In the heart of Brussels, a diminutive bronze statue portrays a small boy, stark naked and nonchalantly answering the call of nature into a fountain. Manneken Pis – the little peeing man – has outlasted several occupations of Belgium’s capital by powerful European empires and has come to symbolise the defiant spirit of its inhabitants.

Now, in the Council of the European Union, Belgian diplomats are taking a similarly defiant stance. Along with a coalition of smaller EU member states, they are in a standoff against the two largest states in the union – and the European Commission itself – over two major aspects of the Commission’s proposals to apply a floor to internally modelled capital requirements.

These smaller countries – in many cases hosts to the subsidiaries of larger EU home-states’ banking groups – want to apply the capital floor to those subsidiaries as well as at the broader group level proposed by the EU. Their intention, in large part, is to avert the eventuality of a host nation needing to resolve another country’s banking failure on its turf.

“If it comes right down to it, and the choice is between a whole group failing or a subsidiary being cut loose, host regulators simply can’t take the risk – and have to protect themselves in advance,” says Rod Hardcastle, a director in Deloitte’s centre for regulatory strategy.

The host nations pursuing this agenda are Belgium, Bulgaria, the Czech Republic, Ireland, Latvia, the Netherlands, Poland, Romania and Slovakia. Separate sources – including a diplomat representing a member state in the Council – identify Belgium as spearheading the move.

“There is one superhost and that is Belgium,” says a senior regulatory affairs manager at a European bank. “They lead the troops, and they mobilise countries.”

If they are successful, it could lead to higher capital requirements for some of Europe’s major banking groups, headquartered in the largest member states.

France and Germany, home to some of these cross-border banking groups, want to avoid this outcome, preferring instead that the floor be applied at the consolidated level only, as the EU proposed.

The output floor is one of the final elements of the post-crisis framework developed by the Basel Committee on Banking Supervision and agreed in December 2017. It requires a bank’s risk-weighted assets (RWAs) – the risk of its business as measured by its own internal models – to always be at least 72.5% of the total RWAs it would generate under regulator-set, standardised approaches. On October 27, 2021, the EC unveiled proposals outlining how the bloc should incorporate these final elements into the Capital Requirements Regulation (CRR III).

The proposals would require banks to calculate the floor at a consolidated level only – using the entire group’s assets – enabling them to recognise offsets between the floor’s effects on different business lines and thereby reducing the level of the floor.

But, in another flashpoint for EU member states, a series of transitional measures would dampen the floor’s impact until the expiry of an extension in 2033 – a deadline some states would like to see extended. Others want the measures deleted altogether.

Both fights will influence whether and when the floor either tightens – or eases.

Home-host standoff

The home-host floor fight is the latest clash in an extended struggle between the two camps over prudential rules. While the host camp can claim victory in the most recent battles – the CRR requires capital and liquidity requirements to be calculated at all levels of consolidation – the Commission’s departure from this precedent has stoked fury among host states.

“They feel that there’s been a kind of a breach of the home-host balance,” says a representative for an EU member state of the Council’s CRR discussions.

“The Commission’s proposal breaks that equilibrium, so it’s a matter of principle to a lot of those member states. It’s not going to move through the Council without those member states accepting the compromise.”

Host nations argue that the proposed application of the floor is a departure from the way capital requirements are set in the EU.

But the main reason they want more capital held at individual subsidiaries is that they do not believe banking groups will support local entities in all circumstances, which could lead to local regulators having to clean up a mess made by another country’s bank.

More capital at the subsidiary level would also ensure that the internal models of local entities aren’t underestimating the risks subsidiaries run locally.

“You can see there being a rationale from the host regulators’ perspective as to whether or not economic risks are going to be underestimated in the subsidiary,” says Monsur Hussain, head of financial institutions research at Fitch Ratings. “The [local] model risks might well be underestimated if the output floor is applied solely at the consolidated level.”

A counterargument to this view is that it could reduce banks’ ability to provide support to a weak part of the group.

“If one institution should get into trouble, the parent won’t be as able to help out this subsidiary [because] capital could be trapped in another subsidiary,” says Louise Sofie Skiffard, head of capital planning at Danske Bank.

The full shortfall doesn’t necessarily have to be met by such subsidiary resources, however. Banks can issue guarantees from their head office against some of their subsidiaries’ risks, which reduces the total RWA of the subsidiary.

You can see there being a rationale from the host regulators’ perspective as to whether or not economic risks are going to be underestimated in the subsidiary

Monsur Hussain, Fitch Ratings

“You can inject more equity into your subsidiary, and you can guarantee some of their risks,” says Adrian Docherty, head of bank advisory at BNP Paribas. “Both are in widespread use but [they] have limits to their application, such as internally set large exposure measures and going through various local legal processes.”

While acknowledging that if subsidiaries have more capital, the risk of their getting into trouble will be smaller, Skiffard says that “application on a consolidated level will make for a more efficient distribution and use of capital”.

Meanwhile, the Commission put forward a redistribution mechanism which would ensure that capital requirements set by the floor are allocated to subsidiaries. The details are vague, but the broad idea is that, if a group is bound by the floor, then the extra capital beyond the internal model calculation should be distributed according to each subsidiary’s contribution to the breach of the floor. Two sources say this was intended to head off opposition from host member states – unsuccessfully, it seems.

The senior regulatory affairs manager at the European bank is disappointed by the hosts’ speedy dismissal of this compromise without carrying out a proper impact assessment of the effect of the redistribution mechanism and how much capital it will allocate to subsidiaries.

“It just shows it is so politicised because this compensation mechanism that the European Commission has put in the proposals [has not even been] considered properly,” says the regulatory affairs manager.

The host member states’ complaints about the redistribution mechanism focus on its complexity – an aspect that is acknowledged within industry circles.

Their own proposals would increase capital requirements for some banking groups and likely result in more banks having the floor determine overall capital requirements than would have been the case with a consolidated-only floor. Although no-one knows precisely what the difference in capital will be for the European banking sector.

The closest answer is a study by the European Banking Authority, presented to the European Commission in March 2020. Among its findings, the report says that one benefit for host member states is that applying the output floor to individual entities would level the playing field between subsidiaries of large foreign banking groups in their jurisdictions and their own smaller national banks.

Goliath vs David

Those in favour of a consolidated-only approach argue theirs is the way the Basel Committee envisages the rule to apply; the standards state that all Basel III rules should be applied at a consolidated level for internationally active banks.

But the joint efforts of the host nations are making them nervous that the opposing camp will prevail.

“There is a level of co-ordination with the host member states, which is why we’re very concerned that we will lose the consolidated approach overall,” says the source.

For large banking groups, the bottom line is that the hosts’ proposals could potentially lead to higher capital requirements – particularly for groups reliant on internal models throughout their businesses.

For the ‘home’ camp, France and Germany house the largest banking groups in Europe and so would stand to benefit from the potential offsets between businesses that are – and are not – affected by the floor.

For example, low-risk businesses such as residential mortgages are assigned low RWAs by internal models due to low historic losses in Europe. That isn’t reflected in the standardised risk weights used in the Basel Committee’s credit risk framework, which is calibrated to reflect the combined jurisdictions’ markets.

Meanwhile, internal models produce high RWAs for consumer lending, consistent with the standardised approach. If an internal model produces RWAs that are above the floor for consumer lending, that gives more room for mortgage RWAs to be below the floor elsewhere in the group.

One disadvantage for the home camp, however, is France’s current EU presidency – a role that charges it with guiding legislation through the Council – which means it cannot be seen to take a proactive stance in any one side of a debate. Its presidency is due to expire at the end of June when the Czech Republic will take on the mantle. It seems unlikely that the Council will be able to agree the final version of the CRR before then.

For the European Commission, an output floor applied at individual level would go against its objective of creating an integrated single market, because it would increase costs for European banks running large cross-border businesses.

“The more internal allocation you have, the more fragmentation in the EU single market and crucially the banking union,” says Scott Martin, a senior manager in Deloitte’s centre for regulatory strategy. “Having fragmentation for cross-border banking groups that are operating within the eurozone is totally against the European Commission’s macro-policy objectives.”

Bottom line

The EBA report submitted to the European Commission in March 2020 shows the baseline impact of the output floor on a consolidated basis – without transitional measures – reflecting compliance with the Basel Committee framework.

The capital increase from the reforms is mostly concentrated in the largest globally active banks in the EU, according to their last monitoring report published in September last year. This is evident in the output floor – for internationally active banks with Tier 1 capital in excess of €3 billion ($3.27 billion), the floor resulted in a 7.4% increase, whereas all other banks experienced only a 2.2% impact.

The deviations from the Basel Committee’s output floor aren’t taken into account in the report, which was published before the European Commission’s proposals.

“In terms of estimating and understanding the actual impacts of the solo-based requirements versus the consolidated level, it’s really difficult to get a firm grasp of those numbers,” says Fitch Ratings’ Hussain.

While the EBA report looks in most detail at the impact of the output floor applied at consolidated-only, individual-only and at all levels, the EBA itself cautions against making conclusions based on its data, due to several limitations.

These include the fact that the data is from a reduced number of European banking groups; in a previous study, published in August 2019, the EBA analysed data from 115 banking groups, whereas the study of the different floor levels uses data from only 51 banking groups. The study also left non-EU entities out of its calculations.

Counterintuitively, it found that capital requirements would be higher overall for European banks with a consolidated-only approach versus a floor-only approach for individual entities. One reason is that a consolidated approach relies on the parent company’s method for calculating RWAs, ignoring methods used at each subsidiary.

A parent entity using an internal model could therefore find itself bound by the consolidated floor on all the group’s assets. By contrast, if individual subsidiaries use only standardised approaches, the application of the floor to each subsidiary would have no impact on each entity’s capital requirements.

The host camp doesn’t advocate an individual-only floor, however, but a floor at all levels – consolidated and individual. As a result, all banks would be caught whether the floor is more binding at consolidated or at individual level. And those banks for which the floor would be binding at both levels would face having a minimum capital requirement set at the higher of the two, and for a certain amount to be trapped in subsidiaries.

Large banking groups would feel the greatest impact on capital requirements if the floor were calculated at both the entity and overall group levels, in tandem with the deletion of the temporary measures. The least impact to them would stem from a floor calculated at parent-entity level, using the whole group’s assets, coupled with the temporary measures becoming permanent.

Transition trials

A series of transitional measures within the CRR proposals, designed to dampen the impact of the floor, nonetheless adds to the unpredictability of the floor’s impact – for various reasons.

Firstly, the European Commission has powers to extend the relief period past 2033. Secondly, one of the measures relies on member states to activate the relief, which opens up potential divergence between EU countries.

“The European Commission proposal allows options that delay the actual implementation of the output floor as designed by the Basel Committee,” says Deloitte’s Hardcastle. “That’s one of the reasons it is quite hard to make an assessment of what the effect of the output floor really will be.”

“If we take into account the proposed transitional arrangements… we can expect that the output floor will restrict the group at the earliest in 2033 when the transitional arrangements are set to lapse,” says Danske’s Skiffard. “It is of course important to state that the outcome of the negotiations is uncertain and may lead to further adjustments based on the final adopted rules.”

The transitional measures also lower the standardised risk weights for counterparty credit risk in derivatives trades, residential mortgages and unrated corporates.

European regulators have warned the Commission against using powers stated in the legislation to extend the measures further.

The Commission proposes to reduce the risk weight for unrated corporates, and to lower standardised risk weights in the output floor calculation for mortgage exposures secured by residential property, which are deemed to be low risk.

It is up to member states to decide whether the residential mortgage relief can be applied to mortgage exposures in their country. It will then be applied to all those exposures, regardless of whether the mortgage lender is based in the country itself.

This leaves concerns that some banks will not be able to benefit from the relief because their local authorities oppose the measures as deviations from the Basel Committee’s rules – or as a potential financial stability risk if mortgage markets in some EU states are seen as overheated.

“What we find extremely bizarre is that the European Commission has put the discretion to apply this mortgage relief in the hands of member states,” says a senior regulatory affairs manager at a European bank. “That will impact the output floor.”

These carve-outs for mortgages and unrated corporates have a very long transition. I think the imperative is that they are made permanent

Adrian Docherty, BNP Paribas

In a leaked document obtained by Risk.net, Germany, for example, states that the EU should be cautious of the potential risks stemming from the sector amid soaring prices but doesn’t propose to delete the measure. Germany’s national supervisor, Bafin, recently announced plans to increase its national countercyclical capital buffer from 0% to 0.75% and set a national systemic risk buffer at 2% on exposures to German residential mortgages. The measures are set to take effect from February 1, 2023.

But before member states are given a choice on the relief for mortgages, they must agree on the transitional measures as a whole. Comments in the working paper from December last year show a stark difference in attitude between member states that want to extend the transitional measures and those that want to get rid of them altogether.

Battle lines appear to be drawn along a similar pattern to those around the capital floor.

Belgium proposes to delete the measures for unrated corporates and residential mortgages but would allow for the SA-CCR relief to last until the Basel Committee proposes an alternative multiplier, rather than the EC being able to make the relief permanent.

Finland and the Netherlands make amendments that would delete all three measures, while the Czech Republic, Ireland and Poland label some or all of the measures as deviations from the Basel Committee’s rules.

At the other extreme, Denmark argues for the transitional reliefs for unrated corporates and residential mortgages to be made permanent. France also supports the possibility of the transitional measures being extended.

“These carve-outs for mortgages and unrated corporates have a very long transition,” says Docherty of BNP Paribas. “I think the imperative is that they are made permanent. They’re necessary in order to make the CRR implementable, but they can’t then be removed at a later date.”

Suggested amendments made by France would set the unrated corporates exemption to lapse when the coverage for ratings in each member state reaches 75% rather than at the end of 2032. Corporates with ratings usually benefit from lower risk weights under the standardised approach.

France’s proposed target is ambitious, especially as it applies to every member state individually. A study published within the EBA’s August 2019 report found that only 21% of bank exposures to small and medium enterprises had external ratings, and only 25% of non-SME corporates.

Germany stresses the unrated corporates exemption is of utmost importance and proposes the EC could extend the relief by five years.

Once these transitional reliefs lapse, the consolidated output floor is expected to bite some EU institutions quite badly.

But if they never lapse, the floor may never come into play.

Editing by Louise Marshall