For the past 20 years banks have constructed ever more elaborate IT systems, which have rarely boosted profits. They must now replace them with simpler, cheaper infrastructures, says Nicholas Carr.
Banks today have a welter of information systems that provide little competitive advantage. But a new era may be dawning.
In 1950, Bank of America launched an ambitious effort to automate its bookkeeping. The project came to fruition 10 years later when, in a televised extravaganza hosted by Ronald Reagan, the bank unveiled its Electronic Recording Machine Accounting computer, fondly known as ERMA. By 1962, the bank had set up 13 ERMA computing centres, servicing nearly five million current and savings accounts.
The computers proved a competitive boon. They allowed Bank of America to expand its network of branches, and hence its market share, much more quickly and cheaply than could its competitors, who still maintained accounts manually.
The successful roll-out of the ERMA system inaugurated an era of massive investment in computer hardware and software by the banking industry – an era that continues to this day. It is estimated that banks and other financial service companies spent more than $350bn on information technology during 2005. According to a recent survey by Information Week, the average bank devotes 10% of its revenues to IT outlays, far higher than the 3% spent by the average company overall.
From boon to burden
Unfortunately, little of that money creates the kind of strategic advantage that Bank of America gained through its early investment in computerisation. Most of the spending goes towards maintaining and upgrading existing systems, ensuring data security and regulatory compliance, and matching the technological capabilities of competitors. Far from being a source of advantage, IT has turned into a cost that all banks must pay but that delivers strategic differentiation to precious few.
Many banks, in fact, now find IT more of a burden than a boon. They have become bogged down in the many complex systems they have built up over the decades. Incompatibilities in hardware and software impede the sharing of data, slowing or distorting the reporting of important management information. Outdated, custom-built applications lock people into inflexible old processes. System conflicts complicate or even sabotage collaboration among units, not to mention mergers and acquisitions.
Worst of all, much of banks’ IT simply goes unused. As Tony Comper, CEO of Canada’s BMO Financial Group, said in a 2003 speech: “I’d hazard an educated guess that the vast majority of the two main end-users in my organisation, customers and employees, actually utilise about 20% of their computing capabilities – and I’m being generous here. The rest of the investment is mostly wasted.”
Mr Comper’s doubts about IT’s power to enhance performance are backed up by formal research. In 1997, Baba Prasad and Patrick Harker of the Wharton School examined the impact of IT investment on financial results. Combing through detailed data on 47 major US retail banks, they found no evidence that spending on IT capital had enhanced profitability, as measured by either return on assets or return on equity.
Indeed, they found that the spending had not even boosted productivity, as the costs of installing the systems outweighed the resulting gains in output. “The easy availability of IT to all banks implies that IT investments do not provide any competitive advantage,” the researchers wrote. “IT investment has zero or insignificant effect on bank profitability.”
A 2002 McKinsey Global Institute study of retail banks in the US, Germany, and France reached a similar conclusion. While early investments in computer automation allowed banks to reduce labour costs substantially, often giving pioneers a productivity edge, the ability to use IT innovation to gain an advantage dissipated as computerisation proceeded. The authors conclude that “most IT investments are now simply costs of doing business and not differentiating”.
Retooling IT
Today, in other words, most banks have similar systems doing similar things. Even new applications, such as internet banking, tend to be replicated and thus commoditised quickly throughout the industry. What distinguishes banks is no longer their technology but the way they manage it.
Here, thankfully, there is good news. We are in the early stages of a sea change in the management of corporate IT. Up to now, banks, like other companies, have had to construct their systems out of expensive, ill-fitting, and scattered parts. Every software application has in effect had to be hard-wired to its own hardware platform. That has led to massive redundancy and inefficiency in the deployment of IT assets.
A study by Hewlett-Packard Labs, for instance, found that corporate server computers tend to operate at less than a third of their capacity. On the labour side, too, most costs go toward simply keeping the complex gear running.
But now, thanks to cheap, standardised components and broadband data networks, a whole new model of IT supply is becoming possible. Through rapidly advancing technologies such as virtualisation and grid computing, on the hardware side, and web services and XML coding, on the software side, banks will be able to turn their fragmented, incompatible systems into a flexible, centralised utility that can efficiently deliver data and applications in many different forms to many users simultaneously.
Déjà vu
In many ways, the coming transformation in IT promises to mirror what happened in factory power production 100 years ago. At the start of the last century, manufacturers had to cobble together their own electric generators inside their plants, operating them at a high cost and at a fraction of their capacity. But when alternating-current networks allowed electricity to be supplied over a network, from efficient centralised dynamos, companies abandoned their old technology platform. Between 1910 and 1930, most manufacturers shut down their proprietary generators and plugged into the utility grid.
Information technology is very different from electricity, of course, but the economics of the supply models are very similar – and in the end its economics that determine business decisions. Already, we are seeing innovative banks beginning to capitalise on the new utility model. Credit Suisse First Boston is recoding data in standard XML formats to simplify the introduction of new derivatives products. Wachovia is using grid computing to tap into the processing power lying dormant in its existing computers, creating a cheap supercomputer to analyse complex securities trades. Scotiabank is reducing hardware and maintenance costs by replacing the PCs at its branches with inexpensive thin-client terminals, delivering applications over the internet from central servers.
These are early examples, but they indicate the kinds of benefits that the technologies of utility computing can deliver. As the underlying technologies continue to improve, a more radical consolidation of IT infrastructure will become possible.
In many cases, banks will simply buy the basic computing capabilities they require from outside IT utilities, freeing up both capital and management time. For a preview of the new world, just look at the raft of sophisticated computing services Google is supplying over the web from its utility-processing plants.
When it comes to information technology, the banking industry today stands at a turning point. The last two decades have seen banks construct ever more elaborate IT systems, which have rarely provided meaningful gains in profitability. Over the next two decades, they will dismantle these systems, replacing them with a simpler, cheaper information infrastructure.
The bottom line
The focus of IT management will shift from trouble-shooting the technology itself to connecting information with processes in a way that can generate the kind of strategic differentiation that leads to outstanding profits.
With its intense transaction-processing requirements and its large pools of capital, the banking industry was the pioneer of the early computerisation of commerce. In coming years, it can be a trailblazer once again. It can lead business into a new world where computers serve employees and customers rather than frustrate them – a world where information technology is flexible, efficient and largely invisible.
Nicholas Carr is the author of Does IT Matter? Information Technology and the Corrosion of Competitive Advantage. He writes a blog at roughtype.com
