AT1 Bonds and Bail-In Risk: What Every Saver Should Know

A Bond That Can Vanish

Imagine lending money to a bank, collecting a handsome interest rate for years, and then discovering that in a crisis your loan can simply be erased — or swapped for shares — while the bank's ordinary shareholders keep something. That is not a hypothetical. It is the design of a specific kind of bank bond, and understanding it tells you a lot about who really carries risk in the financial system.

These instruments are called contingent convertible bonds, or "CoCos." A CoCo is a bond that behaves normally most of the time but can convert into shares or be written down if a pre-agreed trigger fires. The word "contingent" simply means "depends on a condition happening." The most common variety in Europe is the Additional Tier 1 bond, usually shortened to AT1.

Think of it as a safety valve welded onto a bank's balance sheet. When pressure builds, the valve releases — and the people holding those bonds are the ones who feel the steam.

Why Regulators Wanted Them

To understand AT1s, you have to go back to 2008. During the global financial crisis, governments poured public money into failing banks to keep them standing. Investors in bank capital instruments were often protected alongside depositors, which struck many observers as unfair: the people who took the riskiest positions got rescued by taxpayers.

Afterward, the Basel Committee on Banking Supervision published the Basel III reforms in December 2010. The goal was to strengthen both the quality and quantity of bank capital, and to make clear who absorbs losses when things go wrong.

Basel III sorted regulatory capital into tiers. Common Equity Tier 1, or CET1 — mainly shares and retained earnings — is the highest quality and absorbs losses immediately. Additional Tier 1 also absorbs losses while the bank is still operating, on what regulators call a "going concern" basis. Tier 2 capital sits below that, absorbing losses only when the bank has effectively failed — a "gone concern" situation. Total regulatory capital is the sum of these layers.

The crucial new rule was that all AT1 and Tier 2 instruments must be able to absorb losses at the "point of non-viability" — the moment a bank would no longer survive on its own without action. That loss absorption can happen through conversion into shares or through a write-down, and the trigger is the earlier of two things: a decision by the relevant authority that a write-down is necessary, or a decision to inject public funds to prevent failure.

The Mechanics: Triggers, Write-Downs, and Missing Coupons

Every AT1 bond is defined by two things: what sets it off, and what happens when it does.

The most common trigger is a capital ratio. If a bank's CET1 ratio — its core equity divided by risk-weighted assets — falls below a set level, the bond converts or is written down. That level is often 5.125%, though some jurisdictions set it higher; Switzerland, for instance, used a 7% trigger in its regime. The other trigger category is discretionary, resting on a supervisor's judgement that the bank has reached the point of non-viability.

Then comes the loss-absorption mechanism. Some AT1s convert into equity at a preset ratio. Others are written down — partially or entirely, temporarily or permanently. Some Swiss instruments, notably, allowed only a full write-down, with no conversion option at all.

AT1s also carry features that would look strange on an ordinary bond. They are perpetual, with no final maturity date, though they are usually callable after a set period with regulatory approval. Coupons are discretionary and non-cumulative: miss one, and it does not pile up as a debt the bank owes. Non-payment is not treated as a default. And under a rule known as the Maximum Distributable Amount, regulators can restrict distributions — including AT1 coupons — if capital falls below a certain level, though in practice supervisors have tended to trim dividends and bonuses first.

Because of all this, AT1s sit just above common equity in the creditor queue. They offer yields closer to high-yield debt than to ordinary bank bonds, which is precisely the compensation for that position. The market is sizable: by early 2022 it exceeded $250 billion. US banks largely sidestep the instrument, issuing preferred shares to fill their AT1 bucket instead.

The Credit Suisse Moment

Then came March 2023, and the episode that made AT1 a household acronym.

After the collapse of Silicon Valley Bank in the United States, attention swung toward Credit Suisse, a global systemically important bank already weakened by years of losses, misconduct cases, and risk-management failures. On 15 March, a major shareholder ruled out further capital injections, and the share price fell sharply. Emergency liquidity followed from the Swiss National Bank. By the weekend of 17–19 March, Swiss authorities had engineered a takeover by UBS.

As part of that deal, roughly CHF 16 billion of Credit Suisse AT1 bonds were written down to zero. Credit Suisse shareholders, meanwhile, received UBS shares — diluted, but not wiped out.

That ordering startled markets. Conventional corporate hierarchy says creditors rank above shareholders. Here, bondholders were erased while shareholders kept something. The write-down also arrived outside a formal resolution process, relying on contractual terms and an emergency ordinance rather than the resolution framework authorities had spent a decade building.

Supervisors elsewhere moved quickly to draw a line. The Single Resolution Board, the European Banking Authority, and ECB Banking Supervision issued a joint statement confirming that under the EU framework, AT1 instruments cannot be written down before common equity absorbs losses first. The Bank of England made a similar point about the UK creditor hierarchy. The message was: this was Swiss-specific, not a new global template.

The legal story did not end there. In October 2025, the Swiss Federal Administrative Court overturned FINMA's write-down decision, finding that the regulator's action lacked an adequate legal basis — specifically that a "viability event" as defined in the bond documentation had not materialized. The court accepted that the bond terms were convoluted and ambiguous, yet upheld them because FINMA had not required changes before issuance. FINMA has said it will appeal to the Federal Supreme Court. This remains a live legal question, not a settled one.

What This Means for Counterparty Exposure

This is where the story stops being about bond coupons and starts being about plumbing.

Banks are counterparties to each other constantly — in derivatives, lending, settlement, and trading. When you assess a bank as a counterparty, you are asking whether it will still be standing and able to meet its obligations tomorrow. AT1 risk feeds into that assessment, because it tells you how losses get distributed when a bank wobbles.

Here is the tension. The Credit Suisse episode showed that a going-concern write-down can impose losses on AT1 holders while shareholders retain value — a reversal of the assumed hierarchy. It also showed that a supervisor's decision can be contested in court afterward, which introduces legal uncertainty into what had been treated as a mechanical contract term. For anyone pricing counterparty risk, both facts matter.

There is a broader structural question too. Research on ring-fencing — the practice of walling off local operations with dedicated capital and liquidity — suggests that extensive ring-fencing can raise the risk of failure rather than lower it. The reasoning: capital gets trapped in one subsidiary when another needs it, a problem the authors call "misallocation risk." Their modelling found the likelihood of failure could rise several-fold if ring-fencing becomes widespread, because a bank may have enough resources overall but cannot move them fast enough to the right place. That is a reminder that where capital sits, and how freely it can travel, shapes real-world resilience.

A related reform thread is Total Loss Absorbing Capacity, or TLAC — pre-placed resources, including bail-in debt, designed to recapitalize a failing bank rather than let it crash. The idea is that a group failure triggers recapitalization funded by these resources, protecting operating subsidiaries and critical functions. Whether that architecture works smoothly under genuine stress has not been fully tested.

What to Hold Onto

AT1 bonds are not simple yield products. They are regulatory tools with unusual risks: they can be written down, they can stop paying coupons without triggering default, they rank below shareholders in some scenarios, and their triggers can depend on a supervisor's judgement — which is ultimately subject to legal challenge.

Design clarity matters enormously here. Instruments that convert to equity, rather than vanishing outright, appear to align better with the principle that debt should not rank below equity. Ambiguous contract language, as the Credit Suisse case demonstrated, invites litigation and unpredictability — which is costly for everyone, including the banks that issue these bonds and the institutions that hold them as counterparties.

Rules also vary by jurisdiction. Swiss, EU, UK, and Indian frameworks differ in trigger levels, available loss-absorption mechanisms, and creditor hierarchy protections. In India, for example, AT1 instruments under the Reserve Bank of India's Basel III framework carry a minimum CET1 trigger of 6.125%, and eligibility rules for who may hold these instruments are tighter. Nothing here is a recommendation about any security. The point is simpler: these instruments differ meaningfully across borders, and the fine print is the product.