KDPW is not sure how to apply the formula to calculate the capital requirements for custody risk, as proposed in Article 5 of the draft RTS. First of all, it is unclear how CSD’s activities should be mapped to the business lines, as described in Article 317 of the CRR, as non-bank CSDs are not involved in any activities of credit institutions. So-called “agency services” are perhaps the only type of services that are conceivable in a CSD context, but the definition provided in the CRR is different from the central maintenance service as described in the CSDR.

Furthermore, no matter what numbers CSDs calculate using the Standardised Approach, the results seem to be included in the calculations under the Basic Indicator Approach, as proposed in Article 4 of the draft RTS, where custody risk is already mentioned along with operational and legal risks.

According to Recital 10 of the draft RTS, while operational risk generally covers risks relating to securities owned by the same CSD, specific provisions are necessary as concerns custody risk as these refer to the risk of custody of securities in another CSD or in an intermediary in the case of indirect links. We do not understand the rationale behind this statement, at least in the context of non-bank CSDs’ activities. In particular, the custody risk faced by CSDs participants (whether the securities are held directly by CSD or via CSD links) seems to be a part of operational and legal risk of the CSD (represented either by operational incidents affecting assets held via CSD links or claims for compensation received from participants).

Therefore, in our view, Article 5 of the draft RTS creates unnecessary duplication of custody risk treatment and should be entirely deleted (along with art. 3(1)(b) which refers to art. 5). We are convinced that custody risk is already addressed in an appropriate way in Article 4 of the draft RTS.
In our opinion some of the draft standards proposed by the EBA do not adequately capture the risks arising from the activities of CSDs.

Although operational risk is surely the key risk in non-bank CSDs, the operational risk profiles of CSDs are still substantially lower than those of credit institutions, for which the 15% ratio of Basic Indicator Approach was calibrated. As far as we know, average yearly operational losses in CSDs are marginal in comparison with operational losses faced by banks. Therefore we believe that the ratio should be recalibrated to reflect lower risk profiles of the CSDs’ business models.

Furthermore, in our view, it should be possible to adjust the CSDs’ capital requirements for operational and legal risks by taking into account the effect of insurance arrangements covering those risks. For many years such arrangements have been the industry’s standard and a vital component of CSDs’ risk management frameworks. Institutions which are permitted to use Advanced Measurement Approach can, in line with art. 323 of the CRR, recognize the impact of insurance and other risk transfer mechanisms, and then reduce the capital requirements for operational risk by up to 20%. The RTS should allow also those CSDs, which calculate their capital requirements for operational risk in accordance with Basic Indicator Approach, to apply art. 323 of the CRR or a similar formula, if granted permission by their national competent authorities.

Thus, we would recommend the following amendments to art. 4 of the draft RTS:
• recalibration of the 15% ratio to 10%;
• giving all CSDs an opportunity to adjust their capital requirements by recognizing the impact of insurance arrangements.

As far as investment risk is concerned, calculating credit risk on tangible assets exposures, as regards art. 134(1) of CRR, is in our opinion not adequate to the CSD’s business model. In particular, we do not understand why the value of land and property used for the purpose of CSD’s own activities should increase the capital requirement for credit risk.

It should be considered that CSDs’ asset structures differ from those of banks, for which the Standardised Approach for credit risk was established. For example, the KDPW’s risk-weighted exposure amounts for credit risk calculated for tangible assets are significantly higher than those connected with financial resources (invested in cash or in highly liquid financial instruments with minimal market and credit risk, i.e. in accordance with art. 46(3) of the CSDR). In our opinion, the fact that the major part of capital requirement for credit risk in a CSD would be generated by tangible assets suggests that Standardised Approach for credit risk fails to be adequate with CSD’s business model.

Therefore an amendment to art. 6(2) of the draft RTS is necessary, so that tangible assets would be excluded from the calculations of risk-weighted exposure amounts for credit risk. Alternatively, the art.6 of the RTS should in general refer to art. 46(3) of the CSDR and limit the scope of the capital requirements for investment risk only to financial resources.

As regards the method of calculating capital requirements for business risk as at least 25% of the annual gross operational expenses, we would like to remark, that it was established for CCPs. As CSDs operate considerably more stable businesses than CCPs, therefore significantly lower ratio for business risk calculation should apply to CSDs. We have also doubts about the idea of the predefined business risk scenarios for the calculation of the regulatory capital for business risk (listed in Annex II of the draft RTS). In particular, points (e) and (f) of Annex II referring to long and short client balances are not applicable to non-bank CSDs.

Thus, we would recommend the following amendments to art. 7 of the draft RTS:
• recalibration of the 25% ratio to 15%;
• giving CSDs an opportunity to estimate the business risk using any reasonably foreseeable adverse scenarios relevant to their business model (if granted approval by the national competent authority), instead of referring to predefined scenarios (Annex II should be removed).

Finally, as far as capital requirements for winding-down or restructuring are concerned, we are afraid that some of the events proposed in Annex I are not applicable to non-bank CSDs (e.g. “the failure of significant counterparties” or “a severe outflow of liquidity”, as mentioned in paragraph 6). Again, we believe that CSDs should be allowed to calibrate the scenarios for winding-down or restructuring according to their own business models. Therefore we would appreciate more flexible approach in the RTS, going in line with EMIR, both to CSD’s analysis of the scenarios for winding-down or restructuring, and as already mentioned, to business risk scenario analysis.
Anna Zielinska