Exhumed from the GSTPA’s ashes

In the opening plenary of Sibos 2005 in October, Swift chairman Jaap Kamp unveiled how Swift intended to become more commercially focused in the next five years.

One strategic option, according to Mr Kamp, is “to restart the initiative of the securities industry to strengthen global straight-through processing – a sort of Global Straight Through Processing Association (GSTPA) revisited – and align this initiative with [financial market participants group] Giovannini’s requirements”.

This may be laudable perhaps, but does the industry need another GSTPA? Simon Cleary, head of securities markets at Swift, says: “Swift standards and messaging capabilities are actively used across payments, FX, treasury and securities processing around the world. Other providers provide services that support straight-through processing (STP).

“We take a responsibility as an industry co-operative to be the glue that brings various processes and capabilities together to offer a viable STP offering. We’re reviewing this in the context of our 2010 strategy with our board and community to determine specific opportunities and partnerships to best deliver STP. In particular, it is clear to us that, in the securities markets, the post-trade space is one of the key drivers of STP, and this will be an area of specific focus.”

In the beginning

The GSTPA was formed in the late 1990s by a group of banks in response to the lack of transparency and the inefficiencies inherent in the process of buying and selling securities, between fund manager, broker and custodian. The amount of re-keying and faxes, it was thought, was making the process costly.

In December 2002, after the banks had spent €100m on building a transaction flow manager (TFM) to enable central matching, the GSTPA closed up shop due to low take-up by the fund management community and because it had run out of money. The GSTP company running the technology was later sold off to an employee of the company that had won the contract to build it, TKS-Technosoft.

After losing the contract to build the technology that the GSTPA would use, Thompson and the Depository Trust & Clearing Corporation joined forces to build a rival central matching engine, Omgeo, using a slightly different model and building on the fact that many fund managers were already using Thompson’s software, Alert and Oasys.

For some time the industry debated the need for two central matching engines and gradually decided that Omgeo would mainly cover the US and the GSTPA would cover Europe. However, once the GSTPA closed its doors, there was not exactly a rush to the only central matching facility in town. Nine months after the GSTPA closed, Omgeo launched with just five pilot firms and as recently as March this year has finally partnered with six major brokers to launch another version of its central matching engine with a lighter interface, and another pilot.

So why is Swift seeking to resurrect something along these lines again? The grand vision of Swift is that by 2010 it will shift its focus from simply messages to end-to-end transaction management. Alongside that, it is Swift’s general desire to grow its business to include investment managers, and hedge funds. In short, Swift, a co-operative of banks, still believes there is money to be made from automating the buy-side.

Ed Neeck, vice-president in charge of network management at JPMorgan, says: “The basics of STP are still what we really need as an organisation from a global custody standpoint. We need our clients to come into us electronically. Some of our clients have automated fax facilities and they see themselves as automated but [the information] isn’t when it comes to us.”

JPMorgan has been working closely with Swift to find better ways to automate its securities client base. Although multiple solutions must be used for different clients, depending on the size of the organisation, the smaller firms that are still reliant on faxes and manual entry want a single standard that they can use for all their providers.

 Price is right

 When GSTPA and Omgeo were on the table, fund managers were accused of taking a “wait and see” attitude and seemed reluctant to go with either matching engine. Today, little has changed. Much of the problem is achieving a light enough interface that is attractively priced for the fund manager. Swift Lite, and its attempts to woo investment managers, has not been a roaring success either.

“The answer must be to go back to some of those ‘lite’ solutions and to some other solutions that we have been talking to Swift about, and to automate whatever means the investment funds use to communicate and put it in the Swift format,” says Mr Neeck. “Even if it comes down to fax translators that can take text and put it into a message – we are looking at all options.”

Mr Neeck believes that Swift can and should play a role and can offer some practical benefits to automating the process. However, he agrees that, at the end of the day, price and cost to the investment manager are what will matter most. He believes that there may be little or no investment needed on behalf of the investment manager, with the brunt of the cost being picked up by those who stand to benefit most from the efficiencies gained: the global custodians and the broker dealers.

“GSTPA didn’t work because the market couldn’t see the benefits of matching versus the initial cost of outlay to implement it,” says Mr Neeck. “Even today, Omgeo has not seen full take-up by the investment managers, and FIX [Financial Information eXchange – a series of electronic specifications used for trade-related messages] is only really used by the large investment managers. But the problem of STP has still not been fully solved, especially on the settlement side, and that at some point has got to involve Swift.”

Back to square one

While GSTPA was being built, the Securities Industry Association set a deadline for the industry to move to T+1, in the hope that improved processing of securities would enable the securities industry to move to shorter settlement times. The deadline came and went, however, and the initiative was quietly abandoned. After all the predictions of how much would be spent on new STP technology and how much would be saved, little changed. The fund management community continues to use faxes, despite the promised saving of $1.74 for every allocation that went through the GSTPA’s TFM.

According to Andrew Muir, head of financial solutions at Singularity, an IT consultancy in London that connects financial institutions to Swift, the shift in Swift’s focus, from message to transaction, will enable the co-operative to effect greater change on the straight-through process.

“The GSTPA itself is dead, of course. The phoenix that is starting to rise from its ashes is the idea of Swift tracking, monitoring and eventually controlling whole end-to-end transactions, which is pretty much what the TFM was designed to do,” says Mr Muir.

“There will be significant differences in method. Swift will be the engine but not the database. It will do the validation of messages with reference to a transaction model but will not generate those messages, and will probably not own reference data sources as part of its remit. Even so, there are some parallels with GSTPA at the overall ‘vision’ level that are laudable.”

To many, STP neither died nor went away but is still very much on the agenda: every financial institutions’ budget includes it, under a different name: the need to cut costs by doing things more efficiently. The industry continues to grapple with it and yet no broker has ever dared to levy a surcharge on fund managers that are not automated, or turn business away because of the hidden costs in the efficiencies of communicating orders.

Whether Swift can come up with a new way of dressing up STP and finally achieving it will depend completely on getting the buy-in from the investment manager.