Replace Those Aging Core Systems Now

Register now

I often make the call that banks should replace their core systems if they’re old and legacy. I’m usually laughed at when I make this call, but think about it. How old is your bank’s core system?

In some countries, like Poland and Turkey, you’ll probably say 15 to 20 years old. That’s fine. I was recently in Turkey and the bankers there said they had been on a study tour of U.S. banks, and couldn’t understand why the Americans were all talking about the problem with their core systems. Turkish banks have no problem with their core systems. That’s why banks in Turkey, Poland, China, Africa and other emerging, developing and maturing markets are winning all the innovation awards. Their core systems were born after the Internet was invented.

Then we come to the U.S. and Europe.

Large banks with big brands and globalised names have an amazingly complex back office infrastructure, with core systems created through decades of mergers and acquisitions. Many have multiple systems operating for the same processes. For example, one big bank that shall remain nameless has more than 60 payment processing engines and 35 mortgage systems.

That’s a disaster.

First, it demonstrates massive fragmentation and hence security risks, as cyberattackers can find holes in the bank’s defenses. Second, it demonstrates massive inconsistency, so any chance of providing a single customer view is impossible.

But the real point is that I now champion the idea that everything in financial exchange has to be real-time and near free. That's how it is in all other Internet-based markets. In music, travel, taxis, entertainment and more we live in an on-demand market. Thanks to the 3D and 4D printing, we will soon have on-demand everything, with no supply chains hampering our ability to deliver goods now. Everything is now.

I want it now and I want it all real-time and near-free.

And there’s the rub. The banking systems were built 50 years ago, using the then latest technologies, to operate in days with high costs. Swift, Visa, MasterCard and others were leaders of their day, providing overnight batch update systems that allowed counterparties to trust each other in cross-border engagements. Today, that line sounds like such a load of tosh. I don’t need infrastructures that endorse trust to create trust. I just need real-time and near-free clearing and settlement.

That’s the point.

What is batch? What is overnight? What is a counterparty? I don’t care. I just want to deal online now, in real time, and know that there is an immediate exchange of payments and funds that is trusted and near-free. The Internet gives me that, but banks can’t, because their systems were implemented in 1965 and are long out of date.

Most banks in the developed economies of Europe and America began digitalizing in 1965 and have found that dealing with those old mainframe systems is now their biggest liability. As Patrick Jenkins writes in the Financial Times:

In the boom times of the late 1990s and early 2000s, many banks failed to invest in technology, preferring simply to ride the wave and maximise profits. During the financial crisis, meanwhile, there was no time to think of tech investment. And in the post-crisis period there has been no money. The net result across much of the world is a banking system that is creaking at the joints.


And why did we underinvest in keeping up? Because we’re pennywise and that thriftiness has now come back to hit hard as replacing core systems is risky and could be very costly. Just look at the number of banks that have tried and failed or had issues with such systems migration. But it’s not impossible and tools like the cloud allow a bank to migrate the data first, independent of the processors, hence removing a significant amount of the risk.

The question of cost is a different one. What is the cost of not migrating? A business that dies as it’s unable to adapt and deal with the real-time digital age.

The real issue here, to put in perspective, is that too many banks are running systems that were implemented before 1984, the year Mark Zuckerberg was born. They’re going to be dead meat in the age of real-time connectivity in the near-free world of today’s mobile Internet. Just to reiterate: your systems were built before the most successful Internetpreneur (my new word for a creator of unicorns) was born.

If you have a system implemented before 1984 you really are dead meat. C’mon guys. It’s like saying you have systems that were built before the American Civil War. It’s so past, it’s gone. Get with it, guys and renovate, replace, reform and rebuild. If you don’t, then you’re going to be a lovable Dodo. What does that mean? Dead but not forgotten.

Chris Skinner is an author, expert and speaker on banking, finance and fintech.

For reprint and licensing requests for this article, click here.