COBOL (COmmon Business-Oriented Language) was invented in the late 1950s as an easy-to-use computer language for business. It was adopted as a standard for its mainframes by IBM in 1962, as the internet tells me, making it the de facto industry standard back then.
It would be tempting to say that no one uses this horse-and-buggy-whip technology any more, but that wouldn’t be right. The COBOL-mainframe combo continues to be the mainstay for banks, insurance companies and the government.
I’ve never been interested enough in financials to find out for sure. It sems to me, though, that the primary reason is cost. Presumably a company would build a more modern system, run the two in parallel for a while and then shut the mainframes down.
For big companies, this would be a huge out-of-pocket expense, probably making a big one-time hole in the income statement. Worse, from a pragmatic office-political point of view, who knows how top management bonuses would be affected and what the board of directors might think about a long-festering issue no one had told them about. Maybe it gets them thinking of looking for a new CEO.
In the case of government, what I’ve read (and believe) is that it takes ten years for any computer project to go from proposal to implementation.
Why take the risk? Why not cross fingers hoping do disaster happens on your watch and leave the problem for the next guy.
To be clear, I’m not saying this is what one should do. It’s what I think has been done.
An aside:
–in 1990 I was hired by a large financial company to turn around a global fund that had a disastrous performance record. When I arrived, I asked the computer people to create a performance attribution report to my specifications–something I’d had at prior jobs and that could be knocked out on a PC in a week at most. What I really wanted was for the program to be hooked up to a data feed. The programmers said the task would take a year to complete …and cost, in today’s money, $200,000. Welcome to COBOL and/or bureaucracy inc.
Why is this important today?
–there aren’t that many COBOL programmers around any more. People who know the language best from years of working with it are now in their seventies or eighties. So COBOL systems are increasingly hard to get repaired. And…
–DOGE is firing COBAL programmers working for the IRS, and presumably those working for other agencies as well. Those who choose to keep working will likely find that they can make much more money working in the private sector for financial firms. So who’s going to maintain the government’s computers?
Arguably, this phenomenon has no direct stock market relevance. On the other hand, it’s also conceivable that the lack of computer processing power that might result from inability to repair clunky old machines will slow the flow of funds into Washington’s coffers. Another reason to be concerned about higher rates/lower dollar.