Friday, March 20, 2009

Payment Convergence- Vision or Hallucination?

According to a recent study by McKinsey, the U.S. payments industry is a $282 billion business. In a report entitled "Weathering the Storm: Global Payments 2009", The Boston Consulting Group estimates the global payments business at $805 billion, ramping to $1.4 trillion by 2016. It goes without saying that payments is big business. But is big beautiful? It could be, if those that ascribe to a converging payments paradigm are right.

Silo'd be thy name. Payment during early times was as simple as trading bartered goods or exchanging coins. The last two hundred years have seen an explosion in the many ways value can be transferred, including cash, checks, wire transfers, automated clearing houses (ACH), card payments in various hues, mobile- the list is endless. Each of these evolved separately to meet a specific need, creating distinct processing systems and organizations in their wake. Today's financial institutions have multiple systems, with separate rules of governance, and organizational fiefdoms that resist attempts to blur boundaries.

Holy Grail anyone? The idea of a unified infrastructure to process all payment types has been mooted for decades. It is argued that, at a notional level, there is little difference between "capture-validate-clear-settle" in check-speak, and "acquire-authorize-switch-settle" in card-talk. So, would it not be simpler, as some would argue, to have one system that did it all? But given that we have spent the better part of a century perfecting these older systems, does it makes sense to go where angels fear to tread? If it is not broken, why fix it?

Can I call you Sybil? The world, unfortunately, is not that simple. Checks can now be converted to ACH payments forcing an erosion of previously impervious walls. In an even more bizarre twist, checks can be imaged, transmitted, and reprinted as paper substitute checks. Newer forms of payment like mobile and prepaid use the debit or ACH rails as the basic underpinning for moving money. Automated bill payment is essentially an Internet front end to ACH transfers. With payments morphing from one to another with the skill of Dickens's artful dodger, financial institutions are pressed to ensure profitability across payment channels, adherence to disparate rule sets, and risk management that spans silos.

Brave new world. The 21st century has ushered in the need for compliance with a dizzying array of legislation. There is the Patriot Act, Sarbanes-Oxley, Basel II, and OFAC- to name just a few. The post bailout era will likely herald significant new legislation, if I am reading the winds from Washington correctly. The cost and complexity of updating legacy systems one at a time to ensure compliance is prohibitive. Moreover, it is difficult to find skill sets to modify dated systems within an acceptable lead time.

If there is one reason that trumps all others, it is the need for improved customer service. Customers are decidedly unsympathetic to the self-inflicted tribulations of their financial institutions. They need to be able to interact through branches, call centers and the Internet to get information across accounts and payment vehicles. The new "millenial" generation is not likely to have the patience to wait while an operator logs in and out of multiple systems.

End or the beginning? While few would argue against the case for convergence, there are practical considerations to be addressed. Does one begin with the settlement end of the value chain and work one's way forward? The move towards Real Time Gross Settlement (RTGS) systems, particularly in emerging economies, suggests that "the end as a beginning" idea has some takers. Nevertheless, I suggest that the greater payoff is at the other end of the telescope. The conversions from one payment to another take place closer to the point of origination, and that is where the greatest benefits are to be garnered from a unified infrastructure. It is also where the functional and technical challenges are acute.

Buckets or pipes? While there are varied perspectives on how to get there, all are agreed on one thing- do not try to redesign existing legacy systems. There are two approaches to a convergent payments platform. There is the data-centric model based on a central repository for all payments information, and customized one-to-one interfaces with the various payment and core processing systems. An alternate approach is message-centric with a central hub through which all communication between payment, core and other systems is routed. The nirvana is a combination of both, governed by versatile business rules engines that financial institutions can control.

Does size matter? The technology vendors in this space are faced with an interesting choice. Does it make sense to target large financial institutions where the payoff from convergence is likely to be greatest? Or is it better to focus on smaller institutions, where implementations of this kind are not akin to open-heart surgery? The answer depends squarely on the vendor's positioning and business model. If the major contribution to profits is from systems integration and professional services, the high end segment suggests itself. On the other hand, if the model is tilted towards license sales with a modest service component, the lower end would make sense. Technology vendors would be well advised to pick their poison. It is a toss up as to whether it is harder to scale up or scale down. They are both incredibly difficult. This is an instance of clear positioning and alignment at the outset being critically important.

Back to the future. It is perhaps counter-intuitive to suggest that the way forward involves re-engineering legacy processes with the introduction of new convergent systems. Nevertheless, evidence indicates that there is an emerging synthesis between checks and ACH to be followed by online debit. The path is fraught with challenges, some of which I have touched upon.

A mentor of mine once told me that there is a very thin line separating a vision from a hallucination. Where do you think payment convergence lies? Let me know.