The Bail-In of Credit Suisse CoCos: Why Principal Write-Down Made Sense

Credit Suisse (CS) was sold to UBS on March 19, 2023, to avoid its further deterioration from long-lasting distress and widespread distrust, especially after the collapse of Silicon Valley Bank. The most contentious point of the deal may have been the write-down of the principal of its contingent convertible bonds (“AT1 CoCos”) while shareholders retained a minimal claim on the bank.

In this post, we react to the post by Eidenmüller and Paz Valbuena. They argue that the write-down breached the principle of pari passu, which holds that shareholders should be wiped out before bondholders suffer losses. They also argue that the actions of the Swiss authorities with regards to the sale of CS to UBS can be qualified as a bail-out. We disagree with both points. For our analysis, we draw on an earlier post, in which we argued that the actions of the Swiss authorities made sense from a corporate finance, legal, and financial stability perspective.

First, investors in CS did bear a significant part of the losses. UBS agreed to purchase CS by acquiring its outstanding capital for 0.76 Swiss francs (CHF) per share. According to the 2022 CS Annual Report, the book value per share was 11.45 CHF. This means that, compared with the account value of 2022, CS shareholders lost more than 42 billion CHF. Accounting values are clearly imperfect, but they are the metric that matters for determining bank insolvency. In addition to this, the write-down of AT1 CoCos is by any definition a form of bail-in, amounting to 17 billion CHF. The losses of CS have therefore mainly been borne by shareholders and AT1 CoCo holders. The “bail-out” component is limited to the conditional loss guarantee provided by the Swiss government to UBS (up to 9 billion CHF). The Swiss National Bank also provided a liquidity line to UBS to support the acquisition, but this cannot be considered a bail-out, because providing liquidity to solvent banks is the core business of central banks, as explained by Bagehot 150 years ago. This construction is perfectly in line with the policy goal stated by the Financial Stability Board in the aftermath of the financial crisis.

More important from a legal perspective, Credit Suisse has never been declared insolvent. Though the determination of bank insolvency is inherently a contentious, it is self-evident that bank insolvency must be declared before the rules of insolvency can be breached, and this was not the case. In this sense, the European Central Bank and the Bank of England stated little more than the obvious: In bank resolution, the  rules will be upheld. But CS has never entered into resolution.

We also disagree with the claim that AT1 CoCo holders should not be wiped out before shareholders in the future, as this would be a breach of the absolute priority rule. However, AT1 CoCos are supposed to suffer losses in a going concern. This is why they were designed and included in Basel III, and this is why many jurisdictions allow them to be counted as core capital. If the regulatory design does not allow for going-concern loss absorption, AT1 CoCos are useless, and they should not count as core capital. Indeed, reading the contract of the AT1 instruments issued by Credit Suisse, it is easy to realize that creditors agreed to absorb losses in a going concern in the form of principal write-down. This could happen if the accounting value of common equity fell below a certain threshold (7 percent over risk-weighted assets) or if the competent authority decided it could happen based on its legal powers. In exchange, AT1 CoCo holders earned a high interest rate (some AT1 instruments yielded 9.5 percent per year). This implies that the critiques about the breach of priority rules are unwarranted. AT1 CoCo holders accepted the possibility of absorbing losses before shareholders in a going concern.

Eidenmüller and Paz Valbuena also argue that AT1 CoCos should be of the “conversion to equity” (CE) and not of “principal write-down” (PWD) type. In contrast, we believe that the law should grant contractual freedom to private parties, since banks and AT1 CoCo holders are generally sophisticated parties.

Credit Suisse was among the few banks issuing AT1 instruments with a PWD mechanism. This is the reason for the write-down. If the contractual agreements were of the “conversion to equity” type, the Swiss authority would have diluted more existing shareholders while converting AT1 CoCos to equity.

Given the sophistication of all parties, we do not believe that the law should mandate either mechanism as there is no ‘investor protection’ rationale for doing so. If the threat to impose going-concern losses is credible, the market can price it, and parties can make efficient decisions. It is also easy to think of instances where the PWD mechanism is preferable. For example, shareholders may want to maintain the current ownership structure in case of going-concern loss absorption, because a change of control in a crisis scenario could come with significant costs or because shareholders enjoy significant private benefits of control. Shareholders may be willing to remunerate AT1 CoCo holders for the higher risk associated with PWD. Another example is the case for remuneration: Many CS material risk-takers were partially remunerated with AT1 CoCos to decrease their risk appetite, as they face the risk of being bailed in if the bank gets in distress. In this case, a PWD mechanism is clearly a better fit. This also counters the argument raised in the literature (e.g., Berg & Kaserer, 2015) that PWD may encourage excessive risk-taking. Finally, recentpapers suggest that the PWD can also help to prevent ex post risk shifting (“gambling for resurrection”), by reducing leverage and ensuring that shareholders retain equity in case of financial distress of the bank – although there is a trade-off with lower ex ante incentives for shareholders to prevent financial distress.

We do not argue that PWD will always (or even in most cases) be more desirable than CE – we simply believe that the choice should be left to banks and AT1 CoCo investors. In the global market for AT1, the PWD and CE design are equally popular, while the PWD design is predominant in Switzerland (Avdjiev et al, 2020 p. 596). This is a sign that this type of security design may be efficient for some banks.

In the end, our arguments are based on the idea that the failure of banks is not comparable to the failure of any other corporation. From this perspective, the CS case, the first failure of a systemic bank in 15 years, is paradigmatic of the inherent tensions in handling bank failures. Therefore, drawing general lessons can be useful in facing the current turmoil and designing more resilient and effective regulation. Two aspects are worth discussing, as they highlight the uniqueness of bank failures.

First is the interplay between accounting standards and regulation. compliance with capital requirements ultimately refers to accounting standards and practices used to evaluate banks’ assets. Therefore, the incentive of both banks and regulators is to delay the recognition of losses to avoid sending adverse signals to the market and avoid potential runs. This was the case for Credit Suisse, as the SEC highlighted accounting errors months before the sale. The problem seems larger and systemic. The Silicon Valley Bank collapse provides additional anecdotal evidence supporting this claim. The mortgage-backed securities and treasuries of SVB were treated according to the “held-to-maturity” accounting standard, not considering that those assets could be sold before maturity in case of runs. And indeed, the run happened, SVB sold its assets at fire-sale prices compared with their “held-to-maturity” accounting value, and the bank quickly became insolvent. The role of accounting standards in banking regulation is definitely a key vulnerability for any regulatory framework.

Second, and relatedly, it is extremely difficult to determine when non-viability for banks occurs, triggering a bank resolution. As the point of non-viability is closely related to accounting metrics, it is easy to game, and there are incentives to delay and ask for forbearance from regulators.

The actions of the Swiss authorities, however imperfect and vulnerable to criticism, go in the right direction of holding bank insiders – shareholders, hybrid-capital holders, and material risk takers – liable for their actions and investment decisions. This raises a challenge for legal scholars and policymakers: upgrading the current framework(s) in a way that makes it possible to resolve the distress of systemic banks early on, in a smooth and legally sound manner.

This post comes to us from professors Edoardo Martino at the Amsterdam Center for Law & Economics and Tom Vos at the Jean-Pierre Blumberg Chair (University of Antwerp, Belgium).