re: Banks Must Ditch Legacy IT
I've heard similar arguments, for similar reasons, in the engineering computing realm - people feeling the need to try to replace highly technical applications written in Fortran with the language/flavor of the week. Why doesn't this approach work?
1. The folks that created the original applications are probably no longer with the organization - How many banks still have people on staff from 40 years ago, in the same technical positions that they had at that time? VERY few.
2. How well documented is the original application? If you have to go back and re-engineer an application because of a lack of properly documented code, the sheer manpower and man hour requirement to make that happen is going to turn off a lot of execs with the mindset of "If it's not broke, don't fix it." - but then these are the same folks who might start the project with the organization and not be around by the time the delivery is planned 3-4 years down the road.
3. ROI - why spend millions of dollars overhauling an existing system to replace it with something that does the exact same function? As long as existing hardware systems are supported by the manufacturer or a trusted third-party, there's no impetus to make a move from that standpoint and it creates a cottage industry (if one can call it that) that supports nothing but banking organizations.
Ultimately, it comes down to how much money is going to be made/saved by making the decision to move to a new system - and given the fuzzy math behind a lot of those calculations, once you start getting into the 7 figure range and higher, you'd better have some serious backup for those calculations.
That said, my mantra when I was in the consulting game was that upgrades are inevitable - doing them on your own time is much more cost effective than being forced into them by a system failure. Once a major bank starts having major technical failures, I think you'll see the temperature start to rise for other banking organizations to get going on their own modernization efforts.