Reach Out and Grab Someone

By THOMAS MCCARROLL NEW YORK

When American Telephone & Telegraph entered the computer business six years ago, big things were expected to happen. After all, the company had invented the transistor, the basic building block of modern computers, and it had built the nation's telephone system, which is essentially one vast computer network. Industry analysts predicted that AT&T one day would even challenge IBM for market supremacy. The government, which had barred Ma Bell from the business until the phone monopoly was broken in 1984, fretted that it might be opening the way for the giant (1989 revenues: $36.11 billion) to dominate the computer industry. But instead of conquering the market, AT&T has suffered one humiliating defeat after another, racking up huge losses in the process.

Though many expected the company to throw in the towel, AT&T stunned Wall Street last week by proposing the biggest acquisition in the brief history of the computer industry, offering to buy Dayton-based computer maker NCR for $6 billion. When NCR rejected the initial, friendly offer, its suitor shocked the business world once more by launching a hostile takeover attempt. In a face- to-face showdown with AT&T's board members in New York City, NCR management vowed to resist. But industry analysts generally believe that the big telecommunications firm will ultimately prevail, strengthening NCR in the long run. Says Maria Lewis of Shearson Lehman Bros.: "AT&T would make an ideal corporate parent." Still, the deal carries enormous financial risks for AT&T, which had tried to build a computer business from scratch, avoiding large acquisitions. "AT&T will triumph, but it may be a Pyrrhic victory," says John McCarthy, an analyst at Forrester Research. "It's buying itself a land mine of problems."

At least on paper, the AT&T-NCR combination looks like a good match. NCR, known for its electronic cash registers and automated-teller machines, is a leading maker of midsize and desktop computers. With revenues of $5.96 billion last year, it is the fifth largest U.S. computer manufacturer (after IBM, Digital Equipment, Unisys and Hewlett-Packard). What excites AT&T, however, is not NCR's market share but the potential for linking its own long-distance telephone system to NCR's worldwide network of cash registers and ATMs. That would give AT&T significant entree into the exploding business of processing transactions for retailing and financial-services firms. And it fits into AT& T's dream for the 21st century: to wire every household for computerized shopping services.

NCR also suits AT&T's long-term computer strategy. The two companies' machines are largely compatible, using the same operating software, called Unix, invented by AT&T in 1969. As a result, they would be able to integrate their product lines rather than face the dilemma of having to eliminate a system. But more important for AT&T, the addition of NCR would enhance the company's position in its ongoing battle with IBM to establish Unix as the industry standard. Both companies want to replace the technically outdated standard known as DOS. IBM's entry, called OS/2, appears to be the stronger contender. While Unix has been gaining market share, AT&T lacks the credible , machines to exploit the system's rising popularity. Instead, other computer makers using Unix, including Sun Microsystems and Digital Equipment, have cashed in.

AT&T's abysmal showing in computers so far is somewhat baffling. Its scientists at Bell Laboratories have been on the leading edge of computing, playing a key role in developing such technology as the microprocessor. But the company has failed to convert high science into financial success. Its first commercial computers, a series of midsize machines called 3Bs, flopped largely because, at up to $100,000, they were overpriced. The company later formed joint ventures with Convergent Technologies and Italy's Olivetti to make personal computers under the AT&T brand. It also formed a partnership with Sun and made a number of minor acquisitions, including Paradyne, a modem maker, and Istel, which customizes computer systems. But AT&T never managed to capture more than 3% of the market, and losses have mounted to as much as $3 billion. Rumors began to swirl about a possible merger as a quick fix. Among the known targets: EDS, Sperry, Digital Equipment, Wang and Data General.

AT&T had approached NCR in 1988, but the response was the same as today's: no, thanks. NCR only recently revamped its product line, shifting from computers using its own software system to machines that run Unix and DOS. "We didn't want AT&T's computer mess dumped on us then, and we don't want it now," says Charles Exley, NCR's chief executive. In discussions last week with AT&T's chief executive, Robert Allen, Exley warned of the history of failed computer marriages, such as Sperry and Burroughs or IBM and Rolm: "The industry graveyard is littered with mergers that have been outright calamities, and there is no reason to believe this one will be any different."

Allen's task is to convince the NCR chief that this acquisition would not become another tale from the crypt. Although Exley has threatened to resign if "AT&T succeeds in shooting its way into NCR," analysts think he can be persuaded by a higher price. NCR has indicated a willingness to submit, but at $125 a share (for a $8.5 billion total) rather than the $90 offered by AT&T. Wall Street observers think the two sides will settle at around $105 a share. For its part, AT&T refuses to back down. In a letter to Exley, Allen said, "We remain dedicated to the completion of this transaction." But even if AT&T does bag NCR, its problems could be just beginning.

CHART: NOT AVAILABLE

CREDIT: TIME Chart by Steve Hart

CAPTION: PLUGGING IN TO COMPUTERS