In computer years, Cobol must be pushing at least 300 and should be taking it easy and enjoying its place in the computing hall of fame. Yet, while it certainly is not able to do many of the things asked of modern-day IT systems, it’s actually still going strong, alongside its old-timer compadre the mainframe.
It’s an anomaly that the finance industry, so often at the pointiest tip of the sharp end of technology, has such ancient technology at its heart. But behind the carapace of the bright, shiny real-time trading and analysis, mobile and electronic payments, lie some of the oldest code - and programmers - in business. The internet and mobile technology may be driving finance forward, but its foundations are mainframes running Cobol.
Cobol and mainframes have survived numerous rounds of the ‘rip out and replace’ school of dealing with legacy systems. The secret of its survival is simply that although old, the combination of Cobol and mainframes is very good at high-volume transaction processing. There’s also the more pragmatic reason that it can be incredibly risky and costly to replace.
But there’s a real problem. Cobol is over 50 years old and the people who set up the systems are retiring. When they pick up their carriage clocks, who is going to have the deep knowledge of the business logic deep within the bank programs? The problem will only get keener over the next five to ten years.
A 2012 Computerworld survey found that 46% of the 357 IT professionals it surveyed were already seeing a shortage of Cobol programmers, yet more than half (53%) were still developing new applications in Cobol. So it would appear that there is still new life in Cobol in parts of the banking industry.
Cobol is ‘un cool’
On paper, the solution looks easy: if there’s a shortage, encourage new people to learn the skills. But Cobol has a major image problem. A 2013 survey by Cobol tool provider Micro Focus, found that most students considered Cobol as decidedly ‘un-cool’ or even dead and wanted to study Java or C++ instead. So, while 58% of academic leaders thought Cobol belonged on the curriculum, only 27 of universities had it on the curriculum. Only 5% of college-level programmes produced 30 or more students classed as Cobol developers.
Although some of the students thought that Cobol was dead or wouldn’t recognise the code, Micro Focus is keen to point out that 85% of all daily business transactions are powered by Cobol and there are $2 trillion worth of Cobol-based mainframe applications in use today.
You can hardly blame students for wanting to work with newer technology: the kind that will get them a job at Google or the ‘next big thing’, rather than what is perceived as yesterday’s technology. So what are banks to do?
Well, better to tackle the problem sooner rather than later. If they do decide to migrate off the mainframe, it’s going to be a lot easier, cheaper and quicker to do whilst the experienced legacy programmers are there to help with the process. If they decide to keep going with their mainframes, then they need to make sure the baton of knowledge is passed on to the next generation.
Recruiting or retraining staff to fill the gap is not going to easy. Cobol skills are not a the sexy end of the business and in some cases a mainframer with 20 years experience will be paid the same as a Web programmer with five years experience. There needs to be incentives for younger people to specialise in Cobol and for them to see that they have a long-term future in that technology, if they do choose to specialise. Alongside the younger blood, there’s also no reason why the older generation should retire at 65. People are working longer, so make it worthwhile for people to stay. Who knows, perhaps SAP skills will be in the same situation in 25 years time!
As ever, if you offer people the right pay and the right benefits and the right opportunities to create a worthwhile and challenging career, then they will come. But there is no doubt - this is a tricky sell.