On the surface, it is making the right choices. It's decreasing or divesting investments with low returns on capital (hardware) and increasing investments in growth areas.
Hardware is easy to understand: its business is being pressured by moves to cloud computing and by pressure from rivals like Cisco (CSCO). So it makes sense to lighten up. It announced it's selling its global semiconductor OEM business to Global Foundries.
But it's also tough in their two other areas of expertise: Services and Software.
A few years ago, Services was the big growth area. But now customers can outsource parts of the services business to much cheaper competitors. Big problem.
Software has a slightly different problem. Much of the software business is very mature, like database and systems management.
There are parts that are growing: business intelligence software, for example, the most visible example of which is Watson. It leverages IBM's core skills in analytics and big data.
But there's not a lot of fast revenue growth there. See the problem? No growth. Outmaneuvered by smaller, more nimble competitors. Too big to grow.
What's depressing for holders is the stock is cheap, trading at roughly 10 times forward earnings. It has a respectable dividend yield of 2.7 percent.
But investors don't seem to care. They see a company in decline. Thanks to impairments and restructuring charges, operating income is dropping this year to roughly $16 billion from $20 billion in 2013 (ouch!)
The quarterly dividend ($1.10) appears safe, at least for now. IBM's payout ratio is in the mid-20s, not very high, so it seems unlikely it would take such an aggressive move now.
One thing's for sure: the stock is SO oversold that, just on technicals, it should start attracting buying interest. But, after examining the fundamentals, a lot of traders would argue that IBM is the classic example of why no one should buy a stock just because it is oversold.