There is a sense of calm in the banking sector worldwide. The crisis, initiated by the failures of US banks and evident in Europe through the Credit Suisse takeover, has come to an end. However, the shock still lingers in the industry and among its regulators. How resilient is the banking sector in the face of new macroeconomic turbulence?

To answer this question, experts are focusing on specific contingent convertible bonds called AT1. Opinions vary on whether these contingent convertible bonds, with a total issuance volume of several hundred billion euros globally, truly enhance the resilience of banks or could turn out to be accelerants of crisis.

AT1 bonds are a product of the 2007/2008 financial crisis. The idea, proposed by Anglo-American economists from the so-called Squam Lake Group, was to strengthen banks' equity capital by combining the characteristics of debt financing with the ability to absorb losses – well in advance of a crisis, namely in the normal course of business operations. It is not only envisaged that coupon payments can be suspended, but banks must convert them into equity or write them off if a certain core capital ratio is not met.

From a regulatory perspective, these bonds can be considered additional core capital – the abbreviation AT1 comes from the English term 'Additional Tier 1,' under which these bonds are classified. Bonds issued by insurers fall under the abbreviation RT1 (Restricted Tier 1 instruments). The appeal for issuers lies in the fact that these bonds can be issued more cost-effectively than equity capital.

Contingent Convertible Bonds (Cocos) came into the public spotlight once again when Credit Suisse teetered on the brink at the turn of 2022/2023. The situation revealed two issues currently under expert discussion.

Firstly, the Swiss Financial Market Supervisory Authority forced the bank to completely write off its AT1 bonds with a nominal value of 16 billion CHF. Meanwhile, shareholders were compensated at least partially, a matter that will occupy Swiss courts for a long time. There are already hundreds of lawsuits from thousands of investors.

Secondly, experts are concerned about whether the supervisory authorities have set the triggers for a write-down or conversion of the bonds so low that a bank is already insolvent when investors are asked to pay. In an environment where the capital market demands core capital ratios in the low double-digit percentage range from banks, regulatory triggers in single-digit percentages are seen as unrealistic.

Despite these critical points, there is currently no real approach to make AT1 instruments even more loss-absorbing – as this would likely deter investor interest. Thus, the market for Cocos has regained momentum. For example, Sumitomo Mitsui Financial Group placed 1.0 billion euros in April, DZ Bank raised 1.1 billion euros in May, BNP Paribas recently entered the market with 1.5 billion dollars, and Spanish BBVA reported demand of over 3 billion euros for its AT1 bond in June. Even UBS, as the acquirer of Credit Suisse with an outstanding AT1 volume of more than 11 billion dollars, could – with improved conditions – enter the AT1 market.

Yet a seemingly minor aspect plays a crucial role: these bonds theoretically have perpetual durations. In practice, investors demand early redemption from issuers – a demand that banks and insurers often comply with. The core question from their perspective is: How much disappointment can we inflict on investors without losing market trust?

The loss of trust can be fatal, a well-known fact in the capital markets. But what does it mean for AT1 issuers? For example, if they suspend coupon payments, the bonds function as intended by regulators: as loss-absorbing instruments ensuring the ability to manage the business (going-concern instrument).

However, the theoretically usable instrument becomes worthless in practice if suspending payments leads to such a massive reputation damage and crisis of confidence that all bonds of the issuer plummet in the capital market. Investors could assume that the management asymmetrically possesses more negative information, suggesting a far worse situation than publicly visible. In this scenario, hardly any bank would dare to utilize this measure.

One might argue, well, this is a rare case, and besides, banking regulators have a say in it. True. Nevertheless, it requires careful consideration. Moreover, the redemption dates of AT1 bonds, known as RT1 in the insurance sector, work with the same mechanism. What's behind this?

Cocos (Contingent convertible bonds) are theoretically perpetual. This is a prerequisite for their recognition as regulatory AT1 capital. However, the terms of the bonds allow issuers to redeem them after a few years. The decision lies solely with the banks and insurers. In theory. In practice, market participants report a different story.

The pressure on issuers to redeem is enormous in Europe, unlike in the United States. 'One of our competitors recently redeemed their AT1 bonds,' is the message delivered in investor conferences by banks and insurers. This sets a precedent: 'You surely will do the same.'

If someone argues that the instrument is intended differently, that may ensue great agitation. The implication is clear: if a redemption is refused, the entire market presence of the issuer will be under pressure. The conclusion: 'If you are a moderately sized issuer, you quickly get scared and refrain from redeeming the bond.'

Investors have a significant influence. While there are highly specialized vehicles in the AT1 and RT1 markets, often managed by hedge funds from the UK, these funds have limited participation in senior bonds. However, in these segments, they do have influence, posing an immediate threat.

Nevertheless, banks, as continuous issuers in the wholesale funding market, repeatedly need to raise funds, and AT1 specialized funds are sometimes part of larger investor groups – which might also invest in other types of bank bonds. Moreover, a dispute with AT1 investors can harm an issuer's overall reputation.

By choosing to redeem their AT1 bonds, issuers signal: 'Dear credit investors, you matter to us.' The simple but fundamental credo among issuers is: 'We know people have memories.'

Also, relevant from a critic's perspective: major institutions like UBS are significant underwriters of Tier1 capital and thus have specific business interests. Furthermore, acting as an investment bank for others, they would think twice before diminishing their substantial fee income by upsetting AT1 buyers, who might also be their investment banking clients.

Why are investors so eager for redemptions to happen? One might think it's the prospect of an interest rate hike leading to a higher coupon rate for the new bonds. However, this is irrelevant because for perpetual bonds, the coupon rate is frequently adjusted to the current interest rate landscape at the first redemption date. Banks, on the other side, have hedged their positions. What matters is the spread above the mid-swap rate. These spreads were extremely low in the past and have risen since last year – not dramatically, but quite substantially.

As a result, investors receive a higher spread and, therefore, more money overall when the bond is redeemed and replaced by a new AT1 bond. Yet, issuers want to replace their AT1 bonds only if this reset spread is lower than the historical spread of their old bonds.

Additionally, from an investor's perspective, when the issuer redeems the bond, the value, which may have been below 100% before, bounces back to 100%. If the redemption date passes, the bond might drop toward 90% on the stock exchange – leading to uncomfortable questions from fund managers, especially from their own bosses.

How are banks positioning themselves amidst the conflict between economic considerations and investor pressure? UBS made headlines at the turn of the year. They redeemed a $2 billion AT1 bond issued in 2018 in January 2023, the earliest possible date. Market calculations show that the institution will have to pay an additional 2 percentage points: over a five-year period, this amounts to $200 million.

While this is an extreme case given the substantial amount calculated there, many other institutions have also chosen to redeem this year. The list includes Société Générale, Lloyds Bank, Barclays, Banco de Sabadell, and Unicredit.

Critics argue that this is a disaster. They point out that before the 2007/2008 financial crisis, the same practice was established with similar bonds called Innovative Tier1. Redemptions were common. 'Who would have dared to deviate from this norm back then in the really severe crisis?' It was very naive to think that a treasurer would do so. Regulators were forced to let this happen: 'If they had issued a ban, they would have risked a bank run.'

Nevertheless, this is currently not a problem, according to defenders of the redemptions. It is true that some institutions pay a higher reset spread: 'But these are usually increases in the range of only 50 basis points.' Moreover, it is regular practice that banks do not redeem if the increases are too significant. With triple-digit increases, intensive discussions with regulators are to be expected. In short: 'The market has proven itself over time.'

Additionally, defenders of redemptions argue that many institutions do not redeem even when possible. This includes names like Santander, Deutsche Bank, Raiffeisen Bank, and, in one case, Lloyds Bank.

This is not surprising, say the critics of redemptions. Santander traditionally stands strong against investors. However, mid-sized institutions cannot withstand the pressure – even in times without turbulence in the banking market. Their anxious question: What can be expected when a crisis rages and redemption dates approach?

In 2016, it was demonstrated that even a giant like Deutsche Bank could become indecisive. At that time, the Frankfurt-based bank offered to buy back bonds worth €3 billion and eventually repurchased €1.3 billion, merely to demonstrate action to investors and dispel doubts regarding its liquidity adequacy.

This problem can only be circumvented, as per critics of redemptions, if the decision for or against a call is derived from a rule – in other words, if it is deindividualized: 'The decision not to redeem should only signal that the replacement conditions are poor.' For example, it could be stipulated that the reset spread should not be more than a certain percentage higher than the spread of the old bonds – occasionally, a 0.25% increment is suggested, while in some cases, an increase in the Eurozone of no more than 1 percentage point is cited.

Such a reform is far from straightforward. Banks want to maintain their discretion. Moreover, they can threaten politics and even banking regulators, stating that they will restrict their lending because they will have to replace AT1 bonds with even more expensive equity. ECB banking regulators can only prohibit redemption if the stability of the entire institution is at risk – given the limited size of an AT1 bond in relation to a bank's total capital, this scenario is rather unlikely.