Since the 2008 financial crisis, bank regulation has grown at a remarkable pace. Yet much of it has treated only the symptoms of issues like excessive risk. In a recent paper, I address a mismatch in systemically important banks between who decides how much risk to take and who bears its cost and ask what regulation can do about it. Most regulatory tools leave that mismatch in place. Governance is the one tool that can reach it without running into the political, design, or measurement barriers that block the others. I propose using governance measures to give a defined class of bondholders more influence over bank boards.
The Problem: Risk Preferences and Control Are Misaligned
Banks typically fund more than 90 percent of their balance sheets with debt, and their shareholders enjoy limited liability. Equity in a bank behaves like a call option on the bank’s assets. Volatility hurts almost every other constituency, but it benefits the holder of the residual claim. Furthermore, deposit insurance and implicit too-big-to-fail guarantees weaken the incentive of depositors and creditors to monitor bank behavior. When a systemic bank fails, much of the cost falls on taxpayers and the wider economy. They never contracted with the bank and have no way to protect themselves.
The result, which I develop in a companion paper, is a structural mismatch. Shareholders and senior management have the strongest appetite for risk and the most control over it. Creditors and the public prefer less risk but have little say. Conventional “good governance,” which aims to align the incentives of managers with those of shareholders, makes the problem worse rather than resolving it.
Seven Tools, One Recurring Pattern
The paper evaluates seven regulatory alternatives: detailed substantive regulation, capital requirements, extended shareholder liability, recovery and resolution regimes, taxation, the breaking up of large banks, and the regulation of governance. No single winner emerges, but a pattern does.
Some measures leave incentives untouched. Detailed substantive regulation can respond precisely to known problems and is politically feasible. But it is backward-looking, it is costly to enact, comply with, and supervise, and it invites regulatory arbitrage. Removing the tax deductibility of interest could raise capital ratios, but it would not change why shareholders prefer risk. Breaking up large banks addresses size and interconnectedness, but it leaves the risk appetite of the remaining institutions as it was.
Capital requirements do reach incentives. More equity gives shareholders more to lose and narrows the gap between their interests and those of creditors. The difficulty is scale. Increases large enough to change behavior go far beyond what is politically feasible. Recent U.S. experience illustrates the point. In 2023, the federal banking agencies proposed rules that would have significantly raised capital requirements for large banks. After intense industry opposition, the agencies replaced that package in March 2026 with proposals that, by their own estimate, would modestly reduce overall capital in the banking system.
Other measures fail because of their design. Extended shareholder liability raises questions with no obvious answers. Should liability be unlimited or capped? Joint and several? Enforced before or after failure? Owed to whom? A Pigovian tax on systemic risk is elegant in theory. In practice, no taxable proxy tracks systemic risk well, and the right level of tax cannot be measured.
Recovery and resolution regimes are promising because they target the bailout guarantee itself. But they remain largely untested and exposed to political intervention at the very moment they are needed. In the 2023 Credit Suisse case, authorities departed from the conventional ranking between bail-in bonds and equity. That shows how quickly market expectations about resolution can be upended.
Three further problems cut across every option. A regime limited to systemic banks needs a threshold, and any threshold is partly arbitrary. Deciding who should bear the cost of intervention is a Coasean puzzle, because the same scale that creates systemic risk also delivers services the economy wants. Finally, every tool that expands supervisory discretion is exposed to capture. Capture can be material, through the revolving door, or cognitive, as supervisors come to think like the bankers they supervise.
Why Governance, and Why the Board
Governance has the greatest unrealized potential of the seven. It can act on incentives rather than symptoms. It can work through principles rather than detailed rules, which leaves less room for arbitrage. It also operates at the top of the corporate hierarchy. The board sets the bank’s risk appetite and appoints the CEO, so its choices shape risk-taking throughout the organization.
Regulators have addressed governance since 2008, but the reforms have mostly targeted remuneration, board diversity, independence, and alleged short-termism. Some of these reforms have merit, but few address the board’s appetite for risk. Regulators typically require robust processes for setting and monitoring risk appetite while avoiding any view on its level. The weakness lies in how governance regulation has been used, not in what it can do.
Bondholders as a Proxy for Risk-Averse Stakeholders
My proposal is to moderate the board’s risk appetite by strengthening the influence of creditors in board governance. Bondholders would serve as a proxy for the wider group of risk-averse stakeholders, including the public, whose interests are typically not represented. Bondholders are a sensible choice for several reasons:
- They are an identifiable group, unlike “the public.”
- They have their own capital at risk.
- They are typically financially sophisticated.
- They already have a contractual relationship with the bank.
- They supply far more of a bank’s funding than shareholders do.
- Their portfolios tend to be less diversified, which sharpens their interest in monitoring.
One could also give supervisors a greater role in board governance, an option examined by Davies and Hopt. But a board balancing the interests of three constituencies (shareholders, creditors, and supervisors) would be harder to hold accountable. Supervisors also have no capital at stake and face the capture problems described above. Two private parties with money at risk is the cleaner design.
Not every bond fits. The added influence should attach to instruments whose holders have the strongest reason to care about downside risk: unsecured, junior, long-dated or perpetual, subject to bail-in, and as standardized as possible. Bail-in refers to the write-down or conversion of bonds into equity when the bank is failing or likely to fail, so their holders bear losses instead of taxpayers. Existing bail-in debt, such as instruments counted toward the total loss-absorbing capacity (TLAC) of global systemically important banks, offers a practical starting point. The proxy is imperfect. Bondholders care only about the losses they would bear, not the full social cost of a failure. Some institutional investors hold both the debt and the equity of the same bank. Implicit guarantees also dull bondholder incentives in exactly the cases where they matter most. For that reason, the proposal must be paired with credible steps to shrink those guarantees. The claim is comparative. Bondholder preferences sit closer to the public interest than shareholder preferences do, so shifting influence toward them would improve on the status quo.
Changing Who Has a Voice
None of this means abandoning capital rules, resolution planning, or supervision. They remain necessary. The question is whether to keep adding rules on top of a governance model that rewards risk, or to change who has a voice when the board sets its risk appetite. If that appetite moved closer to the preferences of the bank’s risk-averse stakeholders, some of the regulatory complexity built since 2008 could be scaled back, or at least it could grow more slowly.
The paper makes the case for this approach and identifies the proxy. It does not prescribe a particular mechanism, such as board seats. The detailed mechanisms through which creditor influence could be exercised are developed in my 2023 book.
Roni Laakso is a postdoctoral researcher at the University of Helsinki Faculty of Law and a visiting researcher at Aalto University School of Business. He is also senior legal counsel at Nordea Asset Management. This post is based on his recent paper, “Regulating Excessive Risk-taking in Systemic Banks: The Limits of the Existing Alternatives and the Case for Increased Creditor Influence in Board Governance,” available here. The views expressed in this post are his own and do not reflect those of Nordea Group.