Novel bank capital instruments introduced as loss-buffers in the wake of the global financial crisis (GFC) may have hit a wall as European regulators contemplate new ways to shore-up the monetary system.
Indeed, proposals flagged by the European Central Bank (ECB) this April could push the much-misunderstood Additional Tier 1 (AT1) securities – established as a real ‘thing’ only in 2014 – into the too-hard basket for most institutional investors.
Included as part of the ECB’s wide-ranging response to the earlier European Commission’s ‘Call For Advice’ (CFA) on a review of the macroprudential framework in the region, the mooted AT1 reforms do seem to border on the edge of impossibility.
But, as quantum mechanics teaches us, nothing is impossible given the weird and wonderful forces at play in the sub-atomic world; a mathematics-defined universe that is almost as difficult to understand as post-GFC bank capital rules.
In the large-scale reality of classical physics, for instance, balls bounce off walls rather than mysteriously appear on the other side. Barriers, however, are not so clear-cut for quantum particles that are governed by probabilities rather than absolutes.
Based on well-known quantum principles, and indisputable observational evidence, scientists know that sub-atomic ‘balls’ don’t always bounce – sometimes, contradicting commonsense, they jump the impassable barrier in a process known as ‘quantum tunnelling’ (see diagram below).
Figure A : Quantum tunnelling
Quantum tunnelling is more than just a theoretical quirk with the phenomenon responsible for a wide range of physical effects – such as radioactive decay – while also showing up in technology including computer chips, superconductors and microscopic scanning devices.
The GFC was an historic test of survival for the the global banking system: somehow, contradicting commonsense, it got to the other side.
As regulators absorbed lessons from the financial crisis, though, they came to the conclusion that banks needed better ways to cushion against impossible losses rather than relying on taxpayer bailouts and central bank largesse.
During the GFC, regulatory authorities found it particularly difficult to force losses through existing hybrid subordinated debt structures, hindering bank recapitalisation plans. At the time, only a handful of distressed banks suspended coupons on such debt securities with institutions loathe to put investors offside.
More typically, banks cushioned the GFC shock by borrowing through central bank ‘windows’, a series of ‘unofficial’ liability management exercises with governments and discounted share buybacks.
The compromised, and deeply unpopular bank rescue missions, encouraged regulators to design a fairer risk-sharing system that would contain the pain within institutions, cascading losses back through the capital stack including holders of hybrid debt. Out of the chaos, regulators settled on contingent convertible bonds (CoCos), of which AT1 debt is the most common variant, as an important component of the loss-sharing equation.
Since the first issuance in 2014, the AT1 debt market has already suffered through a few real-life ‘stress tests’ including the oil price collapse, a mini taper tantrum in late 2018, the Covid-19 pandemic, the current rate-tightening super-cycle, and last, but not least, the Ukrainian war.
For some time, though, the ECB has raised concerns that the AT1 asset class was not functioning to specification amid overly complex bank capital structures and the marked reluctance of management to stop coupon payments on the debt securities for fear of alienating investors.
The recent ECB submission, discussed above, highlights previously identified ‘flaws’ in the design that deter banks from using such “non-releasable buffers” for additional lending as intended in the original conception of AT1 assets as “going concern capital”.
“… there is a need to strengthen the ability of AT1 to act as going concern capital, specifically regarding the flexibility of payments,” the ECB submission says.
“Improvements could also be made to loss absorption in the going concern perspective (e.g. requiring accounting classification as equity, removing the need for obsolete automatic triggers) and permanence (e.g. by limiting the possibility to call the instrument only if replaced with a CET1¹ instrument or a cheaper AT1 instrument).”
Figure 1: Flaws in the design?
Reinforcing the permanence of AT1s in the capital stack as suggested above would add more rigidity to the current call regime and potentially increase the underlying duration of the asset class as investors price in higher extension risk. All AT1 redemptions already require the ECB approval.
Furthermore, the ECB proposes only allowing profitable banks (or those with “positive retained earnings”) to keep making AT1 or CET1 dividend payments.
Linking coupon payments to bank profitability sounds good in theory; possibly horrible in practice.
It’s questionable, anyway, whether AT1 distributions can be switched off quickly enough during times of need to help shoulder capital stress.
For instance, during the first year of the Covid-19 pandemic, cancelling AT1 coupons would’ve only generated an average capital benefit to banks of less than 20 basis points – a much smaller bang-for-your-buck than the widely introduced dividend ban. Importantly, too, the dividend ban proved to be a temporary setback as most banks have since caught up but an all-in cancellation of AT1 payments would have led to a permanent loss for investors as the coupons are non-cumulative.
Bounce-through: the story of the other side
While regulators and industry debate its future, the AT1 market is currently feeling the effects of multiple financial and geopolitical stresses: rising rates, inflation scares, quantitative tightening, the fallout from the Ukraine conflict and the threat of a looming recession have all conspired to drag the asset class into negative territory over the year-to-date.
Figure 2: Coco Index quarterly return
With both regulatory and market risks mounting on the AT1 market investors may demand a higher coupon on the assets; large banks will likely respond by simply holding more CET1 assets (depressing future return-on-equity expectations) while smaller or non-listed banks are effectively squeezed out of business.
Overall, a shrinking AT1 market would tie up more bank capital in CET1 assets and lower lending – hardly an inspiring result for the eurozone.
And yet there is cause for optimism among the barrage of bad news for the sector.
Firstly, higher rates and steeper curves are eventually conducive to the business of banks, providing a rising floor of support for AT1 assets.
Regulators, as well, should be aware of the risk that across-the-board immediate changes to AT1 trigger-level or deferral language could destabilise the market. We expect any AT1 legislation and ensuing regulations will include a grandfathering period – most likely rolling on until the first call date – similar to the example set by the Swiss banking regulators which introdued extensive grandfathering proposals for existing high-trigger Tier 2 and low-trigger capital in 2015 to smooth the transition to a simplified capital structure featuring only high-trigger AT1s².
Despite all the apparent barriers, the AT1 market could well be poised for a bounceback – or possibly even a bounce through to the other side just where investors, issuers and regulators need it to be.
Available Distributable Items (ADI): AT1 coupons can only be paid to the extent that the issuer has available distributable reserves to do it.
Maximum Distributable Amount (MDA): the concept is easy-to-understand: as capital levels decline, then distributions should be restricted too. We discussed MDA here: https://www.hermes-investment.com/de/insight/fiorino/fiorino-bank-investors-mda/
Point of No Viability (PONV): the point at which a bank is seized by the regulator and resolved. PONV requires all AT1 and Tier 2 instruments to be capable of being converted into common equity or written off.
Going Concern Capital: equity (CET1) and T1 capital
Gone Concern Capital: Tier 2 (T2) capital, absorbs losses when bank hits the PONV.