In the aftermath of the events of September 11, 2001, and regulators' publication of the Interagency Paper on Sound Practices to Strengthen the Resilience of the U.S. Financial System, the Federal Reserve Bank, like other operators of high-value payments systems, began to focus even more intensively on preparedness, continuity and resilience issues. Telecommunications was obviously a key area of concern. As a result, the Bank's Wholesale Product Office (WPO), which is responsible for the Fedwire Funds Transfer Service (Fedwire), developed a proof-of-concept project to use SWIFT's IP-based SWIFTNet services to enhance the network backup capabilities for accessing Fedwire. Rather than reinventing the wheel, the WPO looked beyond its own resources to develop a solution that would leverage investments that payment industry participants were already making. The project, dubbed the "SWIFTNet Pilot" by the WPO, provides a connectivity contingency alternative for Fedwire's largest participants.
The WPO is targeting the SWIFTNet Pilot to the 350 banks that access Fedwire via a computer interface as well as the largest 100 to 150 FedLine banks. These banks, which transfer hundreds to tens of thousands of Fedwire messages per day, represent the greatest risk to the payments system should they be unable to access Fedwire. To put this in perspective, 100 of the 9,300 Fedwire participants now control close to three-fourths of its volume and value -- and concentration is greatest among the largest 25 institutions. The decision to focus on the largest institutions is a break from the WPO's historical practice of ensuring that initiatives benefit the largest number of participating financial institutions. In this case, however, connectivity -- not the application -- is at issue, so the initiative can be limited to the largest Fedwire users without impact to other participants. These target banks are virtually equivalent to SWIFT's U.S. client base, and the Pilot offers them an opportunity to "reuse" their SWIFTNet connections to access Fedwire. Based on these factors, the WPO last year planned, developed and tested a prototype and received endorsement of the concept from key banks.
According to the WPO, 99.95 percent of all Fedwire traffic moves across FEDNET, with access by either dedicated frame relay connections, with ISDN dial backup, or by dial-up connections. Larger banks spread their traffic over numerous connections for contingency reasons. In theory, these banks have line diversity; nonetheless, at the back end some vendors have lines going into the same place. A high-volume bank, therefore, could be left with a large exposure and little recourse in the event of a FEDNET connectivity failure.
The SWIFTNet Pilot Solution
It is this potential gap that the WPO plans to rectify with the SWIFTNet Pilot. The solution will enable large Fedwire participants to originate and receive Fedwire transfers via SWIFTNET using its multi-vendor secure IP-based network (SIPN). The WPO will operate as a SWIFTNET member-administered closed user group (MA-CUG) to control access and services. Banks must develop the capability to reroute payment from their FEDNET connection to the SWIFTNet one in the event of a problem. Moreover, routing Fedwire transfers via SWIFTNet, rather than FEDNET, is not a transparent process and requires that the Fedwire-formatted messages be wrapped in an XML envelope. Banks must use specialized delivery software (developed by SWIFT) to accomplish the XML-wrapping/unwrapping process and to ensure guaranteed delivery and acknowledgement between SWIFT's and Fedwire's very different environments. For the most part, however, the solution relies on standard SWIFTNet technology including SWIFT Alliance Gateway (or an equivalent vendor product) and SWIFTNet Link software. Banks must have both of these products in place to migrate to the IP-based network.
Resilience was the project's major driver. The solution not only had to satisfy the WPO's own standards for Fedwire resilience but also meet the expectations of significant payment clearing banks. In addition, the solution had to conform to Fedwire's business continuity architecture and align with the payment and technology strategies of the target Fedwire banks. SWIFTNet is a prime component of this strategy for many of them. Given that close to 30 percent of Fedwire traffic has either a foreign originator or a foreign receiver and offshore payment transfers will increase once the Fedwire processing hours are expanded in May 2004, the WPO rated globalization and scalability as important decision factors as well.
These considerations become more important if banks decide to use the new access method on a regular basis rather than as backup only. The WPO anticipates that many banks implementing the solution will use the SWIFTNet connection as a delivery mechanism for some portion of their daily Fedwire traffic to increase overall resilience. Doing so makes sense from a cost as well as a contingency perspective -- it lowers banks' FEDNET charges by letting them take down at least one backup connection and potentially reducing ongoing SWIFTNet costs at the same time.
Rollout depends on two major variables: timely deployment by the largest volume users and completion of SWIFTNet migration by the rest of the target population. The pilot has already been delayed by competing priorities such as SWIFTNet migration issues, regulatory considerations and payment system upgrades. Several of the very largest Fedwire payments clearing banks have signed on as pilot customers and are in various stages of readiness. The WPO anticipates that they will be live with the project before year's end 2004.
The WPO deserves credit for thinking outside the box and looking for a workable solution that is practicable for, and provides value to, its largest banks. The SWIFTNet Pilot is a sensible idea both for the Federal Reserve Bank and for those Fedwire participants that can take advantage of the SWIFTNet "channel" via the WPO's CUG. The solution provides a resilient, readily accessible alternative to meet the contingency objectives for Fedwire connectivity. There is little practical downside because the solution uses SWIFT's industry-standard messaging to augment Fedwire's resilience. In February 2004, the Clearing House announced it would allow SWIFTNet access to its new proprietary TCP/IP network for CHIPS (its high-value interbank payment system) and EPN (domestic ACH) payments, which should increase the value proposition for the largest banks.
This article is based on TowerGroup research by Breffni McGuire, a senior analyst in the Global Payments practice at TowerGroup, a leading advisory research and consulting firm focused on the global financial services industry. Ms. McGuire can be reached at [email protected]