Stick around long enough and old ideas become new again. Such is the case with adopting real time core processing in commercial banks.
It’s not a revolution, not even a “groundswell.” It’s simply that with the advance of technology and communications, and changing consumer behavior, the notion of real-time processing has crossed a threshold separating “nice idea” from “becoming reality.”
This distinction may seem puzzling to some banks and vendors. The term “real time” has many meanings, some of which are narrow and relatively simple, some quite rigorous. So, for example, if you define real time as providing retail customers with updated balances as soon as any transaction is completed, whether that be at the branch, an ATM, using a debit card, or on the web, there are banks doing this now, even though the balance in question is a memo post, to be made final later. If you limit the definition to teller transactions being instantly updated to other channels (but not necessarily the other way around), then possibly a majority of banks are “real time” in that sense.
On the other hand, if you define real time as handling any transaction at any channel only once, at which point final posting occurs, then very few U.S. institutions are doing that. It is more common overseas and is used by some direct banks in this country. Real time was also the norm for years with savings and loans and credit unions, and still is in some cases, though the movement to check-based transactions changed things for many of these institutions.
Technology consultant (and bankingexchange.com blogger) Dan Fisher, who was chief information officer at Community First Bankshares, Fargo, N.D., before it was acquired by BancWest, maintains that real time is a lot more perceived, from the customer’s point of view, than actual, from a processing point of view. Most banks, he says, still batch process, and rely on ATM and debit card processors to send a refresh file, sometimes only once a day.
“The reality today is that real time is more of a perception than a reality,” Fisher, president of The Copper River Group, wrote in his March 4 “Beyond the Bank” blog. “It appears to be real, but every now and then the customer is adversely impacted and does not like it.”
24/7 versus batch
Checks and the elaborate apparatus to clear and settle them determined the structure and direction of U.S. banking technology more than any other factor. But as the industry has embraced image processing, remote deposit capture, and branch capture, and as the public has become accustomed to using debit cards, the physical aspect of check clearing has declined sharply.
Combine that with the surging interest in tracking balances via website or mobile device (and the expectation that those balances will be up to date), and you have a powerful force pushing banks to upgrade the traditional way they handle payments and account information.
“Overnight batch mode doesn’t align well with anywhere, anytime banking,” observes Don Free, banking industry research director for Gartner.
By moving to real time, banks could also gain efficiencies, reduce errors, and reduce fraud in contrast to the traditional batch-mode systems that have evolved in patchwork fashion, particularly at the larger banks.
But the process will be neither quick nor easy, for while real-time processing is simple in concept, it is devilishly complex to implement in a large institution. Yet virtually everyone contacted for this article agreed that the conversion is only a matter of time, and is already happening. Case in point is United Bank of Michigan, Grand Rapids.
Fully real time (or close to it)
Dan Fisher’s comment about customers being adversely impacted by incomplete balances was a big reason why United Bank began moving toward a real-time system several years ago.
“One of the biggest complaints we would get from customers was when they would make a deposit at a teller and then go to the ATM later in the day, or even the next day, and see a different balance,” says Tim Lockwood, senior vice-president and chief information officer at the $428 million-assets bank. That prompted a gradual conversion to providing real time balances across all channels—branch, ATM, debit, online, and mobile—which was completed five years ago.
“A large segment of our customers are very concerned about their balances,” says Lockwood, “they like to be in touch.” The difficult economy has only added to this concern.
Currently tellers at some United branches process items at the counter—scanning the items and balancing the transaction—but for other branches with larger check volumes balancing is still done centrally. All checks are truncated at the branches, however.
“We call it real time because we’re [immediately] posting the transaction to the customer’s account,” says Lockwood.
He adds that ACH in-clearings and checks can’t be classified as real time at this point.
This progression did not require a core processor conversion. The bank has been using the Fiserv Premier package, which accommodates real-time updates. Debit card and ATM processing, handled by a different vendor, did require software conversions, he notes, and also required dedicated phone lines.
In addition to Fiserv, other bank IT vendors offering real time processing systems include Fidelity National Information Services, Harland Financial Solutions, Modern Banking Systems, and Open Solutions.
Real-time myths addressed
Frank Sanchez is well-known in banking technology circles as a real-time evangelist. His old firm, Sanchez Computer Associates, developed a real-time system called Profile for large banks. He sold the company to Fidelity National Information Services in 2004, but did not abandon the quest. As president of Fidelity National’s Strategic Solutions division, he continues to be responsible for the Profile system.
Sanchez says “real time banking” to him indicates that a transaction is authorized against a system of record and posted against a system of record at the point of sale, and that no subsequent manual processing is required. This applies equally to payments and nonpayment transactions.
U.S. customers of the Profile system include such direct banks as ING Direct, Charles Schwab Bank, E-Trade Bank, and Ameritrade. “Direct banks see the world in real time,” says Sanchez. He acknowledges that these banks have few if any branches, and fewer transactions than typical banks, but “at the end of the day you have to look at the fact that they have greater efficiencies.”
Having to present a real-time interface to customers in multiple channels is “causing a lot of stress in the U.S. banking system,” Sanchez maintains. “A lot of infrastructure has been built on top of batch systems of record.”
Sanchez ticks off several myths surrounding real-time banking:
Myth One is that there would be a loss of float through elimination of delayed availability. “That’s not the case,” he says; “real-time posting doesn’t mean real-time availability.” Clearing records can be entered as part of the transaction or placed in a proof hold/check hold file.
Myth Two is that if a small business customer comes in with a sack of checks, the teller would have to stop and scan 1,000 checks. In that case, says Sanchez, the teller would simply post the transaction and send the checks to the back office, although he says this issue is becoming less of a factor because of remote capture.
Myth Three: batch and online can’t mix. Modern real time systems, says Sanchez, understand that there are still checks, ACH transactions, and lockbox items that come in as a batch, and that this is likely to remain the case for the foreseeable future. But real-time and batch can coexist.
Myth Four: going real time means you can’t balance, because there is no cutoff to apply to the general ledger. Sanchez’ response is that “there is always a balance” with real time. “Every transaction has a debit and a credit, so you can run [a balance] at any time of the day, like a just-in-time inventory system.”
Sanchez says that the biggest financial benefit from a real-time system would be a decrease in research, reconciliation, and manual processing of exception items.
“Straight-through processing,” he adds, “is a de facto benefit of real time.”
Spend money to lose money?
The discussion about real time has shifted in recent years from what’s technologically feasible to the business case for the investment.
“Changing core systems is not for the faint of heart,” notes Bart Narter, senior vice-president of the banking group at Celent. “You’ve got to have pretty powerful reasons to do it.” The analyst believes there would “absolutely” be back-office efficiencies gained from moving to a real-time, straight-through processing system, most notably the elimination of “all the gyrations in software to make [channels] seem real time.” But a big strike against doing so has been, and continues to be, the question from management: “Why would we spend money to install a new system that reduces revenue?” Meaning: reduction in check-related fee income.
“Banks haven’t figured out the fees,” says Don Free, bluntly, referring to replacing overdraft income. In batch mode, Free explains, banks can sort transactions low to high, high to low, or in check order. Typically they pay high items first, which are likely to be the most important, but which, if an account is under funded, will also tend to create more overdrafts. Free says that in real time that option would be eliminated, and banks haven’t figured out how to replace that income.
Nevertheless, checks are disappearing, and the pressure to move to real time core processing continues to grow, according to Narter. He ticks off two pressure points: 1. the growth of real-time transactions—debit cards in particular, and the risks associated with such payments (once you say ‘yes,’ it’s nonrefutable, even if the transaction doesn’t clear till later); and 2. the demands of consumers for real-time balances.
One other factor offsetting possible fee income loss is reduction in fraud losses.
“Criminals go point to point withdrawing cash,” relates Jim Sizemore, senior vice-president, Bank Solutions Group, Fiserv, alluding to the vulnerability of batch systems. There are safeguards, of course, but real time would reduce that exposure.
Batch and POD won’t disappear overnight
Sizemore believes the majority of banks are already at what he calls the proof model of real time, that is, transactions done at the teller or ATM are reflected immediately in the customer’s balance. Transactions still go through proofing at night, however. The advent of branch capture and merchant capture and image processing has practically eliminated the need for the proof system for physical items, he says, but end-of-day proofing continues to be used to identify stop payments, check signatures, and handle exceptions.
Tom Berdan, vice-president of product management for Harland Financial Solutions doesn’t think most banks are ready yet to skip the proof stage for most transactions, even though they are migrating pieces of their business to real time, which Harland’s Phoenix platform can handle. In the case of a small business initiating an ACH transaction that can be structured to flow right to the core system if it meets certain requirements, “all of the banks I’ve talked to have said, ‘I still want to have some verification’,” says Berdan.
“I think there will still be a batch process for a number of years even though we can do many of the pieces in real time,” Berdan concludes.
Sizemore’s view is that a big transition will come in the next three to five years as banks move to straight through processing.
“We actually believe it will be more pervasive than just transactions,” he says. It will include all the events that happen in the bank—account opening, inquiries, and maintenance.