Congress enacted the GENIUS Act last summer with the promise of bringing payment stablecoins into a federal regulatory framework. In the past few months, the federal banking agencies (the OCC and FDIC), the Treasury Department, and the National Credit Union Administration have been rapidly writing the rules that will determine what that framework will look like.
At Better Markets, we have been active in commenting on their proposals and have consistently highlighted one risk that deserves far more attention: step-in risk.
The concept of step-in risk is simple. Financial institutions, particularly large banks, often engage with “affiliated” institutions that bear their name or otherwise have a connection that is not legal in nature but nonetheless makes it known to customers and investors that the two institutions are affiliated. When the affiliated institution is under financial stress, the “parent” financial institution may have no legal obligation to rescue it, but it does often provide financial support—or “step in.” This is because letting customers suffer losses at the affiliated institution could seriously damage the financial institution’s reputation and overall franchise.
Step-in risk is a longstanding concern of financial regulators. After the 2008 financial crisis, the Basel Committee on Banking Supervision—the main international bank standard-setting organization—developed international guidelines requiring banks and supervisors to identify and manage step-in risk. I worked on developing those guidelines while at the Federal Reserve (Fed). One of the clearest historical examples from 2008 was with money market funds (MMFs).
The Money Market Fund Warning
MMFs are investment products that pay investors interest while promising essentially zero credit risk to the value of the initial investment (i.e., returning the investment dollar-for-dollar) by taking certain precautions such as investing in liquid, short-term assets. Importantly, even though they are offered as a bank-like product, they are not bank deposits, do not receive deposit insurance, and their “sponsors”—often larger financial companies (particularly banks) that associate their brand name—generally have no legal obligation to guarantee their value.
Yet financial history shows that legal separation does not necessarily mean economic separation during stress.
As the Boston Fed documented, during the 2008 financial crisis, at least 21 MMFs would have broken the buck absent substantial financial support from their sponsors. The reason for this financial support was largely reputational. Letting investors lose money in a product bearing an institution’s name could damage the entire franchise and trigger runs from its other products. And it was not just during the 2008 crash: between the mid-1980s and the 2008 crash, sponsors provided support over 200 times to their MMFs. More recently, during the 2020 pandemic stress, both Goldman Sachs and the Bank of New York Mellon provided support to their sponsored MMFs. Note that in both 2008 and 2020, the federal government, through the Fed and the Treasury, ultimately was forced to intervene to provide a public backstop for the industry (absent that more step-in support would have likely been necessary).
The Basel Committee subsequently identified exactly this phenomenon as step-in risk: a bank supporting an entity beyond, or in the absence of, its contractual obligations when that entity experiences financial stress.
The lesson for policymakers is straightforward: risk can sit outside a bank’s balance sheet in good times and come rushing back onto it during a crisis. Now stablecoins could recreate and significantly expand that problem.
Stablecoins Will Make the Step-in Problem Much Worse
Like MMFs, stablecoins are built around an expectation of stable value. Unlike MMFs, stablecoins can be traded around the clock. If confidence in a stablecoin issuer or its reserve assets that back its stablecoin begins to deteriorate, holders of stablecoins have a powerful incentive to get out before everyone else, causing a run on the issuer. This has occurred repeatedly with stablecoins already, and the proliferation of stablecoins that surely will follow after the GENIUS Act would make these runs more frequent.
The GENIUS Act requires stablecoin reserves to consist of assets that connect stablecoins directly to traditional finance, including bank deposits, short-term Treasuries, repurchase agreements, and government MMF shares. Therefore, a stablecoin run would directly spill over to and cause stress in the banking system and Treasury markets. The opposite would be true as well—stress at banks or the Treasury markets would directly spill over to stablecoin issuers. For example, in March 2023, USD Coin fell to roughly 87 cents on the dollar after approximately $3.3 billion of its reserves became trapped at the failing Silicon Valley Bank. A problem at one bank almost immediately became a stablecoin run.
Making matters worse, these stress channels do not run only in one direction. In fact, they can create a negative feedback loop. For example, if stablecoin holders run and the stablecoin is backed partly by bank deposits (likely uninsured), the stablecoin issuer will withdraw those deposits to pay back its users, causing liquidity stress at the bank that held its deposits and—if the stablecoin deposits are large enough—potentially triggering a run on the bank. In effect, a run on a stablecoin can become a run on a bank—a “run on run” dynamic.
The European Central Bank recently highlighted this same problem. Its researchers warned that links between stablecoins and MMFs could transmit stress in both directions and specifically identified the possibility of sponsor support from parent banks when stablecoin issuers come under pressure. In other words, the risk is not limited to whether a stablecoin has enough reserves. The question is also where the losses and liquidity demands go when those reserves are not enough to stop a run.
Adding Megabanks to the Equation
Under the GENIUS Act, a large bank can sponsor a stablecoin through an affiliated issuer, distribute the stablecoin directly, and put the bank’s brand behind it.
Legally, that stablecoin would remain separate from the bank. However, if that stablecoin comes under stress and suddenly trades at 90 cents on the dollar, would anyone reasonably expect the bank not to intervene to bail out the stablecoin holders?
Would bank management allow a financial product carrying the bank’s own name to collapse if providing liquidity could protect the institution’s reputation and broader franchise?
Financial history clearly indicates the answer is too often no.
The Basel framework tells supervisors to look specifically at sponsorship, branding, reputational risk, first-mover incentives, and a history of financial support when assessing step-in risk. As we note in our GENIUS Act comment letters, those factors read almost like a checklist for a bank-sponsored or credit-union-sponsored stablecoin.
Regulators Should Learn From History
Better Markets has argued that banking agencies should explicitly address step-in risk as they implement the GENIUS Act. Unfortunately, the agencies to date have not even acknowledged this concern in their proposals. This approach is inconsistent with prudent rulemaking, and we will continue to press the agencies (including the Fed, which hasn’t yet released its GENIUS Act proposal) to engage on this critical topic. Simply put, not acknowledging a problem does not make it go away.
We outline a few recommendations for next steps:
Banks should have to identify stablecoin arrangements that create material step-in risk.
Regulators should closely scrutinize stablecoin branding because putting a bank’s name on a stablecoin can itself create an expectation that the bank stands behind it.
Stablecoin issuers should also have credible plans showing how they could fail without financial support from an affiliated bank.
And even if regulators cannot eliminate step-in risk, they should make banks hold sufficient capital and liquidity against it.
Kevin Warsh and Ending the Fed’s Bailout Playbook
Fed Chair Kevin Warsh recently told Congress that the Fed “does not want to be in the bailout business, full stop.” The context of Warsh’s remarks is particularly relevant for step-in risk. When Representative Brad Sherman asked Warsh whether the Fed might provide liquidity to stablecoins or cryptocurrencies during a run, as it did for MMFs during prior crises, Warsh said the Fed should do everything it can to mitigate those risks so that it is never put in the position of bailing out anyone, “including crypto.”
Better Markets supports Warsh on his stated goal, but avoiding bailouts requires more than words that promise not to intervene once a crisis has already begun. It requires addressing beforehand the risks through strong and credible rules that make government intervention much less necessary. If Warsh wants to ensure that the Fed never has to create the stablecoin equivalent of the Money Market Mutual Fund Liquidity Facility, the Fed should use its forthcoming GENIUS Act rulemaking—and its existing bank supervisory and regulatory authorities—to address step-in risk now. Requiring banks to identify these exposures, limiting the expectations created by bank branding, ensuring stablecoin affiliates can fail without bank support, and requiring adequate capital and liquidity against residual step-in risk are precisely the kinds of measures that can make Warsh’s commitment credible.
Conclusion
The GENIUS Act was supposed to make stablecoins safer by bringing them inside a federal regulatory framework. But if its implementation allows stablecoin risks to remain outside bank balance sheets during good times while migrating back onto those balance sheets during bad times, regulators will have recreated one of the oldest problems in financial regulation.
Money market funds already taught us how that story ends. We should not need another crisis with stablecoins to learn the lesson again.
For more information on stablecoins and the GENIUS Act, see Better Markets’ comprehensive stablecoin page.
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.