IBM has re-invented itself more times than you remember. They did it when mini computers came along, when the PC hit (an own goal if there ever was one), they did it when the Internet came of age (and when everybody else was still playing catch-up or even ignoring it) and I'm pretty sure they'll be able to survive the transition to the next phase of things (the cloud or whatever is fashionable). IBM is about as stable as they come, they do IT in whatever form it comes. When it was big iron they rolled their own hardware and software to go with it, now it is Linux, so they'll do Linux (on mainframes if you want it), if that changes, they'll change again.
I'd write off a lot of companies long before I'd write off IBM and while they're laying off in one set of divisions they're hiring in others.
So they're not 'magically' going to reverse course, they are gradually going to change course, like they've always done (and like every other supertanker does, move too fast and you'll break things for real, this is not a start-up).
Each and every one of those changes was heralded as 'the end of IBM'. They definitely messed up during the OS wars, I'll give you that but for the most part they rode the waves better than just about any other tech company that has been around this long. Rumors of IBMs imminent demise are most likely vastly exaggerated and any prediction of lay-offs should be henceforth accompanied by some evidence or I'll simply not buy them any more, especially when they entail 100K+ lots of employees.
How are Sperry, Control Data, Data General, Burroughs, SGI, Cray, DEC and a whole slew of others doing these days?