
- Indirect costs – unattractive markets: introduction of AVAs for less liquid positions will make investment into developing and emerging markets less attractive, which can contribute to the slowdown of respective economies development speed. Charging AVAs for complex products may also decrease the innovation on the financial markets and will not help customers to solve their financial needs.