Round Two: What Role Should the Federal Reserve Play in Developing a Faster Payments System?

Friday, April 26, 2019
Authors: 
Brian Knight

This is round two of a multi-part debate series on The Bridge. The debate consisted of two rounds of email exchanges between three participants. An introduction was published Wednesday, April 24, and round one was published Thursday, April 25. The series has been lightly-edited to preserve the original spirit of the email conversation that took place.

Jim Angel

We are in broad agreement: the US economy badly needs faster payments and the Federal Reserve should play a large role in facilitating faster payments. The devil is in the details of what role the Fed should play.

Aaron seems to believe that the Fed can wave a magic regulatory wand under the Expedited Funds Availability Act and make faster payments happen. Were it so simple!

It is not clear at all that the Fed actually has regulatory power to just mandate faster payments. I believe Aaron is referring to 12 US Code § 4002(d)(1) which reads

“Notwithstanding any other provision of law, the Board, jointly with the Director of the Bureau of Consumer Financial Protection, shall, by regulation, reduce the time periods established under subsections (b), (c), and (e) to as short a time as possible and equal to the period of time achievable under the improved check clearing system for a receiving depository institution to reasonably expect to learn of the nonpayment of most items for each category of checks.

Alas, the referenced subsections refer only to check clearing, not wires or other non-check forms of payment. Alternatively, the Fed could use its broad rulemaking authority under Dodd-Frank §805 over payment system risk management and rightly declare that slow payments impose risk on the financial system. However, some may view this as a regulatory overreach.

Even if the Fed did have the authority to just mandate faster payments, a mere rule from the Fed would not be enough. The Fed itself is one of the biggest impediments to faster payments. The Fed is where the different payment providers can settle their payments. Even the bank-operated check clearing and ACH networks depend on the Fed for part of their activities. The Fed’s systems that allow banks to settle payments with each other operate on 20th Century East Coast bankers’ hours, Monday through Friday.

In order for our banking system to speed up and offer true 24/7 service to everyone, the Fed has to operate 24/7.

In order for our banking system to speed up and offer true 24/7 service to everyone, the Fed has to operate 24/7.

George agrees that the Fed should offer 24/7 settlement services and a 24/7 liquidity management tool, but disagrees on whether the Fed should offer a 24/7 RTGS facility. I think that here we have different visions of what the Fed has actually proposed. I see the Fed’s proposal as a modern 24/7 version of Fedwire that would allow competing payment providers to interact with each other on a real-time basis. What is the problem with letting Fedwire operate 24/7? If competing payment providers cannot settle 24/7, they have to work out a system of deferred net settlement in which risk piles up until settlement occurs. Such accumulation of risk is totally unnecessary.

As far as legality goes, only the Fed can offer a service that settles in Fed money at the Fed with no counterparty risk. Thus, the Fed’s proposal easily meets the legality requirement that other providers cannot provide the service.

Rather than slow down faster payments, letting the Fed provide needed infrastructure to competing payment will allow numerous entities to innovate and provide 21st-century service.

Aaron Klein

The Federal Reserve has the legislative authority, what it has lacked is the will. Under the section of the Expedited Funds Availability Act that James cites, the Fed could mandate checks to clear within an hour, as well as cash deposits in ATMs (that is covered by subsection e). In addition, the Fed has broad authority under Section 15 of the Check 21 Act, which states: “The Board may prescribe such regulations as the Board determines to be necessary to implement, prevent circumvention or evasion of, or facilitate compliance with the provisions of this Act.” It is harder to get broader language than that, which is not surprising given that the legislation is based on a proposal from the Fed.

One of the three enumerated purposes of the Check 21 Act is “To improve the overall efficiency of the Nation's payments system.” Combining that broad authority and the purpose of the legislation, Congress has made it clear the Fed has authority. There are probably other regulatory hooks the Fed has at its disposal as EFAA provided and the Check 21 Act affirmed, Congress “provided the Board of Governors of the Federal Reserve System with full authority to regulate all aspects of the payment system.” The Fed has the authority.

We could have a real-time payment system for consumers without the Fed itself operating every aspect of it. That said, there are some critical areas where the Fed needs to up its game, such as making Fedwire 24/7x365, a proposal that I think all three of us agree on. In fact, the slow nature of Fedwire also reduces the ability of securities firms to settle, a separate but also important issue.

James and I are in agreement that the Fed itself is one of the biggest impediments to faster payments. George makes the critical observation that as the Fed continues to play Hamlet, debating whether to create their own system or not, has practical effects on other banks decisions to join alternative payment systems. Economists should be very familiar with uncertainty delaying adoption. The Fed has had the choice to adopt real-time payment technology for over a decade (the UK moved in 2008), formed multiple faster payment groups, and issued multiple studies over many years. The private sector moved forward, the Fed did not. The Fed’s waiting has taken billions out of the pockets of middle and working class families. The Fed’s delay has increased and continues to increase income inequality.

The simplest, most efficient, and fairest thing the Fed can do is use their own authority and change the system now.

Americans who can least afford it are stuck paying billions a year in fees and high-cost loans to access their own money. The simplest, most efficient, and fairest thing the Fed can do is use their own authority and change the system now. As George points out, waiting for the Fed to build its own system will take years, maybe a decade or more. Unless the Fed is willing to use its own funds to pay the tens of billions that it will cost working families to handle that delay, then the Fed needs to fix the problem the best way it can: through adopting regulations requiring real-time payments.

George Selgin

Like Jim, I’m encouraged by the many points on which all three of us agree. I hope by my response today to narrow our remaining points of disagreement still further.

Jim observes, in reply to Aaron, that the Fed may lack the power to simply mandate faster payments. He concludes therefore that it can’t “wave a magic regulatory wand” to get us there. But while both statements are strictly true, the Fed could speed things up considerably by encouraging more financial firms to take part in TCH’s private RTP set up, instead of doing just the opposite, as it does by holding out the possibility of establishing a competing fast payment network—that is, by “playing Hamlet,” as Aaron eloquently puts it. It could, as we all agree, offer 24/7 settlement or liquidity management services.

It could allow RTB member account balances to count towards banks’ LCR requirements, while also allowing interest on those balances. It could encourage further widening of the RTP Business Committee to assure adequate representation of smaller banks. Finally, to encourage participation by certain non-bank payments service providers, it could allow, and could encourage TCH to allow, some of those suppliers to participate in the RTP network.

Jim points to certain advantages of the Fed’s proposed RTGS system, consisting mainly of the fact that it eliminates counterparty and other credit risks inherent in deferred net settlement arrangements. But while RTGS on the Fed’s books, rather than on those of a private clearing entity, may have certain advantages, it is not essential to achieving faster payments.

Therefore, although Jim is correct that the Fed is uniquely capable of supplying final RTGS services, it doesn’t follow that, so far as achieving faster payments is concerned, there is a compelling need for it to do so. Finally, RTGS, as an alternative to a 24/7 Fedwire (deferred settlement) system, is not free of disadvantages of its own, which consist of higher bank liquidity requirements and a correspondingly heightened risk of payment delays and gridlock. Indeed, according to some experts, drawing upon experiences with existing central-bank RTGS systems, such systems don’t really reduce banks’ credit exposure at all. Instead, they merely redistribute credit risk among various payment-system participants.

To have the Fed continue to toy with the possibility of establishing its own RTGS system today is to guarantee an avoidable delay of several years in achieving faster retail payments.

In short, while a good debate can be had concerning the merits of deferred net vs. RTGS central bank settlement arrangements, that debate is largely orthogonal to the one concerning the most expeditious way to achieve faster retail payments in the US today. The two discussions overlap in but one crucial respect, to wit: that to have the Fed continue to toy with the possibility of establishing its own RTGS system today is to guarantee an avoidable delay of several years in achieving faster retail payments.

Brian Knight

Thank you, everyone, for an incredibly informative debate. This was a great discussion of a challenging topic and the scope of issues that were brought up go to show how challenging a question this is. I want to thank our participants for graciously giving us the benefit of their time and expertise and I hope our readers have found this as educational as I have.

Photo credit: Alex Wroblewski/Getty Images

Round One: What Role Should the Federal Reserve Play in Developing a Faster Payments System?

Thursday, April 25, 2019

This is round one of a multi-part debate series on The Bridge. The debate consisted of two rounds of email exchanges between three participants. An introduction was published Wednesday, April 24, and round two was published Friday, April 26. The series has been lightly-edited to preserve the original spirit of the email conversation that took place.

Jim Angel

Thanks for asking me to begin.

  • The US payment system is slow and archaic. This imposes a tax on economic activity that affects all Americans. Other countries have instant payment systems. Why don’t we?
  • The Fed operates on 20th century East Coast banker’s hours. The fact that they are closed for the majority of the 8,760 hours is a year is a huge impediment to the launch of faster payment systems that can interconnect with each other.

The Fed has proposed two actions to facilitate faster payments in the US. From their press release, they call for

"1) the development of a service for real-time interbank settlement of faster payments 24 hours a day, seven days a week, 365 days a year (24x7x365)"

"2) the creation of a liquidity management tool that would enable transfers between Federal Reserve accounts on a 24x7x365 basis to support services for real-time interbank settlement of faster payments, regardless of whether those services are provided by the private sector or the Federal Reserve Banks."

To oversimplify, the Fed is mainly proposing to operate its existing services around the clock. This is long overdue. The first proposal is for the Fed to run a service similar to the venerable Fedwire around the clock. The second proposal would make it easier for competitive payment networks to interact with each other around the clock instead of waiting for the next business day.

To oversimplify, the Fed is mainly proposing to operate its existing services around the clock.

The Fed’s goal is to support the interconnection of private sector payment systems, not to replace those systems. Note the key word in their statement: interbank. When I served on the Federal Reserve’s Faster Payments Task Force (FPTF) it was abundantly clear that the Fed did not want to establish and run its own new faster payment system the way the European Central Bank (ECB) has.

What will happen if the Fed does NOT modernize and operate 24/7?

Numerous bank, fintech, and crypto players have launched payment solutions: PayPal, Venmo, Zelle, Square Cash, RTS, Apple Pay, and Google Pay are just a few of them. Behind the scenes, many of these systems still rely upon the slow 20th century rails of our existing payment system. Payments look instant inside these networks, but it may take days to get a payment out of the network. Most importantly, they don’t interconnect with each other. A user on one network cannot easily send a payment to a user on another network. The Fed’s proposal is to allow these different networks to meet at the Fed to transfer funds to each other around the clock. This is important for facilitating a 24/7 payment system as the different networks need to exchange funds immediately to avoid risk piling up overnight. This is also important for maintaining our global competitiveness as our financial system needs to be open when our trading partners are awake.

If the Fed does not act, we will be left with a slow, fragmented payment system that leaves out many Americans. Even worse, it is likely that network economics will kick in and leave us with one or two dominant networks that will monetize their market power to tax all Americans on every transaction while freezing out innovative new payment systems.

Our economy operates 24/7 and so should the Fed.

Aaron Klein

America’s slow payment system is a major hidden contributor to income inequality. When low-income consumers approach the zero lower bound of their bank account, myriad high costs for short-term liquidity appear that wealthier people never face. Just three of these fees, bank overdraft, check cashing, and payday loans, total over $35 billion a year. Demand for these arise, in part, from the long lag time between when consumers deposit funds and when those funds are available.

The rest of the world is far ahead of America. The UK adopted real-time payments 12 years ago. Poland, South Africa, and Mexico are already there, and soon the European Central Bank will have real-time payments. A Slovakian payment deposited in Ireland will clear before a payment from Minnesota is available in Florida. This is not a problem of technology; it is a problem of leadership.

The Federal Reserve once was a leader in payment modernization. In 2001, the Fed proposed and Congress adopted the Check-21 Act. Since then, the Fed has failed to use its legal authority to adequately keep pace with technology and benefit consumers. Under the law, the Fed has the legal authority to mandate and/or operate a real-time payment system. Instead, the Fed has done neither. That inaction continues to cost working class America’s billions a year.

Modernizing, our payment system will empower Americans to better manage their hard-earned cash and have more of it.

What should we do now? The answer is simple. The Federal Reserve should use its existing legal authority (Section 603 of the Expedited Funds Availability Act) and simply mandate real-time payments in six months. Safeguards to protect against fraud ($5,000 funds limit, existing customers only, etc.) can be reasonably carved out. Financial institutions can choose to participate in existing real-time payment systems, develop their own, or provide the funds to consumers before they actually clear (several already do). The Fed is conflicted in its dual role in operating a payment system while regulating payments for everyone. Modernizing, our payment system will empower Americans to better manage their hard-earned cash and have more of it.

George Selgin

Not so fast.

Let me first make clear my considerable agreement with James and Aaron. I agree that the slow speed of many US payments is harmful, to the poor especially, and that it should be possible for all payments to be processed in hours, if not instantly, rather than in days. I also agree that the Fed, as a monopoly supplier of final settlement services for the nation’s banks, has an obligation to reform those facilities as needed to expedite payments. Finally, I agree that it should do so in part by offering 365-day, round-the-clock interbank settlement services, either by extending the operating hours of Fedwire or by creating a special “liquidity management tool” (LMT) for the purpose.

I disagree, on the other hand, with James’ view that, to achieve faster payments, the Fed must also establish a new all-hours Real Time Gross Settlement (RTGS) facility, for several reasons:

It isn’t necessary. It’s important here to distinguish faster clearing of a payment, which makes funds available more rapidly to the payee, from faster final settlement of dues among involved financial institutions. Reducing the social costs of slow payments is a matter of arranging for faster clearing of those payments. It doesn’t necessarily require faster settlement among banks.

It’s important here to distinguish faster clearing of a payment, which makes funds available more rapidly to the payee, from faster final settlement of dues among involved financial institutions.

The proposed Fed RTGS system is uniquely capable of achieving both instantaneous clearing and instantaneous interbank settlement. But faster—and even instantaneous—clearance itself can be achieved without it. What’s more, a rapid-clearance arrangement already exists. With the Fed’s encouragement, The Clearing House (TCH)—a company owned by a consortium of large banks—launched its Real-Time Payments (RTP) system in 2014. The system maintains a pooled account at the Fed, in which all banks are able to maintain funds. Payments made within the RTP network are settled instantly on the RTP account ledger.

Although only a portion of US banks have joined thus far, in principle all might take part, thereby satisfying the Fed’s “ubiquity” requirement. And although many smaller banks have yet to join, RTP’s fee structure, which allows no volume discounts and is below the Fed’s present same-day ACH fee, actually favors them. Finally, non-bank payment service providers might take part using special purpose banking charters from the Comptroller of the Currency, though that’s likely to take some nudging of TCH by the Fed.

It may not be legal. In so far as the proposed RTGS system provides no essential public benefit “that other providers alone cannot be expected to provide with reasonable effectiveness, scope, and equity,” its establishment would be contrary to the criteria set forth by the 1980 Monetary Control Act.

It will delay, rather than expedite, the establishment of a ubiquitous faster payments network. Because the RTP system already exists, banks might easily comply with Aaron’s six-months mandate simply by joining it. In contrast, it will take the Fed several years to establish a new RTGS system. Yet the very prospect of an alternative Fed-administered fast-payments mechanism has discouraged many banks from joining the RTP network, for none wish to invest in a network that Fed actions may render obsolete.

It will stifle future innovation. While any established payment network enjoys a first-mover advantage, the fast-payments market remains both contestable and dynamic, with many players offering competing—if generally less than ubiquitous—networks. The Fed’s unique privileges, including its status as a regulator of private-market payment service providers, equip it with unique monopoly powers that may ultimately stifle competition and innovation. The history of the Fed’s involvement in check clearing offers an object lesson in this regard.

Photo credit: Jessica McGowan/Getty Images