11 December 2018

ESMA publishes the responses to its Consultation on stress testing guidelines for Money Market Funds

AFG

A general issue AFG wishes to raise is about the too granular level of detail and the amount of the stress tests required. It is important to mention that unlike AIFMD stress tests, this MMFR approach to stress tests contains also a calibrated leg which is totally new and somewhat too cumbersome for a first approach. AFG believes that due attention should be paid on this calibrated part so as to choose a limited number of stress tests that are simple to implement, useful and that allow for less need of recalibration or change every year. On these limited calibrated common stress tests, please bear in mind that a calibrated approach needs to specify as much as possible the parameters. For instance, parameters should be opposable and known like external credit ratings for instance.

In general, AFG insists on the fact that there is a lack of flexibility for the asset manager to choose a more appropriate (for asset managers) methodology for stress tests. Indeed, the money markets environment is very different from the bonds markets, where transactions and bid/offer data are transparent and available on IT providers’ platforms. The consultation document seems to consider that bid/ask spreads are available for money markets instruments which is not the case.

Stress testing should take into consideration the specific features of MMF as an asset class. This is currently not reflected in the proposed stress test scenarios. For example  (i) long-term stress tests for MMFs are not meaningful (ii) considering that MMFs are liquid assets for less than 100% of their NAV is unfunded. 

AFG would also like to stress that it would be helpful keeping the common standardised reporting stress test simple and avoid any supplement to what the level 1 text requires.

AFG members believe that the idea to look through the assets of the underlying MMF and apply stress tests to those assets seems totally disproportionate and unfeasible given that underlying MMFs have all conducted the exercise separately. In line with para. 17 of the Consultation, further approaches could be used.  AFG suggests that the manager should also be able to rely on the stress test results conducted by the manager of the underlying MMF (where provided) or use a standard factor.

AFG would also like ESMA to clarify which part of the stress tests are custom and required from the application of the Regulation to the fund and which are implemented for the reporting purposes and with which deadline, bearing in mind for the second part that all companies, including smaller ones, need time and budgets to implement. There should be no requirement to implement the calibrated stress tests before the first reporting period. Only the March 2018 Guidelines on stress tests under the Article 28 are to be implemented once the MMF is authorised under the MMF Regulation. For the sake of clarity, AFG insists on the fact that funds are not managed with the stress test tool and asset management sector does not work on the same basis as the banking sector with pass/fail mindset. Thus, stress tests should continue to be given the place they have in the fund’s operation as one tool among others and do not lead to requirements that bear disproportionate implementing costs and deadlines.

Full AFG response

____________________

BVI

BVI welcomes an approach that reporting guidance for stress test results needs to be clear and consistent in any case (for example, whether the information should be reported as a percentage amount in relation to the NAV or in text form). This would lead to a better data quality to analyse the outcome by ESMA and/or the ESRB, where relevant. The reporting template mentioned in Article 37(4) of the MMFR and ESMA’s stress test guidelines published in March 2018 already presents a table which is designed to ensure such data quality. 

Moreover, compared to the requirements of the AIFMD, BVI is aware that Article 28 of the MMFR specifies reference parameters and factors for the stress test scenarios which are reflected in the reporting template. However, the reporting rules of both European acts are identical in so far as only the results of stress test should be reported without any specific requirements how to calibrate and measure the impact of the shocks. Therefore, it is of utmost importance to maintain the already implemented principle-based approach in ESMA’s guidelines for the question how to calibrate and measure the impact of the specific shock scenarios. In particular, the design of stresses should account appropriately for the business model and the size of the asset management company. This applies all the more for the situation in Germany where management companies often simultaneous manage many different investment fund types and a money market fund is only one among many funds. Therefore, European guidance must be designed in such a way that the management company is able to adjust the stress test scenarios to the specifics of the managed funds such as size, strategy, pricing and risk models.

However, BVI has the strong impression that ESMA leaves now the principle based approach to stress tests when proposing completely new implementation processes for stress test reporting. In particular, it strongly disagrees to predetermine each factor by specific values and figures in ESMA’s guidelines which should be used in a similar manner by all MMF managers in its reports. This applies all the more as ESMA proposes to calibrate factors by using commercial data such as discount factors for bid-ask spreads. This would lead to the situation that each MMF manager would be required to pay licenses to be allowed use such external data although the data may not be required by the internal stress test design of the MMF manager.

Full BVI response

____________________

EFAMA

The reporting rules of both AIFMD and MMFR are identical, but compared to the requirements of the AIFMD, Art. 28 of the MMFR specifies reference parameters and factors for the stress test scenario. Since art. 28 of MMFR deals in general with the stress testing, with no specific reference to reporting, it seems that the parameters that ESMA will indicate in the second part of the Guidelines could disregard the reporting.

It is of utmost importance to maintain the already implemented principle based approach in ESMA’s guidelines on how to calibrate and measure the impact of the specific shock scenarios. Efama indeed disagrees with having two sets (approaches) of stress tests: the calibrated one on top of the principle based one, which is appropriate and sufficient. In particular, the design of stresses should account appropriately for the business model and size of the asset management company.

As it stands, the proposal is far too prescriptive for principle based guidelines, and may have little bearing to the stresses the MMF is likely to experience. In general, our members have highlighted the fact that there is a lack of flexibility for the asset manager to choose a more appropriate methodology for stress tests. The money markets environment is very different from the bonds markets, where transactions and bid/offer data are transparent and available on IT providers’ platforms. The consultation document seems to consider that bid/ask spreads are available for money markets instruments which is not the case.

Due to increasing intermediation in the investor chain it can be increasingly difficult for MMF managers to know the proportion of investors that are retail and those that are institutional. EFAMA suggests that managers have flexibility in calibrating stress tests which may best suit their knowledge of investors’ profiles. Having a clear picture on the largest two investors can be a rather challenging task, in particular when considering potential retail sub-distribution networks and we believe to rely on most recent and accurate data to formulate reasonable hypothesis in that field for example.

Full EFAMA response

___________________

Amundi

Amundi does not support the idea to develop transversal ST that would apply to all the financial industry on the basis of scenarios and calibration determined for banks and insurers. Typically leverage which is a key issue and a justified focus for banking regulation is not relevant for most funds since only hedge funds will use leverage substantially. The aim of the ST is not comparable and there is no such approach as minimum capital requirement in the fund industry. STs on funds which are not hedge funds  aim at controlling the risks of loss or illiquidity that investors may suffer in a crisis. It does not primarily aim at preventing a fund from going bust which is perceived as a theoretical issue for UCITS and AIF without substantial leverage. Furthermore a transversal ST will not, contrary to a common place thought, help in getting a global view of financial risk throughout all participants. Asset managers (AMs) run funds and mandates for investors who bear the risk. These investors’ duty is to monitor and control their exposure to risks and they generally stress test the behaviour of their investment portfolio as part as their own risk management. When they are banks or insurance companies they use a look through approach that comes at the end to a risk management (including ST) of the portfolios managed by AM. Such overlaps may lead to the erroneous conclusion that risks add up when they do not. If factors of risk may be common, calibration should not be and hierarchy of the most important risk factors will also differ.

Considering the reporting, Amundi believes that instead of detailed data as foreseen in the guidelines the production of one global indicator and/or more detailed comments would be more appropriate. Access to data is available for NCAs when and if they want to further investigate and control. The more data are required the more the report will look like a box ticking exercise and lose the initial objective of ST: better asses risk and prepare to handle crisis.

If standardised ST were made mandatory for the purpose of reporting it is the responsibility of ESMA to make sure that all MMFs will be stressed identically, using the same shocks calibrated identically on factors which are defined exactly in the same manner. In other words, ESMA should produce a detailed reference data basis and a description of scenarios without leaving any space for personal interpretation for these standardised ST scenarios. It is the only way to ensure comparability.

Full Amundi response

_________________________

Full consultation paper

© ESMA