Consumers and businesses around the globe demonstrate a growing interest and need of having funds real-time which only gets aggravated in a cash-strapped economy. Overall real-time payments volume globally grew 63.2% in 2022 to reach a new high of 195B and are projected to hit a mark of 500B+ by 2027*. Upcoming launch of FedNow by the Central Bank is the latest entry in the domestic US market and hopes to see a staggering growth of volume in coming years while current systems, Zelle and ACH-led RTP, clock a miniscule share of the market at 1.2%.

Real-time payments not only serve the need of speed, transparency and efficiency, it is significantly cheaper compared to most of the traditional payment rails. While accepting a card transaction charges a variable basis point to merchants, RTP’s low flat fee offers a greater cost benefit to merchants and institutions which is of great incentive in the current cost-constrained economy. 

Challenges with Real-time payments – 

1. Irrevocable – Real-time settlement and fund transfer do not allow for Refund and Disputes; so cost of fraud is extremely high. 
2. Fragmented – Real-time payment rails are localized and even multiple rails operate in the same country, eg RTP (ACH) and FedNow don’t have interoperability built in, which means you can’t initiate a payment if sender’s and receiver’s FIs are not participating in the same real-time rail. FIs need to integrate with multiple rails for allowing better access to its customers.
3. Cross-border payments – Though India recently announced a move towards enabling fund transfer across the border from a few countries like Singapore however it is a long way to go before cross-border remittances becomes mainstream. As small businesses incrementally do business globally, having real-time funds across the border & eliminating the risk of Fx would be a great boost for their working capital management
4. Government participation – In absence of Government intervention to adopt real-time use cases through G2C or G2B payments or vice-versa like benefits disbursements, insurance payments or tax collection (eg UPI based payment in India) it is hard to achieve the sweet-spot

Real-time payments rails follow ISO20022 message format which offers two-way, data-rich messaging and is slowly becoming the de facto standard for payments message across schemes. It comprises of ~250 structured data elements and Features include – 

• Payments and addendum (additional customer/transaction data) as a single set of transaction 
• Extensive data fields accommodating relevant detailed and supplementary data for remittances , eg purpose & source of the payments, beneficiary details etc
• Message to sender and receivers about transaction completion status
• Sending bills & invoices in pdf/xml format along with payments messages
• No loss of data through the end-to-end payment leg

Exchange of rich data in real-time payments essentially creates an opportunity for the industry to build a more secure and resilient system while creating value for participants through –  

• Strengthening Fraud models 
• Improving straight-through processing
• Multirail strategy with intelligent routing
• Generating new revenue streams

Strengthening Fraud models 

Payments fraud has always been one of the most widely used topics and is of greater importance as real-time rails work on first-time-right principle. Use of AI in tackling frauds is nothing new however a fraud model based on trained datasets from supervised learning would not be comprehensive enough to arrest malicious attempts by full-time hackers as new forms of payments open up opportunities to create new patterns and layering. In addition, legacy fraud prevention strategies typically generate too many false positives that result in costly manual investigations. Hence, a few options that would help elevate trust on the RTP* ecosystem:

• Reinforcement learning – In comparison to supervised or unsupervised learning, reinforcement learning approach has no labels or clusters, but rather simulates episodes that produce strings of rewards which are used as signals for continuously updating the model. Its ability of simulating events and micro-experiments would help construct a more robust fraud model that can protect against unknown situations in a non-revocable payment environment
• Learning Hub – A federated model where all the rails contribute to build a central learning hub by sharing their learnings, keeping data security and regulation in consideration, would complement effort at individual level and will reduce the cost of modeling for everyone. This might need regulation by a central body but would open up immense potential for strengthening steps against fraud and AML across payment rails
• Generative AI – In an ever-evolving fraud world, it is important to build ML models using synthetic data to infuse robustness using variability of patterns. Generative AI can learn the statistical properties of real transactions and generate synthetic data to train the model in anticipation of newer fraud patterns.

Improving straight-through processing

Operational reconciliation in the ERP system is usually cumbersome and manually-intense activity for banks that includes complex business logic and additional remittance information arriving asynchronously makes it even tougher and time-consuming. A Fiserv study shows, an automated solution can quickly reconcile a large volume of transactions, 70 million per day or more, with a high match rate in the mid-90s.

   RTP complying with ISO20022 gives an opportunity to simplify and automate the reconciliation process with categorization of messages as they are grouped into domains and makes it easy to differentiate messages; eg pain.013 indicates a request-for-payment vs pacs.008 represents a credit transfer; so it is easier to identify origination of payments instruction although both would credit and debit same accounts. Each transaction detail uses xml tagging which enables STP processing because of its structured nature.

Multirail strategy with intelligent routing

Fintechs are building APIs for integrations to real-time rails and faster access. As real-time rails don’t have message exchange interoperability, for wider adoption of instant payments in B2B or C2B scenarios, integration with multiple APIs is needed to provide frictionless experience to customers. In the event of a disruption, there must be a fallback mechanism to meet customer desire of transferring funds instantly. Co-existence of real-time systems and traditional rails like ACH or wire need an effective multi-rail strategy to route transactions to the best optimized path. 

IPR (Intelligent payment routing) drives immense value, in particular for account payables (AP) in B2B scenarios, by leveraging the information associated with payments messages and recommending the right-fit, fastest and cheapest payment rail. Moving away from static business rules and using rich information embedded in ISO20022 messages to build an intelligent ML model-based routing would help alleviate the risk of payment failure and would enhance experience for FIs, business and consumers.

Generating new revenue streams

Majority of current use cases for real-time payments have been confined to P2P or B2C use cases and banks have mostly adopted RTP for retail payments. However, B2B payments is a nearly tapped area to be disrupted with instant payments use cases like direct-bank transfer and request-to-pay. Large banks mostly implemented ISO2002 but yet to reap full benefit of the rich data-led intelligence in customizing the product-offering or creating hyper-personalized campaigns. 

  Real-time fund transfer may not be a great source of revenue but the immense opportunity for value-added services to customers or merchants would potentially open up new revenue streams for banks. RTP could be a great alternative for subscription based services as compared to direct debit as it can offer better liquidity to billers and greater transparency and flexibility to consumers. Data rich RTP systems offer better reporting and analytics which could be extremely useful offerings to SMBs to achieve real-time intelligence for their operational or marketing decisions. 

Adoption of real-time payments would require substantial education, interoperability and top-down regulation for this game-changing innovation to become mainstream. While real-time transfer is a great benefit to customers or businesses on its own, a layering through value added services would create greater incentive and use cases for accelerating use of RTP. Building capability to leverage the data gold-mine from ISO20022 at its full potential and driving insights and prediction through smart AI modeling would be potential answers to some of the key concerns industry is challenged with.

* Reference: https://insiderealtime.aciworldwide.com/prime-time-report-23

*RTP acronym has been used to indicate “Real-time payments” in a few places. To specify ACH RTP rail, same has been mentioned explicitly

Disclaimer: This article is solely an individual perspective and does not represent or is influenced by any organization

• SEO Powered Content & PR Distribution. Get Amplified Today.
• EVM Finance. Unified Interface for Decentralized Finance. Access Here.
• Quantum Media Group. IR/PR Amplified. Access Here.
• PlatoAiStream. Web3 Data Intelligence. Knowledge Amplified. Access Here.
• Source: https://www.finextra.com/blogposting/24364/real-time-payments—a-data-goldmine-to-unlock-opportunities?utm_medium=rssfinextra&utm_source=finextrablogs