A “Narrow” Path to Efficient Digital Currency

In this paper, I argue for broadening the scope for digital currency market competition and innovation by allowing for “narrow” stablecoins, meaning ones that, though uninsured, are backed exclusively by Federal Reserve master account balances or, perhaps, by those balances and certain U.S. Treasury Department securities.⁴ While narrow stablecoins needn’t be riskier than either insured stablecoins or CBDC, allowing for them could make for a far more competitive digital currency market. This enhanced competition would in turn result in more rapid and complete financial inclusion, with improved payments-system efficiency stemming from enhanced price and quality competition and more vigorous technical innovation.

Strictly speaking, “digital currency” refers to electronic payments media that can pass directly and repeatedly from one electronic, or digital, wallet to another, much as old-fashioned paper currency can pass from one leather wallet to another.⁵ This ability makes it unnecessary for recipients of digital currency to have bank accounts. Consequently, digital currencies can allow even the unbanked, meaning those who can’t afford to keep bank accounts or who simply prefer not to deal with banks, to take advantage of the speed, convenience, and low cost of digital payments.

Stablecoins are privately issued digital currencies.⁶ They are token-based, meaning that, like old-fashioned banknotes, they are claims upon, or liabilities of, their issuers only and not (like checks) liabilities of the particular persons who offer them in payment: someone offered stablecoins in payment must trust the coins and the firm that issues them; but that person needn’t worry that the payment will “bounce” because the payer lacks sufficient funds.

A CBDC, in contrast, can be either token-based or account-based: that is, it might simply consist of “retail” or “individual” central bank deposit accounts.⁷ The Fed’s discussion paper, for example, defines a prospective Fed digital currency as any “digital liability of the Federal Reserve that is widely available to the public.” Account-based CBDC can also allow unbanked persons to transact digitally, provided that “unbanked” is understood to mean not banked by any private-market depository institution.⁸

The above summary should suffice to justify the PWG report’s claim that, provided they are “well-designed and appropriately regulated,” stablecoins might “support faster, more efficient, and more inclusive payments options,” including considerably cheaper cross-border remittances.⁹ CBDC proponents believe it can offer the same benefits.

Any digital currency also poses risks, including the risk of allowing illegitimate breaches of its users’ privacy or of serving criminal ends. But stablecoins pose a risk of particular concern to regulatory authorities that CBDCs don’t pose, namely, the risk that an issuer might default as a result of losses on assets backing its digital liabilities. In theory, the mere prospect of such a default could inspire redemption runs, not only on a troubled stablecoin issuer but also on solvent stablecoin issuers.¹⁰ Stablecoin issuers’ attempts to meet such runs can result in “fire sales” of their less-liquid assets, spreading losses to other financial firms holding the same assets.

Such systemic risk is, of course, also a feature of conventional systems of deposit banking, where it supplies the rationale for deposit insurance. The assumption that stablecoins might also endanger not only digital currency users but “the broader financial system” leads naturally to the PWG’s recommendation that Congress “require stablecoin issuers to be insured depository institutions … subject to appropriate supervision and regulation at the depository institution and holding company level.”¹¹

But a plan placing the U.S. stablecoin market off-limits to all save insured depository institutions itself poses a serious risk, namely, the risk of a highly concentrated or insufficiently “contestable” digital currency market, dominated by a small number of insured banks (or, if the Fed also enters the market, by them and the Fed).¹² A digital currency market with few participants is likely to result in more limited product variety, lower product quality, and higher fees than one that’s more hotly contested. It is also less likely to be successful in banking the unbanked. Finally, and perhaps most importantly, a stablecoin industry open only to insured depository institutions may fail to provide strong incentives for ongoing digital currency innovation.¹³

Responding to the PWG report, Fed Governor Christopher Waller expressed similar worries.

The view that only insured depository institutions should be allowed to supply stablecoins rests on the assumption that stablecoins pose the same systemic risk posed by ordinary bank deposits. Such deposits are risky because they’re only partially backed by perfectly safe and liquid central bank liabilities. Otherwise, they’re backed by bank loans and other less than perfectly risk-free and liquid assets. These less-liquid and risky assets expose banks to losses when their borrowers default and make it impossible for banks to pay off all their depositors at once in the event of a run. By limiting potential depositor losses, deposit insurance makes runs less likely.

But deposit insurance isn’t the only way deposits can be made safe. Another option, widely discussed if seldom put into practice, is full-reserve or “narrow” banking. A narrow bank is one that fully backs its deposits with cash reserves, generally meaning (in the U.S. context) either Federal Reserve notes or balances in their Fed master accounts.¹⁵ Narrow banking proponents see it as a substitute for deposit insurance, because it would also assure depositors that their money is safe. It has the further advantage of avoiding the “moral hazard” problem, that is, the problem of bank depositors being inspired by insurance to favor banks that take excessive risks.¹⁶

Narrow stablecoins are the digital currency counterpart of narrow banking, and what is proposed here is simply that narrow stablecoin issuers be allowed to participate in the stablecoin market along with insured stablecoin issuers. Though they would be uninsured, narrow stablecoin issuers would not be exempt for other sorts of regulation. They might, for example, be required to meet certain standards aimed at guaranteeing or enhancing their networks’ interoperability.¹⁷ They should not, however, be subject to regulations that would be superfluous given the extraordinary safety and liquidity of the assets backing their digital currency liabilities.

Despite its name, the “synthetic CBDC” plan proposed by the International Monetary Fund’s Tobias Adrian at the Swiss National Bank Conference in 2019 is actually an example of a narrow stablecoin plan. Under it, central banks would offer stablecoin issuers access to their settlement accounts under the condition that they fully back their stablecoins with settlement account balances while insulating that backing from claims by other creditors. Besides being safe, Adrian says, this arrangement “would ensure interoperability” of various stablecoins “by offering a common settlement platform.”¹⁸

Although calling a narrow stablecoin plan one for a “synthetic CBDC” is a bit misleading, there’s a sense in which such a plan represents a compromise between the options of a genuine CBDC and the status quo. Presently, the Fed is a “wholesale” supplier of digital currency only—and one that deals only with licensed banks, the U.S. Treasury, and some government-sponsored enterprises. What’s more, it occasionally denies master accounts to banks that are legally eligible for them. The Fed has been sitting on master accounts applications submitted by Kraken and Avanti, two uninsured cryptocurrency firms that received special purpose depository institution charters from the state of Wyoming in 2019. Avanti plans to offer a token-based stablecoin called the Avit.¹⁹

In a true CBDC arrangement, including the Fed’s “intermediated” CBDC proposal, the central bank becomes a retail supplier of digital media. In the Fed’s version, it uses commercial banks and nonbank payment service companies as its digital currency distributors. Though it calls them “intermediaries,” these private-sector agents are more like custodians that hold and manage retail Fed accounts on behalf of the Fed and its retail customers.²⁰

The narrow stablecoin alternative is a middle course: instead of having the Fed enter the retail CBDC business, it would have it offer its wholesale accounts and services to a broad set of retail digital currency or stablecoin providers—and not just to insured banks and thrifts. The result would be a more competitive and contestable digital currency market and one in which separate stablecoin suppliers would find it easier to interoperate with one another as well as with ordinary banks.

The narrow stablecoin option could be partially implemented without new legislation. Many licensed banks are already allowed to issue digital currency. And according to Peter Conti-Brown, the Fed is legally obliged to grant master accounts to them, even if they aren’t insured.²¹

But as we’ve seen, the Fed has been unwilling to grant master accounts to uninsured institutions. Congress should insist that it obey the law. The Federal Reserve Board is now developing master account application review guidelines for the reserve banks to follow. These guidelines should rule out arbitrary refusals or delays. As I’ve suggested elsewhere, they should also expressly provide that no master account sought for by an eligible, prospective narrow stablecoin issuer be denied so long as the applicant meets other requirement no more onerous than those that insured banks must meet.²²

But the Fed needn’t insist that uninsured stablecoin issuers always maintain 100 percent reserve backing of their currencies. It could safely relax the rules a little, making the stablecoin market that much more competitive, by allowing issuers to partially back their coins with short-term Treasury securities, while including them among counterparties that can use its new standing repo facility to rapidly convert those securities into reserve balances should they ever need to do so.²³

Requiring the Fed to adopt master account guidelines that are friendly to narrow stablecoin issuers and already legally eligible for such accounts is only a first step toward achieving a truly efficient and dynamic digital currency market. Unless the set of firms legally eligible to apply for master accounts is itself expanded, the U.S. digital currency market might still be excessively concentrated and inadequately contested.

That licensed banks alone are eligible for Fed master accounts is a legacy of the days when ordinary banks and thrifts were the only suppliers of private-market dollar-denominated payments media. It’s because these old-fashioned suppliers of substitutes for fiat money “bundle” the provision of payments media with risky lending that such banks can be a source of systemic risk. The Federal Reserve Act’s denial of Fed bank master accounts and Fed wholesale settlement services to all firms save licensed banks takes the inevitability of such bundling for granted.

But new technologies can make the “unbundling” of the various services that ordinary banks offer not only possible but highly efficient.²⁴ This unbundling allows specialized financial technology firms (fintechs) to compete with banks in lending without supplying any sort of digital payments media—or to compete with them in supplying digital payments media without doing any lending.

Although they aren’t banks, fintechs that supply payments media, including digital currency, can benefit by having the same access to the Fed’s accounts and services that ordinary banks enjoy. Today they can only access the Fed services indirectly through bank correspondents. The extra cost of doing this is another important digital currency market entry barrier. More precisely, and more importantly, it’s a major barrier to entry into the market for safe digital currency.

Fintechs’ entry into that market can be relaxed somewhat by making “bank” licenses available not only to ordinary banks but to providers of unbundled bank-like services. Wyoming’s Division of Banking and several other state banking authorities have taken this step, but as we’ve seen, the Fed still hasn’t granted a master account to any Wyoming special purpose depository institution.

The Office of the Comptroller of the Currency (OCC) has likewise made way for unbundled “banks” through its own special purpose national bank program. This program, however, was held in abeyance by lawsuits, all of which have since been dismissed. The program may therefore attract some applicants at last. The suits’ dismissals may also encourage the OCC to revive its 2020 “payments charter” proposal, an alternative to the special purpose national bank charter designed exclusively for fintech payments firms. Congress should encourage these OCC attempts to help fintechs, including actual or prospective narrow stablecoin issuers, gain access to the Fed.²⁵

But Congress shouldn’t stop there. It should also amend Section 13 of the Federal Reserve Act to allow the Fed to grant master accounts to nonbank fintechs. That section’s opening sentence begins, “Any Federal reserve bank may receive from any of its member banks, or other depository institutions, and from the United States.”²⁶ The amendment could simply replace “other depository institutions” with “other payment service providers.” But to counter reserve banks’ abuse of their discretion, it should go further by changing “may receive” to “must arrange to receive” or some equivalent language.

Although granting narrow stablecoin issuers and other uninsured fintechs direct access to the Fed may seem radical, it’s a less radical step than granting that access to individuals. It’s also a step several central banks have already taken. In 2018, the Bank of England, which long allowed only a handful of banks to have settlement accounts with it, made such accounts available to fintechs for the express purpose of making it easier for them to compete with banks.²⁷ The Bank of Lithuania did much the same thing when it established its CENTROlink payment system around the same time.²⁸ Both programs have been quite successful.

Assuming that the proposed reforms allow a highly competitive stablecoin industry to flourish, we must ask: What more could a Fed CBDC contribute? Last August, Fed Governor Waller asked the same question. After critically assessing every possible advantage put forward by CBDC proponents, Waller remained “skeptical that a Federal Reserve CBDC would solve any major problem confronting the U.S. payment system.… Government interventions into the economy should come only to address significant market failures.” Waller saw no reason why the private sector should not be able to achieve all the benefits the Fed might achieve by issuing its own digital currency.²⁹ I share Waller’s opinion, subject to a qualification: in my view, the stablecoin market is unlikely to fail, unless the government doesn’t allow it to succeed.