Will new long-term debt holders effectively discipline banks?

Holders of new long‑term U.S. regional bank debt should recognize that it has the potential for losses in the event of a bank “bail‑in,” which is a desirable feature for regulators as we discussed in our “New Long‑Term Bank Debt to Expose Holders to Equity‑Like Risk” white paper. This addition of “loss absorbing capacity” can minimize the cost of a failed bank to the Federal Deposit Insurance Corporation (FDIC) and to society at large in a more explicit manner than existing debt. The new long‑term debt will effectively require banks to pre‑fund their own bail‑in, with bondholders absorbing the costs of the failed banks and equity investors and bank customers paying higher costs along the way. 

Long‑term debt also has the advantage of potentially reducing the probability of default in the first place. The idea is that the presence of these instruments enhances market discipline. According to the Bank for International Settlements (BIS), “if the risk of bail‑in is more explicit and credible than before the global financial crisis, investors in bail‑in debt instruments should have stronger incentives to monitor a bank’s risk of failure. And by factoring assessments of bail‑in risk into their investment and pricing decisions, investors should be able to exert discipline on banks’ risk‑taking.”¹ 

Put another way, the burden of monitoring bank behavior shifts to the bondholder and away from authorities such as the FDIC, which are signaling they have neither the willingness nor the ability to enforce sufficient discipline.

Channels for creditors to impose market discipline 

Creditors can impose discipline on banks through three key channels:

1. By demanding a higher interest rate when lending to riskier banks, thereby raising the cost of funding and incentivizing banks to reduce risk 
2. By demanding restrictive covenants on debt issuances, restricting riskier activities explicitly.
3. By refusing to participate in new debt issuances altogether, creditors can force banks to raise liquidity from alternative sources—potentially including a backstop facility, such as the Fed’s discount window.² Borrowings from these facilities are usually viewed negatively by creditors and depositors, a property that disincentivizes their usage by banks except as a last resort. The stigma around the usage of these facilities raises the risk that depositors will flee the bank, potentially triggering a catastrophic “run.”

In practice, restrictive covenants are typically not imposed on bank-issued bonds and, therefore, do not vary for riskier banks. This leaves the risk of a catastrophic run and the risk of a higher cost of funding as the two main channels by which long‑term debt holders can impose discipline.

Importantly, these two primary channels require that markets can effectively and efficiently price risk and/or refuse to participate in new debt issuances outright. There are also secondary forms of market discipline, such as higher bank borrowing rates leading to greater supervisory scrutiny, a run on the bank prompting a preemptive bail-in of debt by regulators, or downgrades from rating agencies catalyzing deposit outflows.

Long‑term debt holders uniquely able to monitor bank risk

There are broadly three different types of providers of capital and funding to banks in the normal course of business:

1. Equity investors
2. Long‑term debt investors
3. “Systemically important debt” holders, such as bank depositors

Among capital and funding providers, long‑term debt holders are uniquely able to monitor banks’ risk taking and impose market discipline. To illustrate, consider an equity investor. This is an investor that is willing to invest in a high‑risk bank offering positive expected equity returns, even if the probability of default is high. A long‑term debt holder, meanwhile, has limited upside and tends to be much more sensitive to the downside. Put another way, the payoff structure for equity investors means that their focus is less aligned with minimizing the probability of default than it is for a bondholder.

Similarly, systemically important debt holders—depositors in a bank—are not sensitive to the riskiness of the bank and so view their deposit as “informationally insensitive.” In fact, this fulfills one of the core roles of a bank, the creation of “money-like” claims that can easily and readily be used to make payments and facilitate transactions. This information insensitivity is created explicitly through FDIC guarantees or implicitly by issuing long‑term debt that is junior to deposits in the capital structure. 

Typically, these systemically important debt holders are insured depositors, but they can also include uninsured depositors. As we saw during the failure of Silicon Valley Bank, there were depositors who in some cases had left hundreds of millions of dollars on deposit at the bank. 

As a result, long‑term debt holders uniquely perform an important role from a financial stability standpoint, given their returns align with the goals of financial stability—minimizing the risk of bank failure.

Potential for distortions in measures of market risk

The important assumption in all the above is that long‑term debt holders impose discipline through market actions—either by demanding a higher yield or by refusing to refinance due maturities—that have a direct impact on banks that investors consider to be higher risk.

However, there are risks that the market can overshoot in either direction. For instance:

• What if the informational content in these market actions has been muted or diminished by heavy ongoing flows into passive fixed income investment vehicles? Does a persistent bid for all index‑eligible banks—weighted not by their credit worthiness but by the amount of outstanding debt they have—push credit spreads³ tighter and reduce the power of monitoring?
• Similarly, does contagion across banks punish good banks and bad banks simultaneously because the funding spread for all banks tends to increase as market stress rises? What role do changes in market structure play here? 

Observation 1: For instance, while active fixed income managers can discipline bank risk taking through careful monitoring, passive managers are mandated to buy the entire index. Under the assumption that passive flows exert downward pressure on funding spreads, does this risk blunting the “market discipline” argument that policymakers present for the introduction of long‑term debt? 

Observation 2: If a bank fails, the single‑point‑of‑entry resolution framework assumes that the bank operating company can remain a going concern, recapitalizing itself using resources (such as long‑term debt) that are held at the holding company. However, one important assumption is that the bank operating company can continue to access funding from depositors and wholesale funding markets—or informationally insensitive funding sources. While the FDIC typically provides bridge liquidity through the transition, another assumption is that the bank operating company will eventually stabilize.

Are these reasonable assumptions? Will customers of the bank remain? Even though their claims are senior to long‑term debt, will a repo counterparty, for instance, remain at a bank that has failed, or will that counterparty’s risk committee force it to withdraw funding? If this short‑term funding leaves the bank, can the single‑point‑of‑entry resolution framework work?

Observation 3: The contagion effects from a bail‑in of one bank would likely reprice credit spreads for the whole sector wider, especially if the bail‑in coincided with a period of heightened systemic risk. Would regulators be willing to let this happen? This is particularly relevant because the requirement to have long‑term debt forces banks back into the bond market on a regular basis to refinance upcoming maturities. While we could argue that wider spreads are the result of bondholders imposing discipline, would regulators be willing to tolerate this during a period of market turmoil?

Observation 4: Imagine for a moment that these bonds become index ineligible. This is not inconceivable. For instance, the Bloomberg indices explicitly exclude contingent convertible securities.⁴ The proposed long‑term bank debt gets dangerously close to that line. For instance, one way to think about the new debt is that it is a contingent convertible security, except with a trigger point determined by the regulators (and coinciding with a bank resolution). 

What if Bloomberg decides to exclude these securities from indices? The passive bid would drop away and spreads would widen as only portfolio managers with off‑index allocations to put to work would hold the debt. Put another way, the inclusion of long‑term debt in major bond indices requires passive entities to buy. Does this mitigate the market discipline feature that investors can have by refusing to participate in new issues?

Observation 5: When market spreads are tight, investors do not exert as much discipline in monitoring banks. Findings from the BIS1 show that when investor credit risk appetite is strong, the link between issuers’ credit risk and the bail‑in risk premium (BIRP) loosens. This opens up windows of opportunity for riskier banks to issue bail‑in bonds at comparatively low cost. Yet it also implies weaker market discipline on these banks’ risk taking.

Observation 6: Is it safe to assume that imposing losses on pension funds, mutual funds, and individuals will be acceptable in the event that bail‑in does occur? Experience in Italy and Portugal suggest that the political backlash against regulators could be severe. Which is better, the government explicitly bailing out a failing bank and imposing a cost on society or individual investors bearing the cost of bank failure, which also imposes a cost on society?

Observation 7: Are the appropriate legal and regulatory requirements met by the bail‑in of long‑term debt? In its review of the failure of Credit Suisse, the Financial Stability Board⁵ indicated that a bail‑in of total loss absorbing capacity (TLAC) instruments may have encountered legal challenges related to Securities and Exchange Commission (SEC) registration. Would newly created long‑term debt issued by U.S. regional banks also suffer from the same issues?

Questionable effectiveness in signaling risk or absorbing losses

We have highlighted some market structure concerns around the implementation of long‑term debt for U.S. regional banks as well as some considerations that may limit the effectiveness of the bail‑in of newly created long‑term debt.

The ultimate effectiveness of long‑term debt is important. However, we can envision two possible broad scenarios, and both of them call into question that effectiveness:

1. Long‑term debt is bailed in during a bank failure. In this case, we have concerns about the premium the market would demand for this feature given the presence of passive investors and the pro‑cyclicality of that premium (the tendency for funding costs to increase across all banks in times of market stress) and the risk that losses would be borne by a broad base of investors.
2. The broader consequences of bail‑in (such as legal issues, contagion, and imposition of losses) mean that regulators are ultimately unable to bail in long‑term debt, rendering it ineffective as a loss absorbing resource.