2020 marks the 10th anniversary of the inaugural vintage of European bank Additional Tier 1 (AT1) contingent capital securities. It is also the first time that this asset class has encountered a global recession as the COVID-19 and oil market crises intensify. Against this backdrop, we assess how banks will use AT1 hybrids during the recession and note a combination of prudential relaxations that--in addition to the fact that many banks are entering this downturn with high common equity Tier 1 (CET1) ratios--will help European banks to continue to pay AT1 coupons. These include maintaining an accommodative monetary and fiscal policy, and swift regulatory actions to cancel dividends and share buybacks and to lower variable remuneration.

That said, the risk of instrument-specific noncall or "extension" and coupon cancellation has increased, particularly for banking systems and banks that have failed to recapitalize over the last decade. This relatively benign base case remains dependent on COVID-19 proving to be a short-term, cyclical crisis where timely and effective fiscal stimulus could significantly mitigate the impact on bank asset quality.

Although we expect that European banks will generally continue to pay AT1 coupons, AT1 hybrids can absorb losses on a going-concern basis, not just in resolution. At this point, we largely expect that banks won't need to consider nonpayment of AT1 coupons, supported by statements such as that from the European Central Bank's (ECB's) Prudential Supervisory Board that it has no plans to suspend payments on bank hybrids. However, while we don't see coupon nonpayment as a tool that banks will generally need to use, it could become more relevant to banks that face increased and sustained pressure on their creditworthiness. We reflect these payment risks by applying a gap of at least four notches between the stand-alone credit profile (SACP) of a bank and the ratings on its AT1 hybrids.

The European AT1 Market Flourished From 2010-2020, But COVID-19 Presents A Test

The regulatory overhaul since the global financial crisis sought to create a bank capital structure that would minimize usage of taxpayers' money and automate early-stage systemic or entity-specific crisis intervention via bail-inable securities. The introduction of the AT1 instrument was a core component of this effort and the global AT1 market has since grown to represent a total principal value in excess of $250 billion. Owing to banks' discretion to stop AT1 coupon payments at any time, AT1 instruments have more flexibility to absorb losses on a going-concern basis than previous bank Tier 1 hybrids. The removal of step-up features, which had a higher impact on servicing costs for banks that chose not to exercise an optional call, is also an important feature. European AT1 hybrids are all contractually perpetual instruments, again to give issuers more flexibility to manage their capitalization in times of stress.

European banks account for about 75% of the 275-plus outstanding securities and, by 2020, most of these banks had reached their targeted regulatory capital ratios following a decade of significant capital building. Until early 2020, banks had executed AT1 primary issuance at record low coupons and near all-time tight reset spreads. Net issuance levels were set to fall over the coming years due to the capital levels that banks had already met, and we expected a modest issuance calendar. We also expected occasional noncall refinancing decisions on account of the dramatically improved primary market access.

Chart 1

Chart 2

The AT1 investor base has widened considerably since 2010, despite these instruments being going-concern, loss-absorbing regulatory capital and perpetual in nature. This growth stems from: the reach-for-yield response to sustained quantitative easing and loose monetary policy, improving bank capital ratios, gradual structural standardization and investor mandate acceptance, and the increase in investment-grade rated AT1 hybrids (although many are rated below 'BBB-'). As the credit cycle turns, so the structural and regulatory nuances assume greater systemic significance for investors and for banks' financial flexibility.

The AT1 market has already hit bumps in the road--notably the 2012 eurozone crisis. However, this came early in the market's development when the range of issuers was smaller. As such, we see the COVID-19 pandemic as a true test for the market. It has abruptly reversed the constructive narrative and forced a dramatic secondary market repricing. The end-of-credit-cycle scenario--with deteriorating asset quality and bank capitalization--is not only testing investor appetite and credit market liquidity, but also establishing a new relative value assessment for bail-inable bank debt within the capital structure. More fundamentally, the new economic environment presents the first true test of regulatory resolution regime enforcement and accounting implementation (IFRS 9) while under intense systemic stress. It also gives insight into when regulators and banks might use AT1 coupons as a "going-concern" form of loss absorption and capital preservation.

Chart 3

Table 1

While a protracted weakening in operating conditions could lead to an increase in AT1 coupon risk, we expect the supportive monetary, fiscal, and regulatory response, combined with widespread cancellation of dividend, share buy-back, and discretionary bonuses, will be sufficient to contain such events to a minority. In the scenario of a prolonged material discount in market pricing of AT1s, we anticipate a material increase in the number of noncall events for European banks, because the economic cost associated with not calling will be less material. These pricing and call decisions are also likely to reinforce each other, leading to more frequent decisions not to call and associated impacts on not only secondary but also primary market pricing.

It remains too early to predict the ultimate impact of the COVID-19 outbreak on European banks' core capitalization. However, we expect widespread cancellation of ordinary dividends and share buybacks, as well as significantly lower variable remuneration in 2020. This is likely to represent a sufficient boost to retained earnings to enable AT1 coupon payments to be paid in full and in time. Regulators across Europe have strongly recommended that banks cancel or defer ordinary dividends, and pause share buyback programs, in order to preserve capital. We expect the vast majority to comply. The ECB has estimated eurozone banks would save around €30 billion of capital.

Re-pricing, Re-distributing

The AT1 market experienced a material improvement in pricing between 2010 to February 2020, reflecting a maturing asset class. A virtuous circle of factors included improving capital ratios and AT1 ratings (see chart 4), institutionalization of the investor base (see chart 5), and falling net supply issuance outlook. The extended period of exceptionally accommodative monetary policy and quantitative easing that relaxed investor perceptions of the key AT1 structural risks--namely perpetual duration, coupon cancellation, and principal write down--also had a positive impact on the above themes. In addition, intense global capital inflows into fixed income (specifically alternative, high yield) manifested in powerful reach-for-yield investor behaviour, supporting AT1 demand.

Average AT1 primary pricing tracked secondary market improvements (albeit less relevant than reset spread for most issuers) and lowered the cost of capital for market-opening country champion banks, paving the way for lower-rated banks and banking systems to follow. We see this in the growing number of issuers in the market and a larger range of min to max coupons priced within each currency in any given year (see charts 6 and 7 below).

Since the start of the COVID-19 outbreak, we have seen an approximate average AT1 PNC5 yield to worst in the secondary market rise sharply from all-time lows (4.5% in euros, 5.5% in U.S. dollars) to all-time highs (14%+ in euros, 12%+ in U.S. dollars) as investors face outflows and reprice the inherently risky going-concern capital structure for the recessionary environment. Many European AT1 securities are now pricing a material risk of a noncall event at the next call date and the inherent negative convexity of the structure has resulted in 20-40 point cash losses in the secondary market in recent weeks. Secondary market spreads are relevant for bank ratings only to the extent that they indicate reduced market access for new financing and we note that a decision not to call a hybrid may, on balance, be beneficial to the issuer in a volatile environment.

The COVID-19 recession and related AT1 market repricing is set to alter AT1 primary supply dynamics. We expect market access to remain difficult for much of 2020, which could result in net negative AT1 supply for the first time, as more issuers decide to extend 2020 call decisions. That said, a rapid return to net positive supply through 2021 and 2022 is likely, given a combination of RWA inflation, widened Article 104 AT1 capacity, and unfilled 1.5% AT1 bucket capacity that forces issuers back to the market as soon as the recovery allows. This dynamic, while tough to quantify at this stage, is likely to contribute to an elevated cost of AT1 capital over our two-year ratings horizon.

Chart 4

Chart 5

The changing nature of the AT1 investor base should not be underestimated. As the product became more broadly accepted, so the investor base became less concentrated in a mix of hedge funds, niche contingent capital-dedicated asset managers, and yield-targeting private banking, comprising more mainstream asset managers, pension, insurance, and sovereign wealth funds. Though this was constructive from a pricing perspective, it potentially raises the risk of pricing contagion across other MREL (minimum requirement for own funds and eligible liabilities) asset classes (such as holdco/nonpreferred senior and Tier 2) in the event of coupon cancellation or principal write-down, although these other instruments are gone-concern instruments that only absorb losses in a resolution or bankruptcy.

Chart 6

Chart 7

Chart 8

Chart 9

Chart 10

Current Market Dynamics Highlight Call Risk

Though still rare in Europe, decisions by issuers such as Santander, Deutsche Bank, and Aareal Bank not to exercise AT1 call options have served as a reminder of the inherent "extension" risk embedded in the AT1 perpetual, callable structure--particularly in a period of economic stress. Although many market participants have recognized that refinancing prior AT1 vintages at ever-tighter reset spreads (and/or lower coupons) would eventually reach a limit at which point a "perpetual vintage" would result, the extent of the secondary market repricing has been exaggerated by the speed and severity of simultaneously increased extension, credit, coupon and arguably principal risk. ("Perpetual vintage" refers to AT1 hybrids that were priced with such low/tight back-end reset spreads that--given deteriorating market conditions--the bank is not incentivized to call for long periods, rather than call and refinance at the first call date.)

We currently consider a combination of factors that could drive increased likelihood of noncalls:

• Effective closure of the market for new issuance, which will likely last until at least the senior nonpreferred and subordinated markets have fully reopened.
• Once the market opens, the risk of persistently higher reset spreads on refinancing that reduces the economic incentive to call 2014-2016 vintages.
• A slight erosion in the historical European issuer view that calls--although optional--are typically exercised for reputational reasons (outside of extreme stress).
• Regulatory comfort with banks that prioritize the retention of capital and avoid unnecessary outflows amid current economic uncertainty. Capital conservation has become a more pressing priority for both bank boards and/or regulators, and such junior-investor friendly behaviour may be harder to justify in the short-to-medium term.

Nevertheless, noncall appetite varies among issuers, not least because they may need or want to return to the market in the coming year, for example, to take advantage of the ECB's early adoption of Article 104, giving an expanded role for AT1 in meeting Pillar 2 requirements. In this respect, it may be easier for strong, repeat issuers to return to the market after a noncall decision than for one-off issuers who lack the depth in investor base. (We note that several banks that have previously decided not to exercise optional calls have afterwards been able to access hybrid markets at accommodating pricing.)

Taken together, we anticipate the emergence of a "perpetual vintage" as noncall decisions become a more frequent occurrence. If so, this phenomenon serves as a reminder of the inherent perpetual risk in the asset class and could curtail further institutionalization of the investor base in the short-to-medium term. It also potentially embeds a greater bifurcation of issuer approach to the call decision-making between repeat issuers that are typically more generous to junior investors outside of stress conditions, and less frequent "one-off" issuers that may prioritize immediate economic incentives over investor friendliness--as is usual in the U.S. market already.

MDA Coupon Nonpayment Risks Have Generally Not Risen, But Remain Important

European regulatory capital frameworks invoke particular nonpayment risks for AT1 hybrids

The EU's CRD IV introduced the concept of Maximum Distributable Amount (MDA). This requires regulators to restrict earnings distribution if a bank's capital falls below the sum of its Pillar 1, Pillar 2, and CRD buffer requirements (see graphic). This threshold is variously known as the overall capital requirement (OCR) or SREP-MDA--the OCR being set as a product of the annual supervisory review and evaluation (SREP) process.

Chart 11

Despite a substantial increase in banks' capitalization in recent years, the final phase-in of the systemic risk buffers (SRB) at end-2019 and the steady rise in countercyclical buffers (CCyB) across Europe has meant that many were closer than ever to their SREP-MDA thresholds just as the COVID-19 pandemic struck. And, for some, headroom was set to narrow further through 2020 due to a planned rise in CCyBs--in many cases to within 100-250 basis points of the SREP-MDA. However, recent regulatory actions are likely to ease this pressure somewhat, at least through 2020.

When setting capital policy, bank management teams tend to be highly mindful of the risk of MDA threshold breach, so typically employ a management buffer of 1% or more, on top of the Pillar 2 guidance (P2G). But, they balance prudence with the imperative to meet expectations around shareholder returns. With RoTEs (returns on total equity) already difficult to achieve in the adverse revenue environment, this task becomes even harder as capital requirements rise, and it incentivizes management to trim further any perceived "excess" capital.

European banks would also be forced to stop payment of AT1 coupons if they have insufficient ADIs (available distributable items). Already a modest risk for most banks, we consider that this has substantially dissipated since 2019 as banks can now use a wider IFRS definition of distributable reserves.

European Regulators Have Walked The Walk

Backed by a policy imperative to expand credit to the real economy during a perceived temporary cyclical stress, European bank supervisors have fulfilled their previous insistence that they would cut some of the combined buffers--mainly the CCyB, as well as aspects of the SRB in some cases--in times of systemic stress. As these buffers have dissolved, MDA headroom has expanded across Europe--based on end-2019 ratios at least. Furthermore, the ECB has brought forward implementation of the proposed relaxation in how banks apply Article 104 CRD--allowing banks in all eurozone countries to count some AT1 and T2 instruments toward their Pillar 2 requirement, often easing banks' CET1 requirements by 1% or more.

It also seems clear that while European regulators have pushed banks hard to cut or defer shareholder distributions, they will not stand in the way of AT1 coupon payments at this time. Andrea Enria, Chairman of the Prudential Supervisory Board of the ECB has stated as much when discussing the actions around ordinary shareholder payments. In our view, this is unsurprising. While banks can stop paying AT1 coupons at any time, and payments will be suspended if banks breach MDA or ADI requirements, there is no clear regulatory incentive to intervene at this time, for the following reasons:

• Coupon cancellation would lead to marginal CET1 preservation relative to the amount saved by cancelling shareholder distributions;
• Regulators are trying to solve a real economy crisis, using the banking system as a key conduit for mitigation. Spurring risk aversion toward the banking system contradicts this aim;
• AT1s are issued to a fixed income investor base for whom there is no upside risk (and where the coupons are noncumulative). While they should, and are available to, absorb losses on a going-concern basis, they would become uninvestable except at extreme costs if coupons were stopped too early in a downturn. Banks and regulators will, in our view, continue to balance their unfettered ability to stop coupon payments with deciding when it is the right time to reflect a deterioration and stop distributions.

Regulatory Relaxations Provide Some Respite, Though This Could Be Temporary

The above moves are generally positive for coupon nonpayment risk--particularly for banks such as Lloyds and Barclays that might have moved within around 100-150bps of their SREP-MDA by end-2020. However, this risk has far from disappeared. First, the buffer and Article 104 easing has an unequal effect across Europe, indeed across the eurozone. Some jurisdictions already applied the softer Article 104 approach, and others had not yet implemented any CCyB. So for banks in those jurisdictions, this regulatory easing has no effect at all. Second, some banks have far more capacity than others to flex their discretionary distributions in the form of equity dividends and bonuses. Finally, as banks move through 2020, the extra headroom might anyway be consumed if underlying capital ratios fall. Again, we expect some CET1 ratios to have fallen and some to have risen at the end of the first quarter (see “European Banks' First-Quarter Results: Many COVID-19 Questions, Few Conclusive Answers,” published April 1, 2020). However, the trend through the rest of 2020 would likely be discernibly downward. At this stage, we do not expect to widen our notching on European bank AT1s beyond the typical four to six notches below each bank's SACP. However, bank SACPs could yet move in some cases, and we do not rule out a widening of the AT1 notching in idiosyncratic cases. We may take into account several factors when assessing MDA headroom to evaluate coupon nonpayment risk in accordance with our hybrid criteria (see box).

What's Next For 2020?

The European AT1 market faces a complex year given the web of regulatory requirements within the EU, let alone across other European jurisdictions, the possibly greater divergent behaviour among European banks around call/noncall decisions, and strong regulatory influence over applicable thresholds and banks' call decisions. Bank disclosures around current and future MDA thresholds, capital policy, and ADIs have gradually improved, but remain variable in their frequency, consistency, and content.

Until the disruption caused by the COVID-19 outbreak, AT1 issuance volumes in January and February were strong. While 2020 AT1 issuance was already likely to be down on previous years, many European banks maintain a strong underlying incentive to come to the markets as they seek to:

• Call and replace early vintages of Basel III AT1 issuance;
• Replace legacy AT1 issuance that will have no regulatory Tier 1 capital value once grandfathering ends in 2021, after exercising calls or tender offers; and
• Fill untapped 1.5% AT1 buckets and take advantage of the widened AT1 (and T2) capacity under Article 104.

On the final point, banks may ultimately find issuing Tier 2 more attractive than AT1 should the current pricing environment persist. We note also that in past years when Tier 1 securities traded at deep discounts, banks have taken the opportunity to effect exchange or tender offers, and so book a capital gain. In the current environment, regulators could balk at signing off tender offers that erode capital buffers. However, if these discounts persist, it is possible that the market could see an outbreak of exchange offers, particularly for legacy instruments.

How long the market interruption lasts is uncertain, but we don't expect to see a broad market opening until at least September 2020. We are unlikely to see mass coupon cancellations or principal write-downs in the AT1 market provided the global recession doesn't persist. However, until investors have a clearer view on the near- and medium-term direction of European bank capital ratios, market access will remain difficult. They will also be mindful that the one-year delay to implementation of the revised Basel rules offers only temporary respite. The recent secondary market reaction has been more general than name-specific, but we expect that those issuers that can demonstrate superior resilience under duress will ultimately achieve differentiation in terms of better access to AT1 financing.

Even in the upside scenario of a rapid global rebound, we expect elevated market sensitivity will result in a higher cost of AT1 market access in the medium term: the ultra-tight spreads seen at the start of the year were already looking unsustainable from the perspective of risk-adjusted returns. Given the weak profitability and margin pressures that many banks face, we assume that they will delay their re-entry to the market in the hope that pricing becomes more attractive in 2021.

Related Criteria

• General Criteria: Hybrid Capital: Methodology And Assumptions, July 1, 2019
• Criteria | Financial Institutions | Banks: Bank Rating Methodology And Assumptions: Additional Loss-Absorbing Capacity, April 27, 2015

Related Research

• European Banks' First-Quarter Results: Many COVID-19 Questions, Few Conclusive Answers, April 1, 2020
• COVID-19 Countermeasures May Contain Damage To Europe's Financial Institutions For Now, March 13, 2020
• Bail-in Debt Remodels The Risk Profile Of Bank Funding In Europe And North America, Sept. 26, 2019
• The Resolution Story For Europe's Banks: Life In The Halfway House, July 18, 2019
• EU Banking Reform Package: Enhanced Balance Sheets, Incomplete Market Integration, June 18, 2019

Appendix/Charts:

Table 1

Table 2