CEO Views: A demand for COBOL expertise underlines the fragility of critical infrastructure



By Paul Taylor

April, 2020

“Literally, we have systems that are 40 years-plus old, and there’ll be lots of postmortems. And one of them on our list will be, ‘How did we get [to a point] where we literally needed COBOL programmers?’” - Governor Phil Murphy

I wasn’t surprised in the slightest when New Jersey governor Phil Murphy recently put a call out for volunteers who can code in COBOL, a programming language developed in 1959. The technical debt in financial services infrastructure has been mounting for decades – and is now rearing its ugly head.

A technological time bomb

A surge in US unemployment claims in early April has tested the technical limits of New Jersey State’s mainframe hardware and software. Systems designed well before Neil Armstrong set foot on the moon. To be clear, these are the systems responsible for running 43% of the world’s banking systems, and almost 100% of America’s ATM swipes. I shouldn’t need to spell this out, but the global banking industry is sitting on a technological time bomb.

It’s equally frightening to see organisations forming to try and keep COBOL alive. Retired developers have created technical consultancies, advocating to keep critical systems running on this antiquated language.

To them, I say this: you are keeping critical banking and public sector infrastructure on life support. Customers are suffering, they are not getting access to the funds and services they need. There is always the option to move forward with scalable, modern and resilient technology.

Proponents of legacy systems make a grave mistake in thinking that if a system isn’t broken, there is no need to fix it. In fact, the routine delivery of upgrades in the absence of problems builds resilience into infrastructure. If we don’t continuously upgrade systems, they will become irreparable in due course. Cue New Jersey.

The migration fallacy

These advocates maintain the argument that there’s simply no option to migrate over to modern, cloud native systems. They argue that legacy code is inexorably linked to critical systems, and a migration effort would cause untold damage. This couldn’t be further from the truth.

Forward-thinking, long-term organisations, are already making the move. Our work with Standard Chartered, Atom, Lloyds Bank and SEB demonstrates that cloud native infrastructure in finance works, and it works at scale.

Make no mistake – migration is a significant task for any large-scale IT transformation but it is not an insurmountable one. A well-planned and well-structured migration programme – the type that we employ at Thought Machine – greatly minimises the risk and possibility of any errors.

Cloud native: the engine of global innovation

Cloud native computing is the engine of enterprise IT and consumer technology. The internet, and all of its auxiliary services, would be virtually mute without the power of cloud native computing. This includes the countless education, health and logistics systems which function perfectly on the cloud, today, without a single line of legacy code.

Google is synonymous with consumer internet services. Amazon’s infrastructure is the backbone of the internet. Spotify, Airbnb and Netflix have dominated their respective fields by harnessing the power of cloud native technology. These are the systems and platforms which will prevail in the face of unprecedented shocks.

There will be no call to arms for retired engineers when a business on the cloud faces a surge in demand. Cloud native systems, as always, will dynamically reconfigure, elastically adjusting to the load.

If engineers are called for, they will be well-skilled in cloud native thinking. They will be the types of engineers who, first and foremost, deliver high availability, tolerance, elasticity and resilience. They will deliver innovation and features to improve customer experience - not fight fires. It’s time to migrate, the cost of not doing so is too great.