AT1 bonds as a fire accelerant

The Federal Council is expected to publish its dispatch on the new capital rules for submission to Parliament in April. A purported ‘compromise proposal’ is already being circulated by National Councillor and banker Thomas Matter, which would allow UBS to provide up to 50% of the financing for its foreign subsidiaries in the form of AT1 bonds instead of hard equity, as the Federal Council demands. As the experience with Credit Suisse has shown, the idea is likely to backfire.

The report by the Parliamentary Investigation Committee indicates that CS was already in dire straits by the end of 2022 at the latest and – due partly to the relief granted by the Financial Market Supervisory Authority (FINMA) – was undercapitalised as well. By then at the latest, the AT1 bonds should have been converted into equity, as confirmed by both the plummeting CS share price and its soaring credit risk premiums. UBS also came to the conclusion that CS was ultimately lacking in equity capital. Otherwise it would not have urged the authorities to write off the AT1 bonds during the takeover of its competitor.

This brings us to the crux of the matter: why were the AT1 bonds not written off early on, even though – viewed in the cool light of day – the authorities and CS managers had, from autumn 2022, gradually come to the conclusion that the bank urgently needed additional equity to counteract the erosion of trust among clients?

The honest answer is probably this: writing off the bonds would likely have fanned the flames of uncertainty among investors and most certainly have exacerbated the bank run, leading to the demise of CS. The bank’s executives and the authorities had to assume that an AT1 event would trigger panic on the markets – a classic vicious circle. The Federal Council, FINMA and the SNB therefore felt compelled to keep on repeating, until the very end, that the bank was facing neither a capital nor a liquidity problem. AT1 bonds are essentially intended as shock absorbers. But they can fan the flames when crisis strikes. Everyone in a position of authority must have realised this truth. And this is what prevented a write-off to stabilise CS prior to its collapse, as had originally been planned but not implemented until afterwards.

Given the questionable communication practices by the authorities and bank management in the phase leading up to the demise of CS, the Federal Administrative Court issued a decision in October 2025 dismissing the AT1 bond write-off as unlawful. Although the judges’ decision was a logical response to serious regulatory failings, it might ultimately prove to be an own goal costing Swiss taxpayers billions. It follows, therefore, that a significant hike in equity capital would be the only effective lever to avert damage to taxpayers in the event of UBS going under. The combined balance sheet of UBS is equivalent to 200% of Swiss GDP, which means that no other country in the world is as exposed to a similar risk from a single bank. What is more, it is contrary to international practice for the equity of UBS's foreign subsidiaries to be financed in part with debt capital from the parent company. It would be equally absurd if a private individual applying for a mortgage with UBS were to finance the equity required for the purchase with a loan from another bank (double leverage).

Do higher capital requirements threaten UBS's competitiveness? It is astonishing to note how this fear has gained widespread traction, as there is no scientific or empirical evidence to support the theory. Many cantonal banks have been operating successfully with high equity ratios for decades. If it were true that high equity ratios actually jeopardise competitiveness, UBS would have to – by inverse logic – prioritise awarding loans to clients with little equity, which is a rather bizarre and impractical notion.

Never in the history of the big banks have equity ratios been as low as during the last twenty years. If we follow the logic of UBS, this should have been a blessing for the bank's competitiveness and its shareholders. Even a cursory review shows that the opposite is actually true.

While the Swiss market as a whole (SPI) generated a cumulative return of 217% (5.9% p.a.) between 2006 and 2025, UBS's equity investors, including all dividends, achieved a cumulative return of just 10% (0.5% p.a.). Low equity ratios are evidently no guarantee of attractive returns for shareholders, but only for salaried bonus recipients at the top. While shareholders' equity has cost UBS practically nothing for over two decades, the annual interest rate on AT1 bonds has ranged between 6 and 9%. The statement that AT1 bonds are more favourable than equity should therefore be viewed with quite some scepticism based on empirical evidence alone.

What have we learned from the UBS and CS cases? Low equity ratios increase a bank's refinancing costs, jeopardise its competitiveness and creditworthiness, erode confidence among clients and investors in times of crisis and ultimately lead to a bank run. Banks also preach these accepted wisdoms – which obviously apply to all companies – to clients when they apply for a loan.

UBS's hard equity currently stands at 4.4%. If the demand of the Federal Council is implemented, the quota will rise towards 6% – so essentially a marginal amount. This would also be achievable within ten years simply by reducing or suspending dividend payments and share buy-backs. If the Federal Council had agreed with the pre-eminent professors Anat Admati, Martin Hellwig and Simon Johnson by demanding 20 to 30% hard equity, an increase of a whopping $300bn would have been necessary instead of ‘just’ $25bn to $40bn. As a result, the ‘extreme solution’ demanded by the Federal Council would seem almost like a moderate compromise proposal.

Switzerland has seen two of its largest banks collapse in the space of just fifteen years. UBS bosses assure us that this will never happen again. Really? Let us not forget the famous words of Albert Einstein: “The definition of insanity is doing the same thing over and over again and expecting different results.”