Analyzing the history and legal challenges of Additional Tier 1 (AT1) bonds in bank crises, focusing on cases like Credit Suisse and Yes Bank, and the complexities of resolving bondholder losses and shareholder impacts.
By Krishnan Ranganathan In 2005, an American economist, Mark Flannery, proposed a bond with a self-destruct button. His paper bore a title any gym trainer would love: “No Pain, No Gain.” When a bank gets into trouble, a slice of its borrowings turns into shares or simply vanishes, topping up the owners’ cushion (equity) without a rupee from taxpayers. Lloyds sold the first big batch in 2009; Basel III then made a perpetual version a formal layer of bank capital, Additional Tier 1 (AT1).
Think of it as a car’s airbag: built to burst, but only in a crash, and only if it’s wired in. Lloyds’ bonds ended up in Britain’s Supreme Court where bondholders lost 3-2 in 2016. Lawyers, it seems, are a standard feature of the product. Shareholders walked off with roughly $3 billion from UBS. The usual pecking order (owners lose first, lenders after) was turned upside down. In 2025, Switzerland’s Federal Administrative Court ruled the write-off unlawful.
The bond contract allowed a wipe-out only if state support improved the bank’s capital; the rescue gave Credit Suisse cash (liquidity), not a thicker loss-absorbing cushion (capital). The emergency decree Finma also leaned on lacked a legal basis. The damning exhibit was Credit Suisse’s own email saying the write-down conditions were not met. Finma and UBS have appealed. The Indian rhyme In March 2020, the government, at the RBI’s request, froze Yes Bank, and SBI led a Rs 10,000-crore rescue.
The draft reconstruction scheme allowed the AT1s to be written down; after objections, the final scheme dropped that clause. The administrator wrote off Rs 8,415 crore anyway. This month the Supreme Court, upholding the Bombay High Court, inferred the Centre had made a “conscious decision” to drop it, leaving the administrator no power to use. Same plot as Zurich: a hurried rescue, bondholders zeroed while shareholders survived, and officialdom undone by its own side’s paperwork.
One difference matters. The Swiss court found Credit Suisse still met its capital rules when it died. Yes Bank’s core equity ratio had shrunk to 0.6%, against a 7.375% minimum, after a Rs 18,564-crore quarterly loss. India’s bondholders held a bond that deserved to die; Switzerland’s did not. Did the authorities get it right? On the big call, yes. Basel’s rules exist precisely to make AT1 holders, paid 9% or more for the risk, share the pain when a bank reaches non-viability.
On execution, no. Look at Spain’s Banco Popular in 2017: shareholders were wiped out, the AT1s and the next layer of junior debt were converted and erased, and Santander bought the bank for €1. India let existing shareholders keep diluted stakes while bondholders got nothing. Indian rules permit that order; Europe’s regulators insisted after Credit Suisse that equity goes first. Nor did anyone need to zero the bonds.
The scheme could have converted them into shares at the rescuers’ price of Rs 10. Bondholders had offered to accept as little as a fifth. Right instinct, wrong paperwork, crooked pecking order. Is the court right? In law, yes. Taking property needs the authority of law, as India’s Constitution (Article 300A) says, and the government’s own scheme withdrew it. The global record is consistent: where the write-down power sat in statute, as in EU courts, judges upheld it; where it rested on a decree or a deleted clause, as in Switzerland and India, they struck it down.
A bond that can be erased by official whim cannot be priced. In economics, the result is awkward. With its auditor doubting it could carry on, a 2020 liquidation would probably have left these bondholders with nothing. That is the global yardstick: “no creditor worse off” than in liquidation. Restoring full principal leaves them far better off, at the expense of today’s shareholders; many of the 2020 rescuers have since sold stakes to SMBC.
Nor has the court made AT1s safe; it faulted this write-off, not all write-offs. Blame the drafters, not the judges. Has Yes Bank behaved well? Not before the crisis. Sebi fined it `25 crore for pitching AT1s to fixed-deposit customers as if they were deposits (the penalty is on interim stay). Appealing, as UBS did in Switzerland, was defensible. Its disclosure since helped: core equity (CET1) falls about 2.4 points to 11.4%, still above the 8% floor, and booking the hit would have turned the quarter’s profit into a `5,217-crore loss.
The running bill is nastier: coupons of 9.5% and 9%, about Rs 770 crore a year, plus a fight over 5.5 years of unpaid coupons that Crisil reckons could push CET1 to 10%. A new referee? Will a separate resolution authority, as the withdrawn Financial Resolution and Deposit Insurance (FRDI) Bill of 2017 proposed, be better able to handle failing firms? America’s Federal Deposit Insurance Corporation shows separation can work.
Britain shows it’s not necessary: the Bank of England, supervisor and resolver under one roof, sold SVB’s British arm to HSBC for £1 over a weekend in 2023. Yes Bank’s bondholders were hurt by a hard referee without a rulebook, not a soft one. And the FRDI Bill died in 2018 partly over turf, mostly over fear that its “bail-in” clause would raid deposits. Change the rulebook before changing the referee.
The airbag only deploys if it is wired in. The writer is Former Executive Director, Nomura, alumnus of Harvard Business School, and a visiting faculty at B-schools Disclaimer: The views expressed are the author's own and do not reflect the official policy or position of Financial Express.
