Thank God for Nortel. Without it, Celestica Inc. would be Canada's poster child for tech-bubble euphoria and subsequent monumental failure.
And without Nortel, Toronto buyout king Gerald Schwartz and Craig Muhlhauser, the CEO of Celestica, might be feeling a lot more anxious these days. Schwartz, as head of Onex. Corp., is controlling shareholder of the beleaguered electronic and computer components manufacturer; last December, after six straight years of losses and five unsuccessful restructurings, he turfed CEO Stephen Delaney, and put American engineer and turnaround veteran Muhlhauser in the top job.
Pick some numbers, just about any numbers, and you'll see that the reversals have been astonishing. Schwartz took Celestica public at $25.73 a share in 1998, and the price on the Toronto Stock Exchange soared to more than $120 in 2000. The company's earnings rocketed up as well, from a loss of $48.5 million (U.S.) in 1998 to a profit of $206.7 million (U.S.) in 2000. That put Celestica at No. 49 in Report on Business magazine's Top 1000 profit ranking for that year.
The crack-up that followed was just as spectacular, but the hangover has lasted much longer than the party. Celestica has racked up more than $1.8 billion (U.S.) in losses since 2001, and its share price on the TSX has slid to record lows, near $7 recently. Since 2001, Celestica has also finished in the 900s in The Top 1000 ranking each year, including dead last in 2004, with a loss of $854 million (U.S.), and No. 988 this year. Even Nortel, still troubled, has looked better lately, moving up to 294 from 1,000 last year.
Schwartz, 65, doesn't say a lot publicly about Celestica these days, and he declined a request for an interview, deferring to Muhlhauser. For his part, Muhlhauser, 58, hasn't done a lot differently than his predecessors during his first six months at the helm-more restructuring and layoffs, particularly at a large plant complex in Monterrey, Mexico, that had an operating loss of $75 million (U.S.) last year alone. But he's upbeat-indeed, he's a master of rapid-fire management-speak who can make even painful setbacks sound positive. "Down markets should create the greatest opportunity," he says. "I classify the situation we're in as opportunity-rich."
That's spin, yes, and hearing it, you can almost see analysts rolling their eyes. This spring, several of them hit Muhlhauser with investment rating downgrades, and no one is forecasting a quick turnaround. On the other hand, unless you're a long-suffering shareholder or have tracked Celestica closely, you may not realize how far out of sync with reality analyst and investor perceptions of the company have been over the years. The most intriguing disparity is that Schwartz has been the biggest beneficiary of Celestica's epic share price decline.
Back in the mid-1990s, both Celestica and Onex were on a roll, and the two looked like a perfect match. Celestica had a long history in the computer business as part of IBM Canada Ltd. IBM opened its first office in Canada in 1917. After the Second World War, it built a massive block-long factory in the then almost futuristic Toronto suburb of Don Mills. The plant made mainframe computers and peripherals, and the building-enlarged several times since-is still Celestica's Toronto location today.
In the 1980s, however, IBM planned to close the plant. Eugene Polistuk, the veteran engineer who headed the factory, fended off that fate by focusing the plant's product line on items like circuit boards. In doing so, he entered the electronic manufacturing services (EMS) market. It was a gutsy move-it meant shifting from being an in-house supplier to competing in a much tougher market as a supplier to all manner of electronics companies. But it paid off. Sales reached almost $1.5 billion (U.S.) in 1993, and IBM made the company a separate subsidiary the following year, naming it Celestica-a word conceived to convey optimism and open skies.
Onex, launched in 1984 by Schwartz and a handful of associates, specialized in buying businesses that looked undervalued or in need of restructuring, or were being spun off by parent companies. The Onex model, then as now, is to take a controlling interest, slash costs and streamline operations, and then sell out a few years later at a profit.
In 1996, Onex led a group that bought Celestica for $750 million, including a $200-million investment of Onex's own money. Under Onex's guidance, Celestica soon bought factories in Britain, Ireland and the United States. By the end of 1998, the plant count was 20. In June, 1998, Onex took Celestica public with a $610-million IPO. That left it with roughly a quarter of Celestica's shares, but Onex-and through it, Schwartz-have maintained control with multiple voting shares that have 25 votes apiece. Schwartz's shareholdings and voting stake have varied somewhat since 1998, but at the end of 2006, he still had 78.9% of the votes while owning just 13% of the stock.
Investors certainly didn't begrudge Schwartz those multiple-voting shares at the start. Why would they? Celestica's share price climbed by 20% within a month of the IPO and kept rising for another two years. By 2000, it looked as if Schwartz and Polistuk could do no wrong. That summer, analysts were giddy after Celestica signed a big contract to sell components to then red-hot fibre-optic telecommunications equipment manufacturer JDS Uniphase. Even as Celestica's share price approached $100-well over 100 times the company's earnings per share for the previous year-many analysts maintained a "strong buy" rating on the stock.
Celestica's revenues for 2000 shot up to close to $10 billion (U.S.). Investors were ecstatic, and the company's market capitalization surpassed $20 billion, roughly the same as the Bank of Montreal. Even now, however, Polistuk doesn't think Celestica's share price was excessive, given how fast the company was growing. It had tripled its revenues in three years to almost $10 billion (U.S.) in 2000. He recalls speaking to a conference of investors and analysts at New York City's World Trade Center that year, and forecasting that revenue would double within a few years to $20 billion (U.S.). "The exuberance was so high," he says, "the analysts were arguing with me: Why was I so cautious?"
But Schwartz is a very savvy investor, and the rapid ascent was making him nervous. The value of his stake in the company was close to the value of Onex as a whole-which meant too many eggs were in the same basket. So Schwartz came up with an ingenious hedging strategy for limiting the impact of any decline in Celestica's share price, while maintaining voting control. Onex issued debentures-unsecured bonds-worth a total of $729 million. The debt could be repaid either with cash or 9.2 million one-vote Celestica shares at Onex's discretion.
This amounted to Schwartz placing a strategic bet against his own team. If Celestica shares kept climbing in price, as so many investors at the time expected, Onex could pay back the cash and come out more or less even (minus a few million in annual interest costs on the debentures). But if Celestica's share price declined, Onex could repay the debt in shares worth less than the sum it had borrowed, and pocket the difference. The bigger the price decline, the bigger Onex's gain.
It turned out to be a whopping gain. After the tech bubble burst in 2001, Celestica's losses mounted and its share price melted. By February, 2005, the shares were trading at about $15 apiece on the TSX. Schwartz decided to redeem the debentures for a profit of $560 million for Onex, along with some forward sales contracts that had provided another clever hedge, and which yielded a gain of $191 million.
Schwartz cashed in on both the greed during the tech boom and the fear during the tech bust. Were those strong feelings ever justified? What both the cheering and the wailing have obscured is that, in good times and bad, Celestica has been-and still is-in a gruelling, risky and very unsexy business.
Getting a clear picture of what Celestica makes is no easy task. Here's how the company put it: "Products that Celestica manufactures can be found in a variety of end products such as: networking, wireless, telecommunications and computing equipment; hand-held communications devices; peripherals; storage devices; servers; medical products; LCD televisions; printers; and in-flight entertainment and guidance systems. Celestica also...offers end-to-end solutions including design and engineering, systems integration, fulfillment and after-market services, including managing end-of-life products for its customers.
"Some of Celestica's customers include Alcatel-Lucent, Avaya, Cisco Systems, EMC, HP, IBM, Microsoft, Motorola, NEC, Raytheon, Research In Motion and Sun Microsystems."
Celestica also offered two specific examples to explain what it does: It manufactures Microsoft's Xbox 360 electronic game consoles, and computer and entertainment systems made for airline passengers by Thales Avionics.
Hoover's Inc., a U.S. business research company, has a simpler explanation: Think circuit boards that go into computers, telecommunications networks and other devices. "Celestica may not be the mother of all board makers, but it definitely ranks among the leaders," says Hoover's.
The trouble is that all those leaders, which include giants Solectron and Flextronics, face the same basic problems. They supply components that go into snazzy, high-margin new-tech doodads (fibre-optic networks in the '90s, BlackBerrys and other hand-held devices these days), but most of those components themselves are low-margin commodities. Instead of having a unique branded product it can corner the market on, Celestica can lose orders at any moment if any of its competitors anywhere in the world can supply a component more cheaply.
And there are a gazillion components to compete on. Muhlhauser says that Celestica's Mexican plant produced items with 50,000 different product codes. He's managed to halve that.
Many production runs in this business are short, and manufacturers have to be able to adjust quickly to changes in demand without letting their own inventories of raw materials and finished products get too large. Customers are also getting more demanding, thanks both to competitive pressures in their own industries and to their ability to play competing
EMS firms against one another. Broadly speaking, for EMS
companies, the North American communications and computer industries were strong customers in the '90s, but weaker since the tech bust. The consumer market for hand-held wireless devices has strengthened, but margins there are slimmer.
And over all, the sector's sales are still weak. In 2002, Celestica's revenues plunged by 17% to $8.2 billion (U.S.). Since then, they've climbed slowly back to $8.8 billion (U.S.) last year. This spring, Celestica revealed that it had lost orders from Nortel Networks and Alcatel-Lucent, seeding speculation that accounts with other customers, including Cisco Systems and even mothership IBM, were also at risk.
Muhlhauser declines to discuss specific customers, but he acknowledges that Celestica has tried to concentrate on having fewer, larger clients, and is willing to simply let some others go. Or, as he puts it, "Customer complexity was driving a level of cost that created inefficiencies throughout the operation, so we moved a number of customers out of Mexico, both through planned disengagements and program transfers."
For the most part, Celestica has responded to weak demand the only way battered and desperate Canadian manufacturers seem to know how: Slash and shift. The slash since 2000 amounts to 28,000 jobs, leaving more than 40,000 employees distributed across 40-plus factories worldwide. The shift is to lower-wage countries such as Thailand, India, China, the Czech Republic and Romania. At the end of 2006, 85% of Celestica's employees were in those regions, up from 60% in 2002.
Much of the shift has been driven by customers that are moving their own production overseas. In 2001, 62% of Celestica's revenues came from the Americas; it's now 35%. Meanwhile, Asia has climbed from just under 10% of revenues to more than half. Polistuk says that even before the tech bust, Celestica had forecast that the EMS industry would have to shift production overseas, and had planned to do that gradually. But after the bust, many of its customers were demanding an immediate move. "Suddenly there was a whirlwind, and nobody could move fast enough to make the shift," he says.
The rush to outsource can backfire. Muhlhauser acknowledges that Celestica has lost money in Mexico because it moved too much production there too quickly. So he has cut the work force to 4,000 from 6,000, and is bringing in managers from profitable factories in Asia to streamline operations.
So when will Celestica's top line, bottom line and share price improve? Muhlhauser is quick with the grand strategic concepts. "It's a journey here," he says. He divides that journey into three legs-"stabilize, standardize, optimize." And he is willing to set some dates. "We're in the stabilize period. And we'll be moving, as we get into the second half of the year, into the standardize period," he says. "As we come out of the year, and move into the next year, we'll be moving into the optimize period." Muhlhauser's optimism is relentless. But the factor that may help him most is just how low almost everyone's expectations are for Celestica.
Muhlhauser was certainly put through the wringer in his first 90 days as CEO. In January, he conceded that Celestica's first-quarter 2007 results would likely be "an absolute disaster." Days later, he dismissed five senior executives. The Moody's and Fitch bond rating agencies soon downgraded Celestica's debt, and several analysts lowered their recommendations. Merrill Lynch analyst Steven Fox said Celestica's competitors could "smell blood in the water." By the end of May, the water had at least cleared somewhat. The consensus recommendation on Celestica from 20 stock analysts polled by Zacks Investment Research was "hold"-hardly a ringing endorsement, but maybe a sign that the worst is over.
Transforming large, troubled organizations-and convincing skeptical customers to stick with them-is nothing new for Muhlhauser. He spent much of his early career as a business-to-business sales specialist at aerospace companies like Pratt & Whitney. In 1997, Ford Motor Co. put him in charge of marketing for its in-house parts supply operations, which it consolidated and spun off as a separate new company, Visteon Corp. In 2001, he was hired as CEO by troubled Exide Technologies, one of the world's largest manufacturers and recyclers of batteries. He led the company into U.S. Chapter 11 bankruptcy protection in 2002, and out of it in 2004.
Among the challenges Muhlhauser faces at Celestica is how to buck up employee morale after years of layoffs, first under Polistuk, who retired suddenly in January, 2004, after a disappointing quarter, and then Delaney. Muhlhauser admits he's got his work cut out for him. He figures the work force can be divided into three groups: One-third are engaged. Another third, he says, "have seen guys like me come and go, so they sit down, they watch us, they make up their mind. And so we work on getting them onside." Then there is the third who "don't feel like they want to be part of this thing no matter what we do, and they really are kind of just biding their time. Maybe they're hoping to find another job or hoping to get laid off. And, you know, when I talk to people, I say I'm looking for them."
Beleaguered companies in beleaguered industries often consolidate, of course. Indeed, Singapore-based Flextronics bought its smaller U.S.-based rival, Solectron, in June. Celestica's share price jumped from $7 to $7.50 on that news, and one analyst forecast that the company might be the next takeover target. But the share price drifted back down over the next several days, reflecting investor doubts that anyone would take a run at Celestica.
The EMS sector is so beat up that even acquisition-hungry leveraged buyout firms and private capital outfits that seem to have bought up everything else in sight have steered clear. Of course, few financiers in North America are more adept at LBOs and corporate makeovers than Gerald Schwartz. But after cleaning out Celestica's executive suite, he doesn't appear to have any desire to buy or sell. And, having closed out his hedge, he'll only benefit if Celestica's share price climbs again. He's betting on it.