Why haven’t Instant Payments grown faster?
Instant Payments were too late for consumer use cases and too early for B2B
Early in my JPM tenure I was involved in Real Time Payments (RTP) at The Clearinghouse (TCH). I was also in Zelle meetings where we contemplated replacing ACH settlement with TCH RTP. My McKinsey colleagues helped write the early white papers on why the U.S. needed instant payments and I often worked with those authors. So, I have relevant insight into why things worked out as they have.
Back in 2014, we all expected a truly real-time payment system would reduce fraud and settlement risk, lower payments costs for everyone, and help automate the financial supply chain. The Fed was vocal in encouraging TCH to take this on. TCH RTP has been live since 2017 and growth has been high, but from a frustratingly low base. Other countries are seeing their instant payment systems capture more share, faster. Why?
Timing is everything. Other U.S. payments networks carved away or retained the high-volume use cases of P2P, B2C (disbursements), and C2B (bill pay & commerce). For B2B, the US business community has not embraced the ISO 20022 standard that is a clear differentiator for Instant Payments – so the B2B use case has selective uptake. Finally, despite all its early cheerleading, the government slowed things down in a variety of ways.
This post will walk through the competitive situation, use case by use case, and address the other obstacles. At the end, I will speculate on future adoption scenarios. First, we need to understand why Instant Payments are different from incumbent payment methods.
What makes Instant Payments different?
A word about nomenclature. Real Time Payments (RTP) is both a generic name for Instant Payments and the brand-name for TCH’s implementation. FedNow is a real-time payments system but obviously not “RTP” the brand. In this article I will use Instant Payments as a generic term for both TCH RTP and FedNow. I exclude Zelle, Same-day ACH, and the push-to-card systems. They are all faster than incumbent methods, but lack certain features that are embedded in RTP & FedNow.
Instant Payments are the first totally new payments rail in decades and are digital from the ground up. Such payments have features that collectively distinguish them from incumbent payment methods:
Real-time clearance and settlement. This feature eliminates counterparty risk, which makes it very attractive to regulators. Wires share this feature but that rail is based on 40+ year-old technology, is expensive and cumbersome. Cards have a real-time authorization to make sure funds are available, but settle next day. ACH is entirely batch – even the same-day version. Zelle provides real-time availability but settles next day
Push-only. This is also a risk management benefit. Users initiate Instant Payments from their bank’s digital channels. Since banks do strong KYC and generally have strong cyber controls, these payments are very secure. They are subject to account-takeover fraud and second-party fraud, but more secure than incumbent methods
Built-in ISO 20022 messaging. Most payment methods carry very little data in their settlement message – usually only what is required to process the payment. ACH can send addenda messages with more data, but those are batch and they need to be re-associated with the payment transaction. 20022 is an XML-based standard that global businesses are adopting to automate the financial supply chain
Low-cost. Instant Payments are far less expensive than cards or wires, equivalent to same-day ACH, but more expensive than standard ACH. That still improves economics for many use cases
Irrevocable. Once the sender approves the transaction, it is final – just like wires. That contrasts with “returns” in ACH and “chargebacks” in cards.
These features collectively address many friction points in incumbent payment methods, but, in some use cases, these advantages have associated drawbacks. We will explore these by reviewing each major use case.
Person-to-person payments (P2P)
Countries that have high Instant Payments adoption often start with P2P; it provides high volumes and gets consumers habituated to a new payment system. However, the U.S. got P2P before Instant Payments were ubiquitous. Instead, the banking industry chose to develop a different real-time availability method for P2P. This was necessary to compete in P2P but created a long-term headwind for Instant.
Zelle was incubated at The Clearinghouse but didn’t stay there. The sponsoring banks wanted a P2P system that would anchor consumers on bank digital assets rather than losing that engagement to fintech P2Ps. JPM, BAC, and WFC already had on-us P2P systems that were very popular and growing fast. The other banks generally used Fiserv’s POPMoney, which settled via ACH and charged a transaction fee. The big 3 plus a few others, including Zelle founders USB & COF, had already built ClearXchange (CXC) to get interoperability between all their on-us P2Ps; but CXC settled via ACH with next day availability.
ACH was inexpensive and reliable, but next day availability was an obstacle. Fintech P2Ps gave the illusion of real-time availability because they used book transfers to move funds from one customer’s stored-value account to another’s. The funds couldn’t be used for anything but more P2Ps unless those funds were repatriated to the customer’s bank account – which happened via ACH with next-day settlement. It simulated real-time availability but wasn’t.
Despite high growth, bank P2Ps were losing share to fintechs. In May of 2015, TCH created a task force to build a competitive, industry-wide P2P system by year end. We managed to build the Zelle legal entity and governance system by that deadline, but the Zelle service didn’t launch until 2017 – still pretty fast. Zelle was developed at EWS, another bank consortia, which had more development capacity than either TCH or CXC. CXC was absorbed into EWS as part of Zelle formation.
TCH RTP launched at almost the same time at Zelle, but while Zelle started out with 60% of US DDAs fully connected, RTP took much longer to reach that threshold. Zelle also provided a mechanism to reach non-member banks, while RTP did not.
Only the biggest banks implemented RTP send & receive right away, while most banks started with receive-only. It took several more years to get to that 60% threshold and most smaller banks still aren’t connected.
Zelle achieved real-time availability in an ingenious way. Member banks gave each other a “good funds guarantee” where the receiving customer got access to their funds immediately, while the receiving bank was paid via ACH. Since P2P flows tend to balance bilaterally, the overnight exposure to any one counterparty was limited. The guarantee had been incubated at CXC but never implemented. It gave Zelle a key advantage over Fintech P2Ps – the money is received in your bank account so it can be used for any purpose right away, not just for more P2Ps.
Another key Zelle advantage was its “alias directory” (another CXC innovation). Zelle payments are sent to an email address or a mobile number, not a complex bank account number. The directory associates the email or mobile number to the bank account in the background. The Fintech solutions use aliases of their own, but Instant systems require a DDA account number. That makes them cumbersome for end users.
Zelle’s success starved RTP of a major volume source. The Zelle owner banks were leading TCH members and did this with their eyes open. Later, Zelle allowed banks to settle via RTP rather than ACH, but only a few pairs chose to do so. More may follow, but until that happens, Instant Payments needs to look for volume in other use cases.
Disbursements (B2C)
As I described in the article on Networks, disbursements is a key scrum among payments networks. The card networks, ACH & Zelle all targeted B2C before Instant payments got to scale:
The card networks launched Visa Direct & MasterCard Send around 2015, shortly after TCH announced the RTP initiative. The card solutions leverage mature debit rails to move payments from the “B” to the “C”. They use the consumer’s 16-digit debit card number as the payment account identifier. It is more expensive that ACH or Instant Payments but much less than wires. The card solutions also work for cross-border payments wherever Visa & MasterCard run debit networks
ACH was the incumbent in B2C. For example, Direct Deposit is a disbursement as are pension payouts, dividend payouts, insurance claims payouts, tax refunds, etc. The major friction point is onboarding the DDA account number; but, where there is a recurring relationship, that friction is overcome by the low-cost, universality, and reliability of ACH. Onboarding the account number is also getting easier with Open Banking technology
Zelle launched a B2C option. Zelle not only offered real-time availability, it allowed payouts using the consumer’s email or mobile number. That eliminated the friction of using a long DDA account number or a 16-digit debit card number. Uptake had been slow as the price was relatively high, but prices have come down and usage is growing from a low base
In contrast, Instant Payments are not universal as most smaller banks haven’t integrated yet. Instant also requires the more complex DDA account number rather than email, mobile number, or card number. Disbursements have a light data load, all transactions are already push, and irrevocability isn’t an issue. Even speed isn’t that critical as payouts can be timed to arrive on schedule. None of the unique Instant Payments features differentiate for basic disbursements. And Instant systems don’t work for cross-border payments at all.
Instant Payments have traction in some niche use cases that aren’t pure B2C. Examples include:
Marketplace payouts, where the real-time feature allows intra-day payment. Card networks are less relevant here as most businesses don’t have debit card numbers; the card solutions are also more expensive than Instant. This volume is typically “B2b” – where the “b” stands for small business
Earned Wage Access (EWA). In EWA, an employer advances an employee their accrued wages on demand. Instant works here because the employer has the DDA account number of the employee for direct deposit. Real-time matters because the advance is typically to cover some urgent need. This is really B2B2C as a payroll company usually executes EWA between the employer and the employee
Bill Pay & Commerce (C2B)
Consumer payment methods are entrenched in C2B. Instant has made minimal progress here, but for different reasons in Bill Pay and Commerce.
Bill Pay. ACH (Direct Debits) serve the bulk of the bill pay market, although checks are still relevant. Cards have traction in subscription businesses and in some other verticals – but critically, not in banking where cards are not accepted for loan/card repayments, deposits, investments, etc.
ACH is low-cost, reliable, and universal. While ACH has a “returns” issue, its other merits make it fit-for-purpose. Direct Debits are a “pull” method, i.e., the biller pulls the funds from the consumer’s bank account. Instant lacks such a mechanism until the Request for Payment (RFP) function reaches ubiquity. In that model, the biller sends an RFP to the consumer with full invoice details embedded in a 20022 format. The consumer can then approve an Instant Payment from inside their bank’s digital experience.
While consumers could “push” Instant Payments to pay their bills, the only advantage is the ability to time a payment for the last instant, e.g., before a late fee is charged. For regular, on-time payments, there is no consumer advantage over ACH or cards.
Instant with RFP is also ideal for collections. During a collection call, the agent can initiate an RFP to the consumer in real-time and the consumer can approve an irrevocable Instant Payment while still on the call. However, RFP is not yet implemented by most banks.
Commerce. Cards aren’t going anywhere in U.S. commerce because Instant payments don’t pay rewards or offer credit. RTP could displace some debit volume but it is hard to see how merchants incent consumers to switch – and the savings are not as material relative to debit costs.
Another key challenge is irrevocability. While merchants love this, it conflicts with consumer protection regulations. Consumers have the legal right to refunds for a variety of reasons (e.g., never delivered), but Instant has no processes to support that. The Card system has Chargebacks and ACH has Returns for just that reason.
Business-to-Business (B2B)
B2B should be the core Instant use case. The key alternatives are ACH, Wires, & Checks. Wires are expensive, while checks are slow and fraud-prone. Neither of these methods carry material data. ACH can carry data in “addenda” records, but that is a cumbersome way to operate. The key advantages of ACH are low cost & ubiquity. For mass payments, it is hard to beat – and the slower settlement actual benefits the payer. Further, we now have Same-Day ACH which narrows the settlement gap with Instant – for many users, same-day is fast enough.
Commercial Cards account for a fraction of B2B volume – but that volume is most often C2B-type transactions where the “C” is an employee of a “B”. For example, Corporate cards count as B2B, but they are used primarily for business travel by employees.
“Virtual cards” (VC) have some traction in Accounts Payable, but this is a niche as VCs are costly to the receiver; In contrast, buyers like them because they get a share of interchange. Some buyers will pay faster if the seller agrees to accept a VC. Where a big buyer is dealing with small sellers this can be win-win. But most B2B commerce is among large companies, most of whom will not willingly accept VCs given the cost.
The challenges to Instant adoption will take time to resolve.
Transaction volume. The huge B2B dollar volume translates to much lower transaction volume as average ticket sizes are big
Limits. Instant Payments have a $1 million limit that excludes big B2B transactions
Data. ISO 20022 has not been embraced in the U.S., so the data advantage is theoretical
Ubiquity. While all large banks have embraced TCH RTP, most small banks have not. The small banks are adopting FedNow, but full adoption will take time. Since large companies tend to use large banks, most transactions among large companies can use Instant, but many transactions to/from smaller businesses cannot
Costs. Instant Payments are more expensive than standard ACH and roughly equivalent to Same-Day ACH – so only those transactions that absolutely, positively need real-time can be cost-justified without considering other factors
Request for Payment (RFP) is not widely available. That means certain B2B use cases are not yet accessible
Domestic only. Today, instant systems are not interoperable across borders. So if a use case requires money to cross those borders, Instants are hard to use. Central Banks are trying to integrate, but that is still in the future
All these challenges will work themselves out over time, but another obstacle is that the U.S. now has two Instant Payment systems, FedNow and TCH RTP; those systems are not interoperable and FedNow is not yet fully implemented by its intended audience. We will discuss these challenges next.
The Fed emphasized competition over time to market
I will not comment on whether we should have two Instant systems. The Fed decided we should. In most countries the Central Bank runs the Instant system, so a system like FedNow is the norm. We have done fine with two ACHs and two wholesale methods (Fedwire, CHIPS), so why not two Instant systems? It is a matter of public record that the smaller banks lobbied for a Fed alternative, as they had issues with the TCH approach – and our regulators have always balanced the interests of small banks and big banks.
However, FedNow development slowed down TCH RTP adoption. Mid-sized and small banks preferred to wait for FedNow, which took several years. Larger banks delayed other Instant Payment investments (e.g., RFP) to focus on integrating FedNow alongside TCH. Companies put off adoption as ubiquity seemed further off. Interoperability between the two Instant systems is also an issue: How to take a transaction initiated on TCH RTP to reach a bank that is only on FedNow or vice versa.
Finally, the government diminished the appeal of the C2B Commerce use case by undermining irrevocability. On this topic Zelle is the canary in the coal mine. Regulators and Congress accuse Zelle of not adequately deterring fraud and scams. Fraud happens when a bank account is taken over by a bad guy (ATO) and scams happen when a conman dupes a consumer into sending them money. Fraud can also happen when a consumer shares their digital credentials with a trusted party and that party abuses the privilege (second-party fraud).
ATO is clearly the bank’s responsibility; but scams and second-party fraud are a tougher call. In both cases, the account owner did something they were warned against – either sending money to a stranger or sharing their digital credentials with anyone else. Nevertheless, Washington expects the banks to make the customer whole. The solution is both greater bank vigilance (which raises operating costs) and chargeback processes (which raise operating costs). Zelle responded, but neither Instant system has those capabilities yet.
Can we accelerate progress?
There is no magic bullet to resolve the challenges, but time will address some of them. Smaller banks are joining FedNow, helped by their core processors. Over time this will address the ubiquity problem. Bigger banks are joining both systems.
Interoperability can be addressed by big banks offering correspondent services. The big banks could switch messages from one network to the other when the sender and receiver banks are not members of both networks. This is relatively low risk as the inbound payments are real-time and irrevocable and will always come from a chartered depository. It adds complexity and cost however – the opposite of what Instant is designed for. The Fed could also offer this service.
The government could use Instant routinely for outbound payments. In other countries, government adoption and cheerleading are common (e.g., India’s UPI). Governments’ have big B2C flows (e.g., Social Security) and B2B volume (e.g., Medicare claims). It is not clear if agencies are waiting for FedNow to gain traction or whether they have the same, long implementation lead times as the private sector.
The government could also begin cheerleading for ISO 20022 adoption by businesses, just as they originally championed Instant. If the government itself started requiring ISO 20022 to do business with it, the standard might be rapidly adopted – and with it, greater use of Instant.
The Instant networks could compete with ACH by lowering prices. Instant switch fees cost 4-5¢, much costlier than ACH, particularly for large users. That price gap has limited Instant to niche use cases where speed or irrevocability are required. If Instant is to win share in B2C it needs to be price competitive for routine payments and not just the exceptions. Similarly, more Zelle banks might switch to RTP settlement if the price were lower.
Request for Payment could open up bill pay and other use cases. This is an investment area that may not have a business case in the short run but could create new volume long term. Given where we are in the adoption cycle, this will take considerable time before it has an impact.
All the ideas above have a long runway. In the shorter run, Instant will remain a niche payment method; so why not embrace the niche approach and get smaller pockets of volume across a breadth of use cases?
For example, an early use case was intraday marketplace payouts. Fintechs were monetizing by allowing their sellers to get paid intraday, with a fee of 1%. Where the seller was an individual, like a ride-share driver, Visa Direct & MasterCard Send are the incumbents. But more formal businesses don’t have a debit card number to receive into, and those payment methods charge over 10¢ per item. Instead, marketplaces started using Instant Payments at half that price. Note that marketplaces had the bank account number of the seller for normal settlement — so no onboarding friction.
Instant is also ideal for “me-to-me” payments where a consumer is moving money among their own accounts at different financial institutions. For example, moving money from a checking account to an investment account. The consumer knows the account numbers at both ends so onboarding friction is low.
Another set of niches are what are known in the acquiring industry as “high-risk”. These are lawful businesses that have high chargeback rates. Here, irrevocability is as important as real-time. One example is legal online gambling. While legal, many of these businesses entail “headline” risk; but they need an irrevocable, instant method.
Conclusions
It is frustrating that a superior payment method can’t displace less capable incumbents. Today, Instant Payments feel like Betamax where all other methods are VHS. But the Betamax story is often mis-told. Betamax did have better picture quality, but it lost on recording time – it could record for 1 hour while VHS could do 2, the length of a movie. Consumers cared more about recording time than picture quality so VHS won.
To win, Instant needs to find use cases where it is clearly superior in ways that customers value. Accept that it is more expensive than ACH; instead focus on cases that need data, irrevocability, or speed. Compared to other push payments, it is less expensive but requires an RT + Account identifier rather than a debit card number – so focus on payouts with a business at the receiving end. Etc. This whack-a-mole approach can chalk up small victories while the ecosystem matures.
Key insights
Instant Payments are growing fast, but from a low base. Not all banks have integrated and the two Instant systems are not interoperable.
Each major use case faces adoption obstacles:
In P2P, Zelle is instant as far as users are concerned. Zelle would need to convert to Instant settlement among its member banks to capture these volumes – but this is nascent
In B2C, all payment methods are competing for volume and each has advantages based on requirements for speed, price, data load, & exception processes. Instant has a few niches where its proposition wins, but this does not account for high volumes, yet
In C2B, the incumbents have entrenched economic advantages:
In Bill Pay, billers prefer low-cost ACH, while consumers prefer Cards for rewards. RFP would improve attractiveness, but is not widely deployed
In Commerce, consumers prefer Cards for rewards and borrowing capacity; Instant lacks a chargeback feature for consumer protection
In B2B, Instant will eventually win, but faces obstacles today
U.S. businesses have not adopted ISO 20022 – which differentes Instant
Limits are too low, ceding large transactions to wires
Fees are higher than ACH for transactions that don’t need real time
Small banks are slow to adopt as they were waiting for FedNow
The two Instant networks are not interoperable
Neither system works for cross-border payments
Mass markets are out of reach until the ecosystem matures, but Instant can serve niche use cases where one of its unique features are valued
UK is interesting case study in having to slow down speed of Instant payments. RTP scams led UK Payments to put in a speed bump when adding new payees rather than instant and irrevocable.