In the 60's, the list of companies in the Fortune 500 was so stable that it took 20 years for a third of them to change. Now it takes only four years (source: The Economist). There is a lot of pressure for businesses are to remain competitive.
Will COBOL survive such business needs? ...
Frequently I see discussion threads on the internet asking “is Cobol dead”. A COBOL enthusiast usually gives a definitive “NO” – and goes on to observe such false accusations have been going on for ages, and yet, there is still more COBOL code running enterprise systems than any other language. Another proof is the longevity of companies such as MicroFocus … they are still in business, still making loads of money from selling COBOL compilers and related products.
But do you think there will be any COBOL on starship Enterprise? I think if there were, Mr Data will probably quickly rewrite it, and probably do a good job too (a quick plug for an Intelligent Translation similar to SoftwareMining )
So, you have to wonder whether the big organizations with big legacy systems would adapt in time, or would their legacy system cause them to sink.
There is off-course many who have started modernization (otherwise Softwaremining wouldn’t be in business), but then again, there are many more who jave opting for re-hosting solution.
My personal view is that COBOL won't see a gradual decline … but a sudden death over two or three years. No-one would want the maintenance issues of the "last" legacy systems, and so everyone will rush to get out.
Just before that time, it will peak again with retired developers come out of retirement (again), and others auctioning their services to highest bidder.
Dead or not, there will be plenty of excitement :)