The additional Tier 1 asset class created by regulators became a darling of fixed-income investors, but a controversial move by one of Europe’s largest banks has left questions over whether the instrument is doing the job it was designed for. Elizabeth Pfeuti reports.
In a paper published in September 2013, the Bank for International Settlements explained how investor reluctance to lend to banks at a time of financial stress had led to the birth and continuing development of the contingent capital instrument.
While contingent convertible bonds (cocos) themselves had been around since the currents of the financial crisis threatened to pull down more than one poorly capitalised bank, the arrival of the additional Tier 1 (AT1) instrument in 2013 was something of a game changer.
Structured as a hybrid, the AT1 was designed to suck up first losses while keeping the issuing bank afloat. For those still shaken by the crisis, it gave some reassurance regarding the stability of the sector, and was a cheaper option for a bank than issuing equity. However, despite offering an eye-catching return, it was not an instant hit with investors.
“The AT1 took some time to resonate before it gained a broad investor base,” says Gary Kirk, a former banker and now partner at bond house TwentyFour Asset Management. “It started with the hedge fund community and some of the slightly more entrepreneurial parts of the asset management industry. There was certainly a complexity premium associated with the product initially.”
Mr Kirk says investors were concerned about the trigger into equity or full write-down, and whether they could accept that degree of optionality at all. Because of its structure, the AT1 did not fall into traditional credit indices, making it a tricky buy for many mainstream fund managers.
A new balance
This paucity of investors left issuers with the job of making the deal attractive, and AT1s have been commonly issued with yields in excess of 6%. In 2014, some $116bn was issued through 99 deals, according to Dealogic, with an additional $137bn the following year. Fast-forward to 2019, and after putting up with a decade of low interest rates, the coco investor community has broadened while the bulk of first issuances have been dealt.
“The AT1 has been an off-market purchase for a lot of investors,” says James Macdonald, partner and financials analyst on BlueBay Asset Management’s top decile cocos fund. “But it very much feels like the demand-supply dynamic is finding a bit more balance.” The yields in this asset class are still extremely attractive: in February 2019, UBS gathered $2.5bn for a five-year non-call AT1 at 7%.
“You are buying investment grade-rated issuers with yield in excess of what you can get in high yield, and because of this, the investor base is quite wide and still growing, so liquidity also tends to be pretty good,” says Mr Macdonald.
By the end of April 2019, almost $530bn had been issued by global banks, with close to $336bn outstanding, Dealogic figures show.
Along with plenty of other behind-the-scenes operations to control internal risk levels, this issuance has helped banks increase the quality and quantity of their capital, making them a much more sound investment. Improvements in transparency have also helped banks meet due diligence requirements, according to Mr Kirk, who says there is more clarity on what is on offer.
Save the date
“For many investors looking at the risk-reward available in various securities, moving down the balance sheet into AT1 securities is worthy of consideration,” says Mark Geller, head of the financial institutions group for debt capital markets for Europe, the Middle East and Africa at Barclays. “The chance to buy an AT1 security from a large, well-capitalised, well-known, regulated entity is something that investors have focused on as the market for cocos has developed.”
Mr Kirk adds that the investor community has learned to deal with the idiosyncrasies of the AT1 instrument, with private banks, traditional asset managers and insurance companies among the buyers. “Now the main concern is whether or not bonds extend beyond their first call date,” he says.
This issue was brought into sharp focus when Spanish banking giant Santander decided to not call one of its AT1s in mid-February 2019. Despite being issued as a perpetual instrument, it had been generally accepted that an AT1 issuer would call the bond on the first available date. This meant investors sensitive to duration could put the bonds in their portfolios and assume they were buying short-dated, fixed-maturity instruments.
“As a fixed-income investor, you want two basic things: get your coupon when you are told you will, and get your principal back when you are told you will. The more fixed income-like AT1s are, the more attractive level of spread they will be assigned,” says BlueBay’s Mr Macdonald.
BlueBay, like many others, has a strong incentive to buy bonds that are only callable once every five years, and Mr Macdonald thinks banks should be rewarded for issuing instruments structured in this way. “It gives you at least some certainty for what the duration of the security is going to be,” he says. “If it is callable quarterly, you’re effectively selling call options on the issuer.”
Policy change
But for Paul Deakins, partner at Clifford Chance, Santander’s decision was not surprising, given banks’ necessary strict adherence to Basel III.
“Look at the policy change behind the Capital Requirements Directive IV. Part of it was very clearly designed to get rid of this gentleman’s agreement,” says Mr Deakins. “It was simply something regulators were not comfortable with, looking with hindsight at the financial crisis, and it is now purely an economic decision for the institution in question. Investors should not go into it with any sort of expectation that they would be called on the first opportunity.”
However, some investors viewed Santander’s decision as “pretty bad form”, according to TwentyFour’s Mr Kirk, saying the bank could have done more to steer the market beforehand and that its failure to do so might come back to haunt it.
“I imagine there will be a degree of premium associated with Santander if they rely purely on marginal economics in their decisions to call or extend,” says Mr Kirk. “A lot of the banks we have spoken to say that, within reason, even if it is economically not viable to call, they will still do so, as long as their capital source is still reasonably healthy, which most of them are.”
Mr Kirk cited a move to refinance an existing AT1 by the UK’s Coventry Building Society in March, which he says investors welcomed. The society announced it would tender its existing AT1 at a 2.25% premium to par eight months early, but issue another at a slightly more attractive yield. TwentyFour was a considerable investor in the deal.
“Market participants viewed this transaction as a very smart and efficient way to refinance the deal, and as a result investors can be expected to continue to support this borrower,” says Mr Kirk.
Pricing is key
Even so, Mr Geller at Barclays notes that the market reaction to Santander has been relatively muted, with some significant deals being carried out in the weeks after its decision to extend.
“It is clear some investors would prefer there to be more duration in order to get away from multiple call situations and also to be able to hedge the security efficiently,” he says. “Ultimately though, the value relating to the frequency of the call options can be priced and what is crucially important is, of course, also pricing the front end appropriately.”
This pricing is not yet quite as appropriate as some would like, however, and this could be due to the relative immaturity of the asset class.
“Banks are beginning to listen to us on this, but the difficulty they have is that no real pricing differential historically has been shown in the market between those issuers who have gone down the route of structuring their bonds in a more investor-friendly manner versus those that have not,” says Mr Macdonald. “But the pricing differentials are beginning to show.”
In the months since the Santander decision, its AT1 has been trading slightly below par, according to Mr Kirk, because – now callable every three months – many expect it to be called in fairly short order.
“Had this bond been callable only every five years, as some are structured, then we might have expected to see a significant drop in asset price of that bond,” says Mr Kirk. “The investor holds the cards, because if they don’t like the structure of the deal, they won’t buy it. We are very selective so will always expect a bit of a premium on issuers such as Santander given the lack of clarity.”
Shake-up coming
For Mr Deakins, there is a trade-off between providing investor certainty and preserving flexibility for the issuer. “There ought to be a pricing differential between those two options for one and the same credit,” he says. “And basically, issuers will take the discretion they need, if it comes at a price that is affordable.”
The bigger concern for Mr Deakins is the long-term viability of the AT1 instrument at all. “Adapting to the Capital Requirements Directive V banking reform package, when it gets published later in the year, is going to affect the AT1 instrument,” he says. “The reform is going to require a more extensive disclosure exercise in terms of giving more granular details on how banks’ capital structures are put together, and therefore allowing people to infer from that what the risks are in terms of coupon deferral or decisions not to call a particular instrument.”
This could shake up the sector, which has become, if not complacent, then relatively at ease with the level of risk investors are taking with the AT1.
“We learned from the last crisis that having lots of capital-absorbing losses on insolvency is only any good if you’re prepared to put those institutions into insolvency, which, of course, has some fundamental problems for preserving the critical functions they perform,” says Mr Deakins.
For many investors, the implicit risk in the AT1 is likely never to materialise, despite yielding them a very handsome coupon of up to 8% or more for certain issuers. The in-built triggers of 5.125% and 7% were relevant when bank capital ratios were at 6%. Now they are over 14%.
“The likelihood of you ever reaching that trigger point is so remote that the most likely scenario is that well before you hit it, the regulator has to intervene and apply losses across the bank capital structure,” says Mr Macdonald. In which case, 1.5% of AT1 is very unlikely to be enough to stabilise the situation.
“We think you are being massively overcompensated for the risk, because what the securities were designed to do, they aren’t really doing in this changed environment,” says Mr Macdonald. “Maybe sometime in the future, the regulator goes back and looks at whether these AT1s are doing the job they were initially designed to do, and at that point their design is tweaked.”
Testing needed
Andrea Colombo, head of investment-grade finance for southern Europe, Belgium, France and Luxembourg at JPMorgan, urges investors not to overlook the risk, especially as the sector has not been thoroughly tested. “When you compare the interest rate you get on these instruments to what you get in corporate hybrids, corporates pay considerably less for a comparable rating. At the same time, the AT1 is materially riskier than a corporate hybrid,” he says.
Along with the loss absorption, banks can stop coupons in time of distress, but Mr Macdonald says this would be a very last resort as it would lead to wider ‘bank run’ issues. “These instruments are more akin to leveraged finance instruments than corporate hybrids,” adds Mr Colombo. “You get a good reward, but you shouldn’t underestimate the risk. Although the market is in a good place right now, and there is potentially some upside compared with the last couple of years, things can change quickly.”
Mr Kirk agrees. “There will be a degree of pushback if issuers pull down rates: you are still at the bottom of the capital structure. There is also the ongoing optionality embedded in this product so there will always have to be some premium associated,” he says. “And until we’ve been through a proper cycle and see how these things evolve, there is always going to be some degree of volatility associated with this product as it is part of the capital structure.”
But if a significant event does happen to a bank or there is a wider downturn, the structure of the AT1 might be its undoing. Many investors may be forced to sell their entire holding to get to a neutral position as they are not part of any traditional fixed-income indices. For now, though, with the flurry around Santander’s extension in the rear-view mirror, the AT1 market is steadily moving along.
“We are seeing issuers funding or part-funding the calls they need to make in the next 12 months,” says Mr Colombo, because banks want to limit refinancing risks. This means up to Ä20bn of AT1 in Europe may be called over the next year, plus whatever needs topping up and refinancing.
Mr Macdonald says: “We believe that in the event of a downturn, banks will perform much better fundamentally than they have done historically, but you probably won’t know that until post-event.” Bankers, issuers and investors are therefore awaiting regulators’ next move to see if any tinkering is needed to the AT1 and how it would affect this favoured non-fixed income security.
