Twilight Of The Mainframes

It's just the news Wall Street didn't want to hear. But on Aug. 5, there it was: Nicholas M. Donofrio, head of IBM's $14 billion mainframe-processor business, fleshed out earlier warnings by company officials: Revenues for the Enterprise Systems Div. will essentially run flat through 1994--if not longer.

The implication: IBM's mainframe systems business, based on a design unveiled in 1964 and still the source of about 60% of IBM's total profits, has peaked at last. IBM bulls had counted on one more year of good growth from the latest mainframe, the ES/9000 processor, which Big Blue began shipping last year. But after Smith Barney, Harris Upham & Co. changed its IBM rating from "buy" to "hold" and Donofrio broke the bad news later that morning, IBM stock dropped 2.8%, to 91 5/8 (chart).

Donofrio didn't say so, but matters may not improve even after 1994. Large customers are using fewer mainframes in fewer data centers. Sales of IBM's mainframe disk drives have stalled as customers wait to evaluate a cheaper model they soon expect from IBM. Mounting competition from large-systems rivals Amdahl Corp. and Hitachi Data Systems Corp. means that price discounting that has reached 50% on some deals will only grow deeper. And such competitors as American Telephone & Telegraph Co.'s NCR unit are preparing radically new mainframes that will do much more work for less money than current IBM models.

Still, Donofrio told analysts he's confident of reaching his goal of a 24% pretax profit margin by 1994, up from 16% last year. But not by rolling out technological marvels. While IBM's PC unit prepares to revamp its hardware and organizational chart, he'll cut sales and administrative payrolls and throttle back on research and development spending. That sounds like a strategy for getting the most out of a mature business. IBM mainframes, it appears, have officially matured.

Before it's here, it's on the Bloomberg Terminal.