3(a) does not actually specify how long the duration should be. However, there is a more fundamental issue, in that we do not think using a minimum duration is the correct approach. In a time of stress, duration may not be a factor; if the institution is struggling, each depositor will think of moving, weighing the risk up against the insurance given by the relevant insurance scheme, regardless as to how long they have been with the bank. While we agree the nature of the savings and transactional accounts are different, duration is not a significant risk factor. We therefore suggest the EBA removes this wording from the definition of established relationship.

Another issue would be the definition of “active” accounts. The EBA needs to provide some further details, and also consider outlining what could be used as a potential test of the definition.
We agree with the definition of transactional accounts as stated within the consultations. However, we do not believe this should not be limited to just those that receive a salary paid into them, particularly as the retail deposit category also captures SME depositors, some of whom will be paying out salaries. All regular payment activity should therefore be taken into account e.g. salaries, rents, direct debits, and regular business payments.
The key issue is what constitutes an “established relationship” is individual to each institution. For example, for some banks this may mean long standing savings accounts. For others, insurance products or services such as investment management might be included. We therefore feel this should be left to each institution to determine based on their own business model and internal risk management reviewing what is considered stable or “established”.
We are very concerned by the potential impact of cliff effects i.e. should a deposit go one penny into the next bucket, the entire deposit will be subject to the higher outflow. This is clearly disproportionate and the EBA needs to consider an approach that will avoid these cliff effects.

Further clarity needs to be provided on what currency is to be used, as we are concerned small fluctuations in currencies could result in a borderline deposit moving from one bucket to another on a regular basis.

Therefore we would suggest considering all deposits captured under Article 421(1) as very stable (with the lowest possible outflow rate) and define certain threshold for high and very high deposits which should have a sufficient buffer to avoid above mentioned ‘cliff effects’. In our view these thresholds should be at least EUR 200k for high value deposits and a minimum of EUR 1m for very high value deposits.

We also believe it would be better for the thresholds to be referenced to deposit guarantee levels rather than absolute numbers in EUR. As already stated in the response, a small movement in FX rates could lead to deposits moving in and out of these buckets. It is likely the €100,000 threshold was decided based on EU deposit guarantee schemes. The equivalent in the UK is currently £85k and so should be used for GBP deposits. The very-high deposit threshold could then be a factor of the deposit guarantee level, i.e. 5 times the amount which is guaranteed so for the UK this would be £425k.
We agree with the principle behind what is discussed in the CP, but we are concerned as to how benchmarking would operate in practice. Customers generally go for yield rather than stability.

Any kind of rate correlated criteria should be used for short term deposits only (overnight and initial terms and notice periods <= 30 days). Initially rate driven investments do not reflect the current re-investment motivation, which depend on changed personal preferences (e.g. asset and liability structure, consumer behaviour of the customer and relationship).

In addition we want to highlight that given the current low interest rate environment only a few basis points difference might trigger a change in the regulatory treatment and classify a certain product as rate-driven. This might become even more difficult since quite a few products have a combination of various components which makes it almost impossible to compare. Moreover we think it is difficult to define an appropriate peer group, e.g. comparing globally operating banks does not factor in the given circumstances in the respective home country like economy, house banking relationship or individual/ voluntary DGS. Furthermore, defining a peer within the respective home country seems to be challenging since business model might differ tremendously.

In any case, the EBA need to provide greater detail on how rate-driven deposits will be determined. There are a number of reasons why a deposit could cease to be rate driven, including the end of a bonus period, unilateral rate cuts, not passing on increases in central bank rate or more competitive products being offered elsewhere. At present, it is not clear enough how this definition will be determined.
The EBA needs to provide further details on how firms should decide who is resident and non-resident.

To a certain extend we would agree that non-resident deposits might be less sticky compared to resident deposits. However, this only applies to locally operating banks focussing on their respective home country. For globally operating banks this is entirely contagious to the defined business model which is explicitly dedicated to serve global clients – i.e. clients can have various accounts in different countries with different currencies related to their personnel lifestyle (e.g. multiple residences).
As he rules are yet to be finalised, it is not possible to give an accurate indication of the costs. However, one issue which is likely to cause significant cost is backdating the rules to clients who already have deposits. This will be a laborious process that is likely to prove very expensive, and the occurring costs for implementation will outweigh the proposed benefits.
BBA has no comments.