European Banking Federation (EBF)

The scope of variable remuneration seems clear. However, the draft Guidelines do not use a discount rate to scale up 25% of the variable remuneration. Instead the formula uses a discount rate to scale up 100% of the variable remuneration and then reduces that value by a function that takes into account the ratio of the threshold (25% or lower as defined by Member States) to 1 minus the threshold to derive a small increase in total variable remuneration. In addition, the draft Guidelines do not result in 25% of the variable remuneration being deferred for a minimum of 5 years, but rather in 25% of the increased variable remuneration being deferred for 5 years. Consequently, since 75% of the increased variable remuneration is greater than 75% of non-increased variable remuneration, the employee also receives an increase in the amount of non-long term variable remuneration.
EBF does not believe that the incentive and subsequent deferral of compensation for at least 5 years that is provided by CRDIV should also result in an increase in the residual variable compensation. Instead EBF believes that:
• Only the prescribed percentage of the variable remuneration that would be paid if no long-term incentive plan was included (25% or lower) is subject to the long-term minimum 5-years deferral incentive scheme.
• There should not be any change to the balance (i.e. 75%) of the total variable remuneration that is payable in less than 5 years.
The suggested factor is appropriate, however, there are concerns regarding the use of the most recently available data. In section 7 it is stated: “For paragraphs 8 to 10 the most recently available data should be used when remuneration is awarded”. This implies that the institution, when entering the earning period for the variable compensation, is not able to calculate what the actual discount rate allowed is or should be. It is common practice in many institutions to agree upfront the terms & conditions for variable remuneration, including what the maximum ratio between fixed and variable remuneration should be for the earning period. This would not be possible if the discount factor remains unknown until the award is concluded (i.e. after the earning period in question), as suggested in the draft guidelines. We believe it is important to withhold the possibility for institutes to communicate the discount rate upfront.
We therefore suggest that the last sentence of section 7 is rephrased as follows: “For paragraphs 8 to 10 the institution can choose between using the most recently available data when the maximum ratio between fixed and variable remuneration for the year is agreed with the employee or when the variable remuneration is awarded”.
This change would have limited effects on the actual discount rate, since the inflation in a specific Member State does not significantly differ over such short time, and further, Member State inflation as such represents a small fraction of the totality when calculating the discount rate.
Yes. However, for remuneration awarded in countries outside the EU, long-term government bonds may have a risk weight. We suggest that Section 10 should thereby be rephrased to:
“For remuneration awarded in a third country the following should apply:
a. if the remuneration is paid in a currency issued by a third country, institutions should use equivalent official statistical data available for the country issuing the currency or should use the average yield on the EU government bonds with a residual maturity of around ten years, as published by Eurostat;
b. if the remuneration is awarded in a currency issued by an EU Member State, institutions should use the average yield on the EU government bonds with a residual maturity of around ten years, as published by Eurostat.”
A fixed incentive factor does not seem appropriate. The theoretical maximum increase in total variable remuneration is 33.33% (if the maximum threshold of 25% is chosen). In practice, the actual total increase in total variable remuneration will range only between 14% and 22% because of the way the formula has been designed, based upon inflation and government bonds rates ranging between 2% and 5%, and with a fixed incentive factor of 10% p.a. In the event that the threshold is lower than 25%, the increase in total variable remuneration is broadly reduced pro-rata.
The EBA formula for 5 years deferral (assuming that inflation and government bonds rate are ignored) works as follows: an incentive rate of 10% p.a. results in only 10.5% increase in total variable remuneration. A stress-test of different incentive rates reveals the following:
• 50% p.a. results in 27.7% increase in total variable remuneration;
• 100% p.a. results in 32.2% increase in total variable remuneration;
• 500% p.a. results in 33.3% increase in total variable remuneration

The above demonstrates that the increase in total variable remuneration is not proportional to the size of the incentive rate. Consequently, banks are unlikely to consider extending the deferral period significantly in order to only slightly increase the cap for variable compensation.
EBF suggests combining the factors one and two (for inflation and inflation risk) with at least one additional factor. The factor should be multiple of factors one and two combined with a multiple as an incentive and maybe a minimum factor. In addition, we believe that there is no need for difference between factors three and four (incentive for long-term deferred variable remuneration and incentive for retention period). Instead, one factor should be offered for deferral and retention, and this should be either a progressive or – preferably – an exponential factor (i.e. either 2% for the first year, 4% for the second and so on or – preferably – 2^n, where n is the duration of the deferral and retention period).
It is recognised that a retention period of x years counts for less than a deferral period of x years due to adding half the % in the incentive factor and due to that the retention period is not taken into account in the exponential (n being the vesting period only).
Still, EBF questions the purpose of requiring the usage of retention as such on top of deferrals. It is recognised that retained variable remuneration is linked to the performance of the institution. However, this also goes for deferred variable remuneration, where deferrals have clear advantages such as malus provisions tailor-made for the position, the relevant business unit and the institution.
Further negative aspects of requiring retention (as oppose to deferral) are:
• The need for clarity on whether retained money is taxed in different Member States and whether retention applies to pre- or post-taxed amounts vested;
• National labour law in Member States often prevents adding claw back provisions for vested variable remuneration subject to retention, therefore, this option cannot be used in practice;
• The administrative burden of ensuring that shares (or similar) are kept during retention period is significant and costly.
Moreover, the minimum retention period is defined as 2 years, for which there is no rationale. The same applies to the definition of this factor as 1% p.a. (i.e. 10% of the fixed incentive factor). The impact of the factor results in minor additional variable remuneration for the material risk of those additional retained years. This is illustrated in example 1 of EBA consultation paper that the employee is invited to give up EUR 26,525 that would have invested in a minimum of 3 years for EUR 30,000 (an increase of EUR 3,475 or 13%) that will vest in 5 and 6 years with further retention of 2 and 3 years.
An alternative approach to retention requirements would be that retention periods could be compensated by the institution through prolonged deferral periods of not less than half the length of the relevant retention period (applying the ratio between incentive factors for deferrals and retention respectively).
The application of the discount rate seems clear. Nevertheless, European banks believe that:

1) The portion of the award that is required to be vested for a minimum of 3 years deferred (EUR 15,000 in the example above) should remain as EUR 15,000 and, thus, unaffected by the long-term incentive:

• EUR 100,000 Award without incentive
• EUR 60,000 Immediate pay-out
• EUR 25,000 25% max of the award
• EUR 15,000 Balance to be vested for minimum 3 years

2) A discount rate should only be applied to the 25% (or to the lower chosen by Member States) of the total variable remuneration; and

3) The total non-discounted variable remuneration should remain unchanged. In the example, the amount remained at EUR 75,000
• EUR 60,000 Immediate pay-out
• EUR 15,000 To be vested for minimum 3 years
• EUR 75,000 Sub-total remains unchanged
It is evident that the cost to benefit ratio will vary between banks, as they are dependent on the number of the relevant identified staff and the workload necessary in order to fulfil the respective transparency requirements. Overall, the costs associated with these requirements should be largely administrative and not material. However, too detailed public disclosures do not only trigger additional administrative burdens / costs but also encourage employees to compare and question the institution’s numbers compared to peers. For instance, when basing the discount rate on the most recently available data on inflation and interest rate for EU government bonds (and these fluctuating), institutions will arrive at different discount rates despite having identical deferral models, as the dates for awards will vary. A preferred approach would be that, instead, relevant information is reported to the respective Member State supervisory authority.
Example 1 addresses several complex areas. There seems to be a minor calculation error on the total amount of variable remuneration, where EUR 132,797.96 is stated instead of EUR 132,622.73.
While some of EBF members consider this example clear, there are members who believe that even though the example shows the calculation of the values to be included in the calculation of the ratio of variable remuneration to fixed remuneration, nevertheless, it fails to explain that the amount not subjected to the discount rate that will be paid in less than 5 years (i.e. EUR 120,000) is not the amount that could be paid if no 5 year deferred compensation is paid, which would be EUR 132,623.
Of this amount, 60% could have been paid out immediately (50% in cash and 10% in capital instruments, in total EUR 79,574) and 40% (EUR 53,049) would have been vested for a minimum of 3 or 4 years. This 40% can be split into 20% (EUR 26,524.5) that could be increased and 20% that will be excluded. In return for agreeing to exclude the 20%, EBA proposes that the 40% of EUR 132,623 (i.e. EUR 53,049) can be increased to EUR 70,426, an increase of EUR 17,377. This increase is in fact attributable to the increase in the amount vesting in a minimum of 3 or 4 years (EUR 13,902) and the additional amount (EUR 3,475) that is awarded for changing EUR 26,525 vesting in 3 or 4 years into EUR 30,000 vesting in 5 and 6 years.
The table below is quite explicative:
No incentive Change Incentive
Cash EUR 79,574 0 EUR 79,574
Min 3 yrs vesting EUR 26,524.5 + 13,902 EUR 40,426

Min 3 yrs vesting EUR 26,524.5 - 26,524.5
5 year vesting 0 + 30,000 EUR 30,000
+ 3,475
Total EUR 132,623 + 17,377 EUR 150,000

In EBA example 1, the employee sees that 20% of the EUR 150,000 (i.e. EUR 30,000) is deferred for 5 or 6 years, however, the employee does not see that, without the incentive, he could have received EUR 26,524.5 vesting in a minimum of 3 years. More specifically, according to EBA formula, the employee would give up EUR 26,524.5 that would have vested in a minimum of 3 years, the vesting period would extend to 5 and 6 years with 2 and 3 years retention, and for this change in terms & conditions he would receive additional EUR 3,475. It can be reasonably concluded that this is a very limited benefit.
EBF considers this example as confusing and believes that it needs to be made simpler and clearer. In fact, the present value formula is not general, but applies when the ratio is 100% (and not, for instance, 200%). Furthermore, longer deferral periods are not favoured in the pro-rata model where discounted variable remuneration does not necessarily decrease with the deferral length (is rather a U-curve).
Yes. In addition, we would also suggest including examples which assume the maximum ratio between variable and fixed remuneration at 200%.
Whilst we generally agree with the impact analysis of the proposals, there is the feeling that the draft Guidelines do not provide the best means of incentivising long-term deferral structures due to a lack of practicality, and complexity. The discount heavily depends on external factors and the resulting volatility makes it difficult to explain, and not reliable enough to be used in real life compensation strategies. As an alternative, the regulator could decide on a fixed discount table of 10%, 15%, 20% and 25% if certain criteria for deferral and retention periods and / or structures are met.
Konstantinos Karountzos