THE STRUGGLE OVER SPERRY'S FUTURE Once IBM's biggest competitor, Sperry Corp. has sunk to No. 7 in the U.S. computer business. As earnings hit the skids, the * company confronted an identity crisis: Should it keep making computers? When three top executives fought over the question, one lost.
(FORTUNE Magazine) – WHOEVER RENTS Sperry Corp.'s midtown Manhattan headquarters when the company moves out next year will inherit a battlefield. Sperry spent three years there at war with itself, as dissenting executives fought Chairman Gerald Probst and the managers of the computer business. At stake: what kind of company Sperry should be and whether it would be worth more broken up and sold. This fall the dissenters lost. Vincent McLean, Sperry's chief financial officer, resigned; a few other senior officers also left. The spoils quickly went to the victors. Sperry announced it will move its headquarters to the Philadelphia suburb of Blue Bell, the computer unit's home, and named computer boss Joseph Kroger its new president. While the rebels have been banished, the questions they raised still haunt the company. Sperry has a proud yesterday and a lackluster today. It built the world's first commercial computers under its former brand name, Univac. Sperry still makes mainframes, those multimillion-dollar machines that do everything from keeping track of airline reservations to managing government bureaucracies. It also has a steadily growing defense and aerospace operation that is a leading supplier of navigation and weapons-guidance systems. However, the company that was for years IBM's biggest competitor has fallen to a distant seventh among U.S. computer makers. Ranked No. 69 on the FORTUNE 500, Sperry had $5.8 billion in revenues over the past four quarters. But its return on shareholders' equity, an anemic 9%, is well below the FORTUNE 500 median of 13.6%, and its stock is trading sluggishly below book value. Sperry's biggest problem is its biggest business. Computers currently generate about half its revenue and two-thirds of its profits. But the technological struggle to keep up with competitors is consuming a huge share of Sperry's assets and R&D money. IBM and leading Japanese computer makers have been improving the dollar-for-dollar performance of their mainframes at a breakneck 20% annual rate. Sperry plans to spend over $2 billion on computer R&D in the next five years -- 75% of the company's total R&D, but a puny 10% of what IBM is likely to spend. So outgunned is Sperry that Ulric Weil, a security analyst at the Morgan Stanley investment banking firm, thinks that, without a lucky breakthrough, the company is unlikely ever to become an above- average performer. Sperry shows signs that it may be changing. It is quietly negotiating with Hitachi for help in building its next generation of mainframes. The deal could be Sperry's first step toward abandoning the risky business of computer manufacturing and becoming a potentially more profitable software and services company -- and a sales agent for Japan. For the most part, however, Sperry is groping for direction and throwing out confused signals about where it intends to go. It downplays the Hitachi deal and insists it will keep building mainframes. Yet it is also shifting emphasis from manufacturing to electronic systems integration -- piecing together, mostly from components made by others, complicated office and weapons systems. Inspired partly by AT&T's entry into the computer business and IBM's acquisition of Rolm, Sperry would like an alliance with a telecommunications company. Such a move would broaden the product line and expand Sperry's markets. Meanwhile, the company is prowling for acquisitions; it has $430 million in cash from a debt offering and the recent sale of New Holland, a farm-equipment manufacturer (hay balers, combines) that Sperry had been trying to unload for years. FORTUNE has reconstructed the debate over Sperry's future through interviews with company executives, former employees, and Wall Street sources; with one exception, the eight outside directors declined or ignored requests to be interviewed. The central figure in the executive-suite struggle is Chairman Jerry Probst, 62, a crusty executive who made his career in computers and is steadfastly devoted to that business. When Probst became chairman in 1982, Sperry was a financial shambles. It held hundreds of millions of dollars of long-term leases on mainframes that had been underwritten with short-term debt. For years the practice of borrowing short and lending long to finance customers had boosted earnings. But Sperry owed a dollar of debt for every dollar of shareholders' equity. When short-term rates soared, profits plummeted from an all-time peak of nearly $8 per share in 1980 to less than $3 two years later. PROBST LACKED the financial background to nurse Sperry back to health, so he hired an expert: Vincent McLean, 54, chief financial officer and strategic planner at NL Industries, a New York oil services company about half Sperry's size. McLean came to Sperry in the same dual role and was named the only * inside director besides Probst. McLean, who will take early retirement next March, is smart and buttoned-down, a Midwesterner who has an MBA from the University of Michigan. But while he was at ease among the fast-dealing investment bankers of Wall Street, he lacked the cunning of a corporate politician -- a skill that, as it turned out, was essential for survival at Sperry. Soon after his arrival he described himself as a gadfly; suspicious insiders dismissed him, recalls a Sperry alumnus, as ''a good tightwad.'' By all accounts, McLean did an effective and at times innovative job of cleaning up Sperry's balance sheet. With a series of shrewd domestic and Eurodollar debt offerings, for example, he extricated Sperry from its precarious short-term borrowings. But McLean's role as corporate strategist led to conflict with Probst. Corporate planning was a novelty at Sperry, which had been organized as a confederation of semiautonomous businesses. In the beginning, Probst was all for planning, which was done by two powerful executive committees, one for Sperry's computer and defense operations, the other for the machinery business. Probst headed both. McLean didn't take long to identify the problems of the mainframe business. He challenged the computer group's forecasts and questioned the wisdom of keeping more than 70% of Sperry's assets in computers. His doubts about the computer business crystallized in 1984, in the form of a financial restructuring plan that Sperry commissioned from investment bankers at Salomon Brothers. At Sperry it was known as Project Trinity, after the three businesses Sperry was then in. McLean's private name for it was Project Skirt, based on a salty wisecrack Sperry executives won't divulge. Skirt went beyond the routine liquidation analyses that investment bankers do for clients whose stock is depressed and who worry about corporate raiders. PaineWebber had already done that sort of study for Sperry. Known as Diamond, it was available for Sperry's board to examine in the event of a raid. Skirt, by contrast, was hardly discussed in the boardroom. Management had ordered it to assess how Sperry might reward shareholders by breaking up the company. The plan reviewed several scenarios. The one McLean favored would have dumped New Holland and then split the computer and defense businesses apart by selling or spinning one off. The computer group consumed tremendous amounts of money thrown off by the defense business -- in some years hundreds of millions of dollars. In fact, computers were known around Sperry as ''the ultimate poison pill.'' The phrase describes a scheme concocted to drive off raiders by making a hostile takeover financially ruinous. In Sperry's case, the threat of ruin came from the mainframe business. The defense and aerospace group, on the other hand, had $2 billion in revenues last year and had landed major contracts to supply navigation systems for the Navy's Trident missile submarines and weapons systems for Canadian combat ships. It usually earns 20% a year on assets. Judging from the value of independent defense electronics companies, McLean estimated the unit's worth at $2.5 billion -- a sum close to the total market capitalization of Sperry today. Eventually Project Skirt became synonymous with the idea of splitting apart the computer and defense groups. It wasn't merely the pet of McLean and his planners; at least three other corporate officers gave it tacit support. But Skirt didn't stand a chance with Probst. Although he had commissioned the study, he dismissed the results as a financial contrivance. Probst expressed no interest in Skirt even in January 1985, after General Motors, a prospective suitor few chief executives would ignore, asked about buying Sperry's defense business. GM, in pursuit of Hughes Aircraft at the time, was exploring other ways to get into defense. When McLean suggested that Probst might want to meet with GM executives, Probst wasn't interested. ''Why,'' Probst asks, ''would we sell a business that's in a market so many people want to get into?'' Paradoxically, as McLean was succeeding in cleaning up the balance sheet, he was losing influence with Probst. Probst distrusts investment bankers and thinks financial people ought to know their place in a company. ''This is not a business school exercise,'' he says. ''These brilliant MBAs ought to roll up their sleeves and get to work in the factory.'' The clashes between McLean and Kroger over the future of the computer business annoyed Probst, who insiders say shuns confrontation. In the fall of 1984 he replaced the two planning committees with an office of the chairman, promoting Kroger to McLean's level and giving him a seat on the board. Kroger's promotion meant serious trouble for McLean. Kroger's motto for the computer business is ''survive and move ahead,'' and he and Probst saw eye to eye. Kroger, 51, began as a computer salesman at Sperry and worked his way up. Known for decisiveness, he has the strong jaw and commanding presence of a central-casting chief executive. He also has a reputation as a macho, vindictive trench fighter who demands unquestioning loyalty from subordinates. In recent years he has shaken up the computer unit, driving out more than a dozen vice presidents and breaking the hold Sperry's engineering department had on product planning. Looking back, he says simply, ''Every time I got promoted I eliminated a level of management.'' Those who know Kroger speculate that he may mellow now that he has almost reached the top. After his promotion Kroger started dealing directly with Wall Street, bypassing McLean. With Probst's blessing, the computer group staged its own presentation for security analysts last December. The idea was to explain Sperry's plan for making the computer business, in the words of First Boston analyst Steven Milunovich, ''a prince and not the toad that it has been.'' ALTHOUGH IT failed to change many opinions, the presentation was remarkable. To show that Sperry was serious about staying in mainframes, the computer group unveiled, in unprecedented detail, product plans for the next five years. Kroger and two lieutenants pragmatically reviewed the computer group's faults -- its wimpy profits, eroding market share, and weakening grip on major customers. They declared two long-term financial goals: raising annual revenue growth from near zero to more than 15%, and hiking return on net assets from 9% to 17%. Ambitious as these goals seem, they may not be high enough to satisfy shareholders. Sperry's internal calculations suggest that the computer unit's return on assets would have to exceed 20% for the stock to trade sustainably at book value. Under the plan, growth is to come mainly from software, services, and small computers that will be sold to Sperry's established customers. Sperry makes no small machines of its own, but markets IBM PC clones, built by Mitsubishi, under the Sperry label. It also resells small computers made by NCR, Computer Consoles Inc., and other companies that all share AT&T's Unix system as their basic operating software. To strengthen its links with the customers who are supposed to buy such gear, Sperry believes it must keep building mainframes. Moreover, it thinks that it gains significant marketing advantages from the mainframe business. Sperry's big computers use programs that won't work on other makes, so that once customers buy they become a captive market for Sperry's lucrative software and service sales. Most customers will pay premiums even for machines that are late, rather than abandon their software investments. Kroger disclosed to the analysts that, beginning next year, Sperry would start selling a new generation of smaller, less expensive mainframes based on sophisticated customized chips. He is counting on the machines, with code names like Liberty, Saturn, and Swift, to keep computer revenues strong. Project Skirt died soon after Kroger's presentation. Last January Probst deleted the plan from the agenda of Sperry's annual management review. But Skirt's defeat did not prevent merger and divestiture rumors from swirling, and Sperry's financial staff made it clear to the company's investment bankers that they thought Sperry would be worth more broken apart. Another source of rumors was Sperry's search for a telecommunications partner. One advocate of that idea was said to be Sperry director Willard Boothby Jr. A managing director of PaineWebber, Boothby declined to be interviewed by FORTUNE. But in 1984, according to a Sperry source, he inquired on Sperry's behalf whether AT&T Chairman Charles Brown might be interested in an alliance. Brown said no, but word of the overture leaked out. When ITT Chairman Rand Araskog expressed interest in Sperry last January, Probst, McLean, and Kroger temporarily put aside their differences and scrambled to strike a deal. The talks, spearheaded by McLean and Kroger, did not last long. Araskog withdrew, partly because of pressure from large shareholders who wanted to see ITT make divestitures, not mergers. Asked how the merged company would have been structured, Probst growls, ''The market never let us get that far.'' Probst blamed the collapse of the talks on Wall Street leaks; he told McLean not to talk to Sperry's investment bankers without first asking permission. When Burroughs Corp., a $4.9-billion-a-year rival in the mainframe business, showed up as another suitor, Probst handled the negotiation his own way, paying little heed to his financial advisers. Burroughs Chairman W. Michael Blumenthal made his first proposal to Probst at his New York apartment in April. Though Probst worried that Sperry's customers might view a merger as weakening Sperry's commitment to supply them with mainframes, he did not reject the idea. He met with Blumenthal and other Burroughs officials several times. At least once he brought along Kroger who, to one Burroughs executive, seemed opposed to the idea because it might threaten his job. Meanwhile Probst, as this executive tells it, appeared willing to let the talks go on forever. ''Probst didn't know the rules,'' he says. ''We pointed out that merger discussions take on a life of their own, and said we should get on with it.'' In early June Blumenthal insisted in a letter to Probst that the dithering cease. Rumors of the merger were beginning to circulate and Burroughs stock dropped 9.5% as security analysts puzzled over Blumenthal's intent. On June 14, without having reached an understanding with Probst, Burroughs made a formal offer to the board to acquire Sperry by a stock purchase and gave the company three days to respond. The proposal assigned Sperry shares, then trading at $55, a value of $65, or $3.7 billion in total. (Sperry shares recently stood at $46.75.) Sperry's board requested more information, the deadline lapsed, and Blumenthal withdrew. The board was also busy behind the scenes, FORTUNE has learned. Treating the Burroughs offer as potentially hostile, it looked for white knights and deployed enough golden parachutes -- over 150, worth an estimated $50 million -- for a company many times its size. In September, Sperry struck out on a different course when Probst and Kroger commissioned PaineWebber to help negotiate the deal with Hitachi. Further isolating McLean, Probst instructed the investment bankers not to talk to him. Distressed to be barred from negotiations that affected the company's future, McLean resigned. THE HITACHI DEAL is an attempt to forestall a disaster with Sperry's largest computers. The company needs help with a successor to its current mainframes, the 1100/90 series. Those machines were introduced in late 1983, two years after IBM's 308X series, with which they compete. IBM has already begun shipping the first of its next generation of machines, the so-called Sierra (see Selling). So far, customers bound by software to Sperry have tolerated the technology lag. But if the company falls further behind, they might feel forced to break the chain and switch to IBM to keep up with competitors who use more advanced computers. To hold its position, Sperry's internal studies show, the company must substantially soup up its mainframes by 1988. But its engineers won't be ready with their next-generation computer -- code-named Mercury and discussed during Kroger's Wall Street extravaganza -- until after 1990. According to sources close to Sperry who are not involved in the negotiations, the company would like to buy circuitry for an interim machine from Hitachi. Sperry needs help because it stumbled in developing an important area of electronic technology. Crucial subsystems of modern mainframes, notably the instruction processor or logic, comprise dozens of exotic high-speed chips. For a machine to work efficiently, these must be densely packed. The high- speed circuits generate a lot of heat; without cooling, a mainframe would melt down faster than you could say Three Mile Island. So packaging, connecting, and cooling the chips has become an art. According to Sperry engineering veterans, the company blundered around 1982 when it underestimated the problem and cut the budget for packaging research. It eventually caught the mistake, but still didn't comprehend the complexity of packaging. Instead of scrambling to catch up, Sperry bet on a long shot, taking a $42-million stake in a start-up company called Trilogy Ltd. Trilogy was founded by the renowned computer architect Gene Amdahl, who was experimenting with a technology known as wafer-scale integration. His idea was to cram onto the surface of a mammoth, English-muffin-size chip the circuits that occupy dozens of chips in today's computers. Each Trilogy wafer would do so much that Sperry wouldn't need to package many together. But the wafers fizzled. Sperry took a $24-million write-down on its investment in the summer of 1984, and with astounding aplomb shrugs it off publicly as a worthy try. Privately Sperry was frantic about Trilogy's failure to come through with big chips. ''Trilogy was our Linus blanket,'' recalls an engineer. It was time to dial H-I-T-A-C-H-I. The Japanese company has chips and packaging technology it could use to design and build subsystems for Sperry's enhanced 1100/90. For that to happen, however, Sperry would have to share with Hitachi details of its computer architecture that have been kept secret for two decades. The project would require such intimate collaboration that Sperry might be forced to turn to Hitachi for help with next-generation Mercury and all its successors. Hitachi also has up-to-date disk drives -- dishwasher-sized peripheral devices for storing data -- that Sperry needs for its customers around the world. Disk drives typically account for 40% of the revenues and profits from / a mainframe installation. Sperry's efforts to manufacture the units, most recently in a joint venture with Control Data Corp., have failed. Sperry views the arrangement it is seeking with Hitachi as a stopgap and plans to build Mercury without Japanese help. Thus the deal won't reduce the assets committed to the mainframe business or the roughly $100 million that Sperry earmarks for mainframe hardware development each year. The first effect of a Hitachi deal could be to confront Sperry with a marketing dilemma. Covenants would probably keep Hitachi from invading the market with Sperry- compatible mainframes of its own. But the more Sperry serves as a sales agent for other manufacturers, the less control it has over the evolution of its products -- and the less it can assume that customers will stay loyal. Asked to reflect on the prospect of leaving the mainframe manufacturing business altogether, Kroger muses, ''You'd ruin your image. You could claim to be a systems integrator, but why should anyone buy from you thereafter?'' If it sells Hitachi's disk drives and builds mainframes with Hitachi subsystems, Sperry will learn the answer. Kroger and Probst are seeking a new image for Sperry. With McLean out of the picture, they have turned to schemes for expanding -- and fusing -- Sperry's two businesses. Probst says his main decision will be what new businesses to buy. Earlier this year, he held an informal competition between defense executives and Kroger's group to devise the best acquisition plan. But no matter how much -- or little -- Sperry acquires, big questions still hang over the company. Kroger acknowledges that it is ''probably questionable'' that the computer unit will achieve its five-year financial goals. Sperry still doesn't seem clear about what it is doing, nor does it have much time to decide before the doubts that plagued the company in New York follow it to Blue Bell. BOX: INVESTOR'S SNAPSHOT SPERRY SALES (LATEST FOUR QUARTERS) $5.8 BILLION CHANGE FROM YEAR EARLIER UP 22% NET PROFIT $263.9 MILLION CHANGE DOWN 2% RETURN ON COMMON STOCKHOLDERS' EQUITY 9% FIVE-YEAR AVERAGE 9% RECENT SHARE PRICE $46.75 PRICE/EARNINGS MULTIPLE 10 TOTAL RETURN TO INVESTORS (12 MONTHS TO 11/8) 28% PRINCIPAL MARKET NYSE Explanatory notes: page 205 |
|