AT1 rush shows bank investors have short memories

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

People buying hotcakes. Monkeys grabbing bananas. Pick your metaphor: fixed-income investors cannot get enough of the niche bank capital instruments called AT1s, or additional tier one bonds.

Eight months ago, when Swiss giant Credit Suisse failed and regulators wiped out all $17bn of its AT1 bonds, the prophecies were clear: the AT1 market — previously seen by investors as a high-return, low-ish risk niche that gave banks an extra buffer of capital — was done for.

With the Swiss authorities having apparently acted on a whim to ease Credit Suisse’s forced $3.2bn sale to arch-rival UBS, faith in the investability of the product evaporated. This was not only because of the dramatic wipeout itself but also because there was an inversion of the established hierarchy that would normally see equity investors punished first. A swath of legal action against the Swiss authorities is ongoing. “This market is dead,” one Swiss financier told me at the time. “No investor will trust an AT1 issue ever again.”

“Ever” for bank investors, it turns out, is a very short time. This month, in an epic plot twist, UBS itself issued a substantial $3.5bn of AT1s. They will initially look very like the old Credit Suisse bonds, though UBS says the terms will be amended once shareholders agree. It says that rather than being wiped out in a crisis situation, they would convert into equity, bringing Swiss AT1s into line with the norm elsewhere.

The UBS issue commanded extraordinary demand. The order book topped $36bn, or 10 times the ultimate issue size, as investors queued up for the 9.25 per cent coupon. Other banks are capitalising on the fervour: last week Barclays began marketing a new AT1, again to a rapturous reception.

Investors say they now regard the Credit Suisse case as idiosyncratic. This is partly because of the nature of Credit Suisse’s demise — it suffered a severe loss of client confidence, and a rapid withdrawal of customer funds. It is also partly the result of the Swiss context. Legal terms for AT1s and the application of them differed from those elsewhere.

“We can’t get enough AT1s now,” says an executive at one large asset manager. The basic logic is sound. The macro environment might be tricky with geopolitical turbulence, low economic growth and higher interest rates that are likely to trigger more loan defaults. However, the banking system looks strong, underpinned by capital cushions that are a multiple higher than they were when the global crisis hit in 2007-08.

It will be an irony not lost on European bank bosses, such as Barclays’ CS Venkatakrishnan, that the hotcake demand for its AT1s is in sharp contrast to the weak demand for its equity.

But do AT1s fulfil their planned regulatory purpose? They did, in a way, in the Credit Suisse instance: for all the controversy, $17bn of funds was recovered from the bank’s implosion that in turn helped facilitate the rescue deal with UBS and mitigated taxpayer risk.

For all kinds of reasons, including the quirks of the Swiss context, they did not really work as global regulators intended, though. AT1s, also known as contingent convertible or “coco” bonds, were originally designed to recapitalise a bank that had been dented by big losses. In Credit Suisse’s case, losses had yet to materialise and capital ratios were strong. Policymakers had underestimated the interaction between customer confidence, liquidity and viability.

The Swiss National Bank concluded in its annual financial stability report a few months after the Credit Suisse debacle that its AT1s were “not effective”. Coupons could have been withheld but were not, in part because that would have highlighted fears about poor performance. Similar vicious-circle logic might apply if an equity conversion was triggered, leading to a stock market panic. Regulators elsewhere now admit privately that AT1s may not actually be fit for purpose at all.

In the US, AT1s in the European sense do not exist, though more straightforward preference shares — which avoid the complex triggers and Catch-22s of cocos — do. What happens next may be down to the Basel Committee of global regulators. Last year, it concluded in a report that “robust empirical conclusions regarding the loss-absorption capacity of AT1 instruments cannot be drawn at this stage”.

If, in light of this year’s events, it takes a negative view of AT1s as a viable capital instrument for the long term, there may not need to be another Credit Suisse-style event to snuff out today’s market exuberance.

patrick.jenkins@ft.com