Banks have failed to exorcise their technical gremlins

January 29, 2014 5:05 pm

Banks have failed to exorcise their technical gremlins

By Martin Taylor

The systems architecture of the typical lender is like a gothic house of horror, says Martin Taylor

Ido not know why Lloyds Bank’s computer systems crashed at the weekend. I do not know why Royal Bank of Scotland’s systems crashed last year. I do not know much – anything – about computers, frankly. So why am I writing this piece? Because I do not entirely believe the simple story that it is all about under-investment, and that the banks are useless at IT.

In the mid-1990s I had a position of responsibility at Barclays. This is such a long time ago that in any other industry my reminiscences would be irrelevant to today’s problems. But retail banks today are horribly like they were 20 years ago. This vulnerability perhaps makes my memories more relevant than they might otherwise be.

It was my unfashionable view in the mid-1990s that Barclays and its competitor banks were rather good at IT – it was one of the few things they did really well. I did not say this often for fear of universal derision. (No one who works at the Financial Times says: “Isn’t it amazing that we manage to publish a newspaper every day?” Believe me, it is amazing.) For banks, IT is what consultants call a “core competence”. Which simply meant that you had better be good at it or you might as well pack up. And we were, all of us, fairly good at it.

How can I say that? Well, when you consider the volumes going through the payments system, the gigantic number of credit card transactions, the volume of cheques cleared (now there is a technology that ought to be put to sleep) and the infinitesimally small number of errors, you have to accept they are doing something right. No one thanks you, of course, when their salary arrives in their bank account, though they scream blue murder if it does not. Success goes unnoticed and unrecognised. This crucial economic function is in general carried out to an enormously high standard, especially when you compare it with the catastrophic performance of almost anything in finance with a human component. (I know this is unpopular but it happens to be true.)

In the end the gigantic mainframe-supported systems of the past, a crucial barrier to entry into banking markets, may become an intolerable weakness

In the 1990s IT did not enjoy especially high status within banks. Real men, then as now, wanted to meet property company bosses at the Savoy Grill and lend them huge amounts of money that were unlikely to be repaid. Insiders took the computer staff as much for granted as outsiders did. Yet even then there was very little to a bank but its systems, its regulatory licence and its reputation for confidentiality and competence. Wise bankers need to take care of all of these.

Investment in IT had to ensure there was enough capacity to meet demand. One autumn day in the mid-1990s the Barclays systems came close to a complete crash because the marketing team in charge of signing up student customers had hit on an unexpectedly alluring special offer and – to its incredulous delight – had taken double the bank’s usual market share. Unfortunately there was not enough computing capacity to process the new accounts. With shameless disregard for natural justice, marketing blamed the whole fiasco on IT.

Over time the systems architecture of the typical big bank, especially if it has grown through merger and acquisition, has departed from the Palladian villa envisaged by its original designers and morphed into a gothic house of horror, full of turrets, broken glass and uneven paving. Pull it all down? Well, one day, yes. In the end the gigantic mainframe-supported systems of the past, a crucial barrier to entry into banking markets, may become an intolerable weakness.

Already in the 1990s the idea that technological disruption might undermine the established operators was very much in the air. Unbelievably from today’s perspective, telephone banking was seen as a game-changing threat. (It was simple, though, for the incumbents to install copycat versions). Then came the pure virtual banks, some of which turned out to have pure virtual capital situated in pure virtual jurisdictions.

These days, starter banks should be able to provide a much better customer experience than the incumbents, with systems that can be bought pretty much off the shelf. Customers are getting used to the idea that their accounts might exist only in cyberspace. And the big banks have indulged in an orgy of brand-value destruction, which may end up having consequences well beyond even the stupendous conduct costs they have racked up.

Ironically, the regulatory pressure that the banks have brought upon themselves may have made the industry just unattractive enough to deter really ambitious new entrants. As the economy recovers, though, and the interest rate environment normalises, all this may change.

In 1995 I visited the Barclays IT headquarters in Cheshire. On no account was I to wear a tie, they told me; geeks were free spirits. I went in an open-necked shirt and found they were all wearing ties. They would not take them off, either. Even then, you see, it was hard to get right.

The writer is a former chief executive of Barclays