I was... "fascinated" by this quote: "That is good news for start-ups such as Metro Bank, which wants to open more than 200 branches in Britain by 2020. With such grand designs, isn’t Vernon Hill, one of the bank’s founders, tempted to build an IT system of his own? “I hate programmers,” replies this dyed-in-the-wool entrepreneur. “They only cause trouble.”"

I am not surprised.

The relation between IT departments and the rest of the company are often weird, it even more complicated when the IT department has two jobs rolled into one : administrating the system and developing them. And they work by ticket for all these activities.

They have been told to run IT like a business on its own so everything is done by contract. But IT never can make a profit, and reinvest in cost saving or modernization measures. Purchasing off the shelf software eases and clarifies a lot of all that. Just having you IT services provided by an external company often does not.

It's a good thing that bank doesn't use any software.

You have to go with the lesser evil.

Is he wrong? /programmer

META: Why did the title change? I cannot edit the title of my own submission anymore... Was it merged with a duplicate submission? If so, then there's also http://news.ycombinator.com/item?id=1541336 (which was created just a few minutes after my submission)

I believe sometimes mods change titles to be more accurate or neutral. I've had it happen to me before, too.

Having worked in operations for two major banks, I can confirm the article. We used software in DOS emulation and old fashioned command prompts to process trades, which were essentially done by hand. This was during the wave of consolidation leading up to the financial crash. Why improve when you can buy the competition?

It's not just big banks, small community banks (and credit unions) are at the mercy of convoluted system vendors as well.

A system I know of includes many barely-integrated parts, most of which are from different vendors: - A core processing system

- An online banking system

- An add-on to the online banking system to provide "enhanced" logon security to comply with an FFIEC mandate

- An add-on to online banking that provides online statements

- An email-alerting system (think "your account is overdrawn")

- A loan origination system

- A loan servicing system

- A CRM system

- A document archiving system

- A mobile banking "solution"

The WTFs you find in these systems will boggle the mind, and seamless integration is maddening.

Here in Germany, I haven't found a bank that has any email functionality or a post-2001 web interface.

Maybe the reason why it takes 3 days to transfer funds to another account really is incompetence rather than greed.

If you want to gain some domain knowledge in this area, especially for smaller community banks, be sure and check out the excellent 5th comment.

One of my previous jobs involved applications used by some banks, and I got to see how archaic some of the hardware still in use was. In one bank - to remain anonymous but is a well known name - they were still using a 25 year old computer system, which was older than I was at the time.

I know of one dearly-departed bank that used a mainframe system so ancient that there was only one company left that supported it. Since the early 90's, that company had had just one customer.

What is wrong with mainframes?

The problems banks solve are the reason mainframes exist. Billions of data points that must be processed 100% correct 100% percent of the time. It seems wrong to demonize mainframes as old and outmoded.

It's not the idea of a mainframe in and of itself. It's that the longer you have such a system in place, the more difficult it becomes to update it.

The State of California still runs all its payroll processing via an ancient mainframe running COBOL. Making changes is difficult, fewer and fewer programmers are available to perform the work (and so it costs more) and the problem just keeps getting worse. Apparently they have been trying to replace it since 2004, but the legislature doesn't want to provide the $100 million that is supposedly required.

When it takes 2 banks of mine about a week to decrement a number in one bank's database and increment a number by the same amount in the other bank's database, there's something horribly wrong somewhere.

Banks fear the complexity of IT. Now they are pushed to get into the challenge of offering new services that require a complex setting, so they smell problems in the near future.

I agree with this article, but I admit that the Little Britain reference in the title is what made it for me.

http://en.wikipedia.org/wiki/Carol_Beer