Want to Save a Billion Dollars?

Saving a billion dollars should excite management, yet there appears little enthusiasm for the call recently by one of the major banks for their competitor institutions to share IT. Admittedly the call came from an investment bank, but it could easily apply to the custody world as well. IT expenditure in the custody business globally can be estimated at around $2-3 billion per annum. And much of it is duplicative to the extreme.

We have high-profile areas where we share IT already. SWIFT is the classical, and most visible, example, but usage of all market infrastructures invariably involves some IT and business process outsourcing.

The advent of cloud computing will surely be a catalyst for further IT cooperation. At the least it may lower the barriers to entry for new players, or existing ones without a mega IT budget appetite, by giving them access to sophisticated, and especially niche, applications.

There are two fundamental barriers to IT sharing. The first is human. Even organizations find it hard to move to single platforms for like functions. This is partly because silos triumph over group value, and partly because owners of extant technology refuse to move to common operating models with their peers elsewhere. The second fundamental problem is the delusion that IT is a major differentiator; it is not. The core differentiators in banking are service quality, people support, product reach, geography and value for money. To some extent these are linked to size of IT wallet as each need powerful IT tools.

But technology could be changing accessibility to these. Component-based technology platforms are now the norm. Securities services technical architecture is not about a custody platform or a fund administration platform, but about a series of modules that can be found in many other business areas: trading, asset management, derivative clearing, risk management, etc. The moment we all moved to that modular architecture, we created the potential for shared remote applications. And the individual applications, if they meet demand from different market segments, could be delivered at a price per measure of use that makes smaller players as effective as some of the larger ones.

Such a modular vision does, though, require the right delivery mechanism. SWIFT, especially, and other vendors such as Reuters or Bloomberg, have colonized the financial sector connectivity or information spaces. Yet they have not really penetrated the e-banking world. From a message content perspective, standard messages prevail, although there still is the ridiculous, and unnecessary, incompatibility between key vendors in some core areas. A key strategic issue for SWIFT, or one of its current or future competitors, is whether it can, or wishes to, extend its reach into the core client e-delivery and intranet facilities of its users.

The challenge is substantial, for such connectivity would demand the creation of key partnerships between application vendors and the network providers, far beyond the impressive list of approved vendors operated by SWIFT. But such connectivity becomes logical if cloud, or its next generation, applications are to become the standard in the market.

The value for users is undoubted. Duplicative technology has no real sustainable competitive value. It may reduce competition between small and large players, but that differentiation, as I noted earlier, is unlikely to be sustainable for the long term. It is a value of the last century, perhaps, or, at least in our fast-changing technological age, of the last decade! Duplicative technology does not allow major firms to differentiate or gain competitive advantage for the very reason that it is duplicative!

And the further problem of the growing modularity and specialization of much modern IT architecture of full securities service firms is that it is hard, and extremely costly, to ensure resilience and latency across the multiple gateways and diverse applications of the modern day hi-technology environment in which we operate.

Looking at infrastructure we can note two trends. Generally speaking, the post-trade infrastructure has worked well and managed the stress of highly volatile throughputs in markets. Trading markets have had more problems, mainly because time is of greater essence in such environments. As we look at shared technology, a rational approach is needed. The answer to shared technology is not a cloud, or son of cloud, containing an end-to-end securities infrastructure. Cloud applications (or similar shared IT platforms) have to be logical. Intuitively, shared technology should be built on a progressive basis, with non-cash-linked, such as regulatory reporting, applications being first to market. But long term, in the securities services world, there is little differentiation in core SMAC, international market fund administration or derivative clearing applications. And the change suggested is more in IT architecture than IT applications; the cloud and other advances change the mechanism for accessibility or delivery rather than the tools of production.

At a time of sharply falling margins, settlement turnover declines, resulting from de-leveraging and the rise of central counterparties as well as low-term asset appreciation or new capital formation, a radical rethink is needed. The choice for the industry is simple. Look at imaginative ways to reduce cost and potentially business risk from development problems, or just revert to the time-honored method of cutting the IT development budget, reducing staff and skills. And also, do learn from history and be patient, use competition and avoid ten-year public sector mega projects that become a money-guzzling pit for little value added.

The challenges are huge. The global prize, across banking, could be a multi-billion dollar saving of development spends. And, even more significantly, we could see a trend to shared technologies enabling markets to reach the elusive nirvana of greater harmonization.

But how do we get there? And has the industry the imagination and courage to consider such a quantum step?