From the June 2007 issue of Treasury & Risk magazine

Wired Efficiency

Like most treasurers, Debbie McSheffrey, treasurer of Atlanta-based Premiere Global Services Inc., can tell you: While sending money by wire turns out to be the fastest way to move money, it also turns out to be the slowest--and most imprecise--way to deliver usable remittance information. Applying wire payments is extremely manual and error-prone, notes McSheffrey, who is a member of the payments advisory group of professional organization Association for Financial Professionals (AFP). "It's a real challenge to figure out what a wire is paying," she says. "A parent may pay on behalf of a subsidiary. One wire may pay multiple invoices from multiple divisions. Too often we have to call the sending bank or our customer to try to resolve what is being paid, which is a waste of time on both ends."

And if there is a dispute, the system is even more cumbersome and expensive. When a wire payment is delayed pending investigation, it can trigger a late payment fee or deny a prompt-payment discount, McSheffrey points out. While those penalties could be reversed, that requires even more manual intervention, she adds.

All this may be about to change. After years of grumbling by corporate treasury staffs--and pressure from the AFP-- the two domestic wire networks, Fedwire and the Clearing House Interbank Payments System (CHIPS), are expected to announce by the end of June a dramatic overhaul in how remittance information is transmitted. The new approach would afford wire payments the same kind of processing by computer and automatic posting common for many ACH and check payments that go through lockbox operations and then arrive in electronic transmissions.

Presently, any remittance information goes into a free-text field that accommodates up to four lines of 35 characters each. Even if the information is supplied, it's so unstructured that it has to be read and applied manually, explains Arlene Chapman, senior consultant for technical services at the AFP. In June, the wire networks are scheduled to sort out options for structuring and standardizing remittance data and announce a plan for fixing the problem.

There are three leading candidates, according to Ken Isaacson, assistant vice president and Fedwire Funds Transfer product manager at the New York Fed, who heads up the project, along with Hank Farrar, senior vice president of The Clearing House, responsible for CHIPS.
1. The STP 820 has widespread support, but it would take time. "The STP 820 would require us to add fields to our existing message format, but not completely rewrite it. It would take many months, if not years, to do that, but many corporations are already receiving 820s and we're still talking to software vendors about how easily they could support it," Isaacson says.
2. The ISO 20022 would require a complete rewrite of existing message infrastructure, which could take as long as a decade. A subset of the ISO 20022 is mandated in Europe for SEPA (Single European Payments Area) payments and used for some other retail transactions but is not yet used by high-value networks, Isaacson says.
3. A simpler, quicker, cheaper solution would be to keep the existing free-text message field and get all players to agree to a "usage convention" specifying the order in which remittance information would appear. They also would agree to certain machine-readable prefixes to identify data elements, Isaacson explains. "That would take almost no change on our part," he says. "We could do it in a day, but other parties would need to make modifications in how they order and code the information."

AFP and its members are pushing the STP 820, a stripped- down version of the EDI 820 already widely used to carry remittance information with ACH payments, says Anita Patterson, director of treasury operations at Manheim, the Atlanta-based auto auction house and chair of the AFP payments advisory group. "It's a more corporate-friendly variation," she says. However, some bankers and staffers at CHIPS and the Fed favor the ISO standard because it is global, Patterson adds.

There could be two winners. "The short-term solution could be domestic, but eventually we'll need a global solution," Farrar explains. "We have an increasingly global economy, and a big part of the increase in cross-border payments will be wires, so global interoperability will be necessary."

While the most conspicuous beneficiaries of structured wire remittance data would be receivers of wire payments, senders and banks also have a stake in supporting straight-through processing since they also get drawn into manual investigations, AFP's Chapman notes. Companies of all sizes in all industries get wires and support the project, but those with the heaviest inbound wire traffic have the highest cost of manual processing and would save the most money, she explains. With improvements in remittance information and increased straight- through processing, banks could improve revenue if wire payments become more attractive.

There also would be a fraud-control payoff. Statistics show that check fraud is rampant, while wire fraud is rare. The more wires displace checks, the lower the fraud losses would be for corporations and, especially, for banks, Chapman suggests.

A fix will require more than just action by the wire networks: Payers and receivers, the sending and receiving banks and the software companies that provide the corporate and bank systems all have to support any solution with changes to their own systems. So, while June will mark a significant move in the right direction, a final fix will take a while longer.

While implementing standards may be daunting, most agree that it must be tackled. CHIPS and the Fed commissioned a survey of 381 corporations, augmented by eight focus groups. Sentiment was clear. "They told us they want more structured, standardized information," Isaacson summarizes. "94% said it would be valuable; 65% indicated that it would be very valuable and that they would be willing to pay about $1.67 more per wire to get it." And an AFP survey showed that 80% said they would be "very likely" to supply such information if the fields were there to carry it, Chapman reports.

"We got the message," Farrar says. "We need a consensus solution that all parties will say 'yes' to quickly and get on with the business of making it work."