The trouble with Danish, squared

Business Mayor

23-04-2025

Unlock the Editor's Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
When a piece of financial regulation has the word 'Compromise' right there in the name, you can tell that it's going to cause problems. And the EU's 'Danish Compromise' on bancassurance groups is doing so at the moment.
As businesses, banking and insurance are two great flavours that go well together, like dill and herring. But as accounting systems, they are individually distasteful and even worse in combination, like liquorice and ammonia. This is the root of the problem of regulating bancassurance groups; there is no sensible way to consolidate the sets of accounts without massively distorting the regulatory ratios of one side or the other.
The hardcore, no-compromises approach (and the one required by the Basel Standards) is simply to take the equity of the insurance company and deduct it from the tier one capital of the bank. This makes sure that there is absolutely zero double-counting, but it's very harsh and doesn't really reflect the underlying economic reality. The European approach is more lenient, because it allows the bank to treat its insurance subsidiary as just another risk-weighted asset and hold some capital against it. That's somewhat controversial, but it's the law in Europe and it's not wholly indefensible — the insurance capital doesn't disappear just because the owner is a bank.
It makes quite a difference, in a stylised but reasonably representative pro forma calculation of the total capital ratio of a banking group with a material insurance subsidiary:
But this loophole ended up being a little more generous than anyone had realised. Last year, an obscure posting on the European Banking Authority's Q&A Blog created what the team at Mediobanca called 'The Danish Compromise-Squared' and set in motion a train of events that are now causing a little bit of controversy.
Basically, if you allow the Compromise, then the basis for the capital requirement is the (accounting) book value of the insurance subsidiary. But if you do this, then what happens if the insurance subsidiary itself makes an acquisition? Particularly, what if it acquires a fund manager?
Fund management companies are always difficult for banks to buy, because of what's known as the 'goodwill hit'. The market capitalisation of an asset manager is usually a lot bigger than its tangible book value, because it's made up of intangibles like brands, management contracts, the services of skilled employees, relationships with advisers, and all the other things which make it possible to charge hefty fees for debatable performance.
The difference between tangible asset value and the price paid is recorded as 'goodwill' on the balance sheet, and it's pretty settled regulation that goodwill has to be deducted from a bank's regulatory capital. But it isn't deducted from accounting equity, and accounting equity is the basis of the Danish Compromise.
This makes it much more capital efficient to carry out any acquisitions in this sector through your insurance subsidiary rather than on the balance sheet of the parent bank, if you have previously gained the DC treatment. As the EBA makes clear, there's no real basis to 'look through' the accounts of the subsidiary and pick out things that would be subtracted if they had been acquired in a different way. And this makes a real difference — let's add an acquisition of an asset manager at 3x book value to the picture:
It certainly feels like 'One Weird Trick — Bank Supervisors Hate It', and they do. The ECB never liked the original Danish Compromise, and seemingly likes the extended version even less; in a recent interview, supervisory board chair Claudia Buch said that 'Our interpretation is that it's intended to be applied to the insurance sector and not to, for example, asset management undertakings.' So far the ECB has already told Banco BPM that it is not going to be allowed to use this method for its acquisition of Anima in Italy. BNP Paribas sent out a press release last week saying that as a result of the ECB's recent expressions of opinion, they had updated their internal analysis of the effect of acquiring the AXA IM investment management business to assume that it would have a negative 35bp of capital ratio, rather than the 25bp initially estimated.
But it gets more complicated, because the ECB isn't actually allowed to make policy like this — it's a supervisor, not a regulator. The same interview, Buch admitted that 'this is the role of the European legislators and the European Banking Authority as drafter of technical standards'. Banco BPM even put a question in to the EBA about whether the treatment was allowed, but they rejected it, saying that it would take 'deeper and broader consideration' than they felt able to give in a short time span.
It feels like this is a bit of a regulatory hot potato that nobody wants to catch. And in many ways, the 'DC-Squared' might be quite defensible, because 'goodwill in asset management subsidiaries' is actually quite a high quality intangible, particularly when compared to things like capitalised software development costs. After all, Barclays managed to raise more than $10bn by selling Barclays Global Investors to BlackRock in the absolute teeth of the global financial crisis. Saying that the goodwill is worth literally nothing feels wrong. Read More Inside Apollo's alleged grim-reaper gamble
Even the BNP press release might be a clue that the ECB isn't fully committed to full deduction. The Mediobanca team estimate that if they had to fully deduct goodwill on the AXA transaction, they would be talking about something closer to 65bp of capital impact, rather than the 35bp mentioned in their updated analysis. So perhaps they are anticipating that there is some halfway house to be achieved with the ECB. If that happens, we would be looking at a compromise with respect to the compromise on the Compromise, which surely ought to be some kind of world record.