Cisco: Behind the Hype

CEO John Chambers still thinks his company can grow 30% a year. But critics question its aggressive accounting

The questions are brutal and surprisingly frank. Each one is a challenge Cisco Systems (CSCO ) CEO John T. Chambers would not have heard even a year ago. "What can I do to combat the increase in politics and empire-building inside Cisco?" wonders a female employee. "Will we see more layoffs?" inquires another. "Does the Internet now have negative connotations?" asks one man.

Such candid employee question-and-answer sessions have been a staple of Chambers' management approach since he became CEO of Cisco seven years ago. But on this cool November morning in San Jose, Calif., in an auditorium filled with more than 200 staffers, the inquiries betray a sagging morale and unease that until recently were unknown inside this high-tech stalwart.

Once one of the world's fastest-growing companies, Cisco is struggling to find its footing amid a severe economic downdraft. The company is in the red. Its sales are slumping, and its market valuation has fallen by some $430 billion, to $154 billion, since March, 2000, one of the deepest losses of shareholder wealth in history. The decline puts many of Cisco's employee stock options under water and limits the company's ability to acquire new companies.

Yet even after this comeuppance, Chambers, 52, continues to insist in public and private that Cisco can still "change the world" and increase revenues as much as 30%-plus a year. Just as the spending boom of 1999 and early 2000 may have been overdone, he says, the dramatic shrinking of technology budgets in 2001 was also exaggerated--and temporary. He points to new markets, from voice-over-Internet systems and wireless networking gear to storage-networking devices and optical equipment, that he believes can fuel Cisco's next round of hypergrowth. "There is a confidence that if we execute right, we have a higher probability of winning now than ever before," says Chambers. "I believe the best years are in front of us."

But skeptics abound, starting inside the company auditorium where Chambers is getting the hard questions. Many are wondering if the ultimate Internet salesman of the 1990s is simply in denial. After all, the Cisco that dazzled investors and business observers at the height of the high-tech bubble now looks like a company that caught a spectacular wave. Its success and Chambers' own rise as a New Economy visionary were a function of a near mania for tech stocks. In those heady days, all the stars seemed to align in Chambers' favor: a massive Internet buildout, a breathtaking stock price, and an era that allowed companies to push the boundaries of acceptable accounting. "Cisco was at the right place at the right time and exploited it very well," says Robertson Stephens senior technology analyst Paul Johnson. "Chances are they won't be able to do it again."

There has always been a good deal of myth-making where Cisco is concerned. At the height of the Internet frenzy, it was the very embodiment of the age. When it came to Cisco, everything seemed faster, bigger, and better. Its sales and earnings growth were second to none. It sold more sophisticated gear over the Internet than any other company as it raced to fill a demand that seemed unquenchable. It could close its books in a day, thanks to its powerful information systems. For 43 quarters in a row, Cisco met or beat Wall Street's hungry expectations for higher earnings. For one brief, heady moment, it became the most valuable corporation on the planet.

Now, in the cold light of the tech downturn, some of those myths are being exploded. As academics and other experts reassess the tech era, they are raising serious questions about how well Cisco and other highfliers actually performed, even at the height of the boom. They say that a history of aggressive accounting, from massive write-downs of assets to the use of now-banned "pooling of interest" accounting for acquisitions, makes it hard to gauge how much many of these companies actually earned from ongoing operations. These bold accounting techniques became standard operating procedure at many companies during the boom. But few applied them as deftly or with greater effect than the much-emulated Cisco.

To be sure, Chambers' company is hardly an Enron in the making. Cisco's accounting practices, while aggressive, are not illegal and have been certified without qualification by PricewaterhouseCoopers LLP, its independent auditor. With $19.1 billion in cash and equivalents, more than double that of competitors Alcatel (ALA ), Lucent Technologies (LU ), and Nortel Networks (NT ) combined, the company boasts a pristine balance sheet with zero debt. Plus, Cisco still holds a dominant share of the $15 billion market for corporate network gear and has recently boosted its market share in some key product categories, including routers. Long term, analysts believe the company can achieve 15% to 20% annual earnings growth.

But Cisco's core business of supplying Internet infrastructure, while solid, is nowhere near capable of generating the 70% growth rates of yesteryear. After all, the corporations and telecom carriers that lined up for Cisco's gear in the '90s are spending much less to expand networks. Many of the telecom players are laden with debt and overcapacity. Some analysts believe the current down cycle for overall corporate capital spending could last years, even when the recession ends. Meanwhile, those that do want to spend can pick up some bargains. For example, eBay Inc. lists more than 3,000 Cisco products that are being auctioned for much less than their initial prices. "The bubble companies that fueled Cisco's growth have largely been wiped off the face of the earth," says Scott C. Cleland, chief executive of the Precursor Group, a high-tech consultant that recently bought $27,000 worth of Cisco gear for $7,000 on eBay. "Chambers' core business has matured, and all the growth markets are speculative." Indeed, in many of the markets that Chambers points to as engines of future growth, Cisco faces large and entrenched competitors. In others, such as voice-over-Internet, it's becoming clear that the potential first envisioned will take much longer to materialize.

Yet while the bubble has burst elsewhere in techdom, Cisco continues to exist in its own separate universe. Legions of true believers still cling to the hope that Cisco will reemerge as the high-tech bellwether that will lead the sector out of its slump. Their continuing faith is why Cisco stock trades at a spectacular 95 times estimated 2002 earnings despite losing money in 2001. A new study by Robertson Stephens estimates that even if Cisco grew at 20% annually over the next 10 years, to $100 billion in sales, and sustained operating margins of 15%, an investor who bought the entire company at its current market value would earn a measly 3% return a year, based on projected cash flow. They'd be lucky to do so well. The chances of hitting that 20% growth target now seem a stretch.

For all the disquieting signs, an almost dreamlike view seems to prevail among the company's Wall Street supporters. In a 90-minute conference call between Wall Street analysts and Chambers on Nov. 5 to discuss first-quarter results, few analysts acknowledged that Cisco had posted a net loss of $268 million or that sales had plummeted 32%, to $4.4 billion. Nor did any of the analysts grill Chambers on the $290 million Cisco disclosed that it earned by selling "excess inventory" written off as worthless only seven months earlier. (At the time, critics wondered if the huge sum included some still-good equipment that could be sold later, inflating the bottom line. "We have no plans to use it, period," a company spokesman told BusinessWeek last spring.) Without the gain, Cisco's net losses would have been nearly half a billion dollars.

It's not just the numbers that deserve greater scrutiny. Part of the Cisco myth revolved around the company's super-sophisticated information systems. Cisco was supposedly using the Internet to bind together its suppliers and contract manufacturers into a seamless whole, pointing the way to the corporation of the future. In fact, Cisco's "network organization" did little to soften the impact of the downturn--or save Cisco from the disastrous inventory buildup.

Similarly, its vaunted acquisition strategy was supposed to make the world Cisco's lab, allowing it to pluck promising technology after someone else had borne the development risk. But many of the companies Cisco snapped up with its inflated stock came with steep price tags, reaching as high as $24 million per employee. Often, high-priced talent walked soon after the deals were done. Many of those companies have yet to deliver new products or markets or contribute to Cisco's bottom line.

Indeed, there are serious questions about how successful many of the high-tech companies were during their halcyon days. Michael Porter, a Harvard Business School professor, is now completing a major research project on the era's massive distortions in financial reporting. One astounding finding: He and his colleagues can't figure out how much money Cisco really made in the 1990s. After spending billions on buying companies, Cisco routinely wrote off massive amounts of acquisition-related charges--$5.4 billion in the past five years alone--making it nearly impossible to piece together how much Cisco actually invested or how much it earned on its capital. Says Porter: "When the historians actually plow through all the data, we will likely find that even during its so-called heyday, Cisco wasn't nearly as profitable in terms of return on invested capital as many believed."

Chambers and his top executives strongly defend the company's accounting practices. "I've got an extremely conservative chief financial officer, and we have been conservative all the way through," says Chambers. "We provide a tremendous amount of information. The more information we provide, the more questions it raises."

Trouble is, even the most sophisticated forensic accountants have had difficulty burrowing through the company's less-than-transparent financial reports. Only recently, for example, the Maryland-based Center for Financial Research & Analysis reported that a 3% growth in revenues from the previous quarter occurred only because Cisco reserved less for bad debts. The center used Cisco's own numbers released in its Nov. 5 conference call. Now, the center is backing away from its report because Cisco says it has actually been putting aside more money for uncollectible receivables, not less. However, not all of its offsetting "revenue adjustments" are broken out in its filings. Concedes Chambers: "We've got to be able to articulate [our numbers] better."

Chambers may be chastened, but he hasn't lost his confidence or charisma. In his upbeat, rat-a-tat style, Chambers moves through a standard stump speech. He says the company's customer focus, culture of empowerment, prowess with acquisitions, and dominant market share will allow it to overcome the odds. "Let's not kid ourselves," he says. "This market is coming to us. It is going to consolidate. Who in this industry is better at consolidation than Cisco? And who, by the way, has the assets that Cisco does?"

Even with questions about the company's past, no one doubts that Chambers can reestablish Cisco as a viable and solidly profitable company. But Cisco's current market valuation is largely based on Chambers' highly optimistic view of the company's growth potential and Wall Street's willingness to accept its more bullish unaudited, pro forma financials. Chambers thinks he can reach his lofty growth targets by taking greater share from rivals and conquering new markets. But the chances of finding another bonanza like network switching, which helped propel Cisco's fast-track growth in the '90s, are slim.

Cisco insiders acknowledge the difficulty of returning to the growth rates of its glory days. "When you're a really tiny company, you can jump into one new market and it can make you a lot bigger," says Michelangelo Volpi, Cisco senior vice-president in charge of Internet switches and services. "You can't do that at Cisco anymore. You can't take a $20 billion company and say one market is going to make us a $40 billion company. In each of those markets, there will be entrenched competitors. They'll know the business well, and they'll be tough."

It's a far cry from the '90s, when the company averaged 70% annual sales growth. But that torrid pace covered up a multitude of weaknesses. Cisco was famed for systems that were supposed to give its managers unparalleled access to real-time data. But when business weakened, those systems proved to be flawed.

Just how flawed became apparent shortly before midnight on Saturday, Jan. 27, last year, as the company prepared to close out its 2001 second-quarter numbers. Analysts were expecting a profit of 19 cents per share, but a slowdown in orders and shortages of certain parts would bring this quarter down to the wire in ways that would belie Cisco's reputation as a superefficient producer in the New Economy.

That night, at the company's main San Jose warehouse, employees scrambled to load as many boxed-up machines onto trucks as possible, thus enabling them to be counted as "sold" for accounting purposes before the quarter ran out at midnight. "We had guys running with parts, trying to run them up to the trucks," recalls Larry R. Carter, chief financial officer, who monitored the action from a computer terminal in the warehouse that tallied shipments in real time. One harried staffer, in his haste, fell headlong to the concrete floor in front of Carter. Ten minutes before midnight, Chambers called Carter's cell phone from Davos, Switzerland, where he was speaking at an economic summit. "Did we make it?" Chambers asked. After a few more minutes of watching the numbers come in, Carter had to pass along the bad news: Cisco would undershoot expectations by a penny per share, its first Wall Street miss in 11 years.

The news shaved 13% off the company's market value in one day, but there was worse to come. It turned out that Cisco's networked-manufacturing model was not nearly as accurate as Chambers had boasted. Only 40% of what Cisco sells is actually made by the company. Instead, a network of suppliers and contract manufacturers delivers an unusually large chunk of Cisco-branded merchandise direct to customers. This business model was supposed to keep fixed costs to a minimum, eliminate the need for inventory, and give management an instantaneous, real-time fix on orders, shipments, and demand.

The highly hyped systems, however, failed to account for the double and triple ordering by customers tired of long waits for shipments. So Cisco began to stockpile parts and finished products. "We made a conscious decision when our lead times were 12 to 13 weeks to build inventory, because we were leaving a sizable amount of revenues on the table every quarter," says Carter. Soon inventories were growing faster than sales. The result: When a weakening economy brought capital spending to a near halt, Chambers found himself stuck with billions of dollars of inventory he didn't expect to have. In April, he wrote off $2.2 billion of excess inventory and cut 18% of Cisco's staff, or 8,500 employees.

Despite the massive write-off, Cisco still holds surprisingly high levels of finished-goods inventory: 26 days of finished inventories in October, three days more than it had in April when it took its big bath of a write-down. It's just another question mark about demand for Cisco's products. Dennis Powell, vice-president for corporate finance, calls the change in finished-goods inventory "inconsequential" and points out that overall inventories have fallen to $1.3 billion from $2.5 billion in the past nine months.

To improve manufacturing and sales forecasting, Chambers says, Cisco has added new checks and balances. It now disseminates more information about demand and product backlog not only to contract manufacturers but increasingly to the hundreds of suppliers who provide the parts for contract manufacturers. The idea: to make sure everyone works off the same information, making it less likely Cisco will see any inventory surprises. In addition, Cisco has added new ways to gauge customer demand. For instance, not only does Cisco cull sales projections by region and business unit, it now double-checks this against monthly customer surveys of expected spending up to 180 days in the future.

The inventory debacle was the first major miss Cisco had. Indeed, there was good reason for its obsession with meeting Wall Street's expectations each quarter. The continual earnings gains kept Cisco's stock price high, which in turn allowed the company to make acquisitions and lure talent. That obsession explains the creative ways Cisco found to account for its purchases and tally its earnings. Among the $5.4 billion in acquisition-related charges it took in the past five years, for example, the lion's share--$3.8 billion--covered "in-process research and development." In some cases, these expenses nearly equaled the cost of the acquired companies. By taking these charges up front, a move that is supposed to reflect that the acquired research and development may prove worthless regardless of the sum paid for it, future earnings tend to look a lot rosier. The reason: If the technology ultimately pays off, the earnings are unencumbered by their true costs, which otherwise would have been amortized as goodwill.

Like many companies, Cisco's financials have long been boosted by the use of "pooling of interest" accounting for acquisitions. By employing this accounting, Cisco significantly reduced the impact of many acquisitions on its books and avoided potentially troublesome write-downs on the acquired assets in future years. When Cisco bought GeoTel Communications in 1999 for some $2 billion in stock, it recorded only $41 million for the deal on its books--roughly the amount of GeoTel's shareholders' equity. Abraham J. Briloff, professor emeritus at Baruch College, estimates that in the two fiscal years ended in July, 2000, Cisco "suppressed a grand total of $18.2 billion in costs" by using this method of accounting. "It inflates their subsequent results to the extent that they avoid having to charge off everything from inventory, patents, licenses, plant, equipment, and goodwill," says Briloff. Cisco disputes Briloff's analysis, maintaining that the success of many of its acquisitions might not have led to write-offs and could just as easily have resulted in an increase, rather than a decrease, in goodwill.

One of Cisco's most frequent accounting tactics comes up every quarter when the company directs shareholders to its unaudited pro forma earnings numbers as the best gauge of profitability. In the first quarter of fiscal 2001, for example, Cisco reported pro forma earnings of $1.4 billion, nearly $600 million over its net income. It arrived at that tally by excluding such ordinary and important costs as acquisition expenses and payroll taxes on stock-option exercises. In the fourth quarter, Cisco's pro forma income of $163 million was 23 times its actual net earnings.

While liberal use of more malleable pro forma earnings is not a rarity, most high-tech stalwarts, from Microsoft (MSFT ) to Oracle (ORCL ), do not report such unaudited numbers. The Securities & Exchange Commission recently cautioned companies and their investors about the potentially misleading metrics, warning companies that they could face civil-fraud lawsuits. "It's like students deciding the process of how they're graded so that they can always get an A from the teacher," says Howard Schilit, who directs the Center for Financial Research & Analysis. Cisco's Carter maintains that pro forma numbers more accurately portray the company's operating results because they exclude volatile charges, such as acquisition expenses. "The more information you give investors, the better," says Carter. "It's sort of like `What do you like, vanilla or chocolate? We're going to give you both so you can choose."'

Even the quality of Cisco's pro forma earnings is deteriorating. Well before Cisco's write-down in April, the company was relying on nonoperating income for a growing portion of its pro forma earnings. By the January quarter of 2001, for instance, interest and other nonoperating income had increased as a percentage of Cisco's pro forma results for each of the past five quarters, rising from just 8.8% to 14.9%. In the latest quarter, more than a third of Cisco's pro forma earnings come not from operations but from interest and other income.

Accounting issues aside, Chambers' real test will come in the next few quarters as Cisco attempts to wring greater profits out of existing products and step into new growth areas. Given the company's valuation and its current size, Cisco will have to dominate not one or even two but several of these new markets to meet its growth goals. That's no small feat given the slew of big-name competitors, from Nortel Networks to Nokia, going after these same markets. And as the corporate-networking market--Cisco's bread and butter--continues to mature, the pressure on Chambers to conquer new markets will only increase.

The most promising growth industry could be voice-over-Internet systems. By using Internet-based phones, or so-called IP phones, a company no longer has to invest in building and servicing a separate voice network. In addition, IP users can easily combine voice and Internet features--such as accessing e-mail via telephone--since they spring from the same network. About 7% of the new underlying phone-switching systems, or PBXes, sold to corporations are Internet-based, made by companies like Cisco. That's expected to jump to 19% of the $10 billion market by 2004, according to International Data Corp. "Voice alone could represent $10 billion [in revenues] a year for us," in the next five or six years, says Volpi. Reaching $10 billion so quickly may be more than a stretch goal. After all, the overall market is still tiny and Cisco's corporate customers have been slow to adopt this technology. "It's a slow road ahead," says International Data analyst Tom Valovic. The optical-networking market, pegged as another future growth engine, has much less potential. Cisco already trails a host of competitors, including Lucent, Alcatel, Nortel, Fujitsu (FJTSY ), and Ciena (CIEN ). Although Cisco claims a strong position in the local rather than long-haul part of optical networking, even the local segment suffered its first-ever sequential decline in the third quarter of this year.

The wild card in Cisco's growth plans is its effort to woo telecom companies. These massive purchasers of networking gear have largely stayed loyal to traditional vendors such as Nortel and Lucent. Cisco, at its peak, was generating 40% of its business selling to telecoms, many of them upstarts challenging incumbents like Sprint Corp. (FON ) and Verizon Communications (VZ ). But as many of these outfits faltered, Cisco has become more dependent on established corporations. Upstarts, however, continue to haunt Cisco. El Paso Global Networks Co., for example, announced in October it was putting the brakes on a massive billion-dollar network, of which Cisco was slated to be the primary supplier. The reason? Just too much capacity in the market.

But Cisco's struggles with carriers have gone well beyond market conditions. To get back on track, Cisco is rethinking its approach to this market. First, it is aiming nearly all of its resources on the 50 biggest carriers. Second, it's trying to forge relationships beyond the executive suite with the networking and operations people who often influence which vendors are used.

Perhaps Chambers' biggest challenge is the switch from leading hypergrowth to grappling with issues of cost control and productivity that will restore the company to profitability. Some question whether the cheerleading CEO who has only known 50% growth since taking over the helm in 1995 can stomach an environment of disciplined cost cuts and layoffs. The decision to lay off 8,500 employees last March weighed so heavily on Chambers that he mulled over its consequences on his treadmill one night at 2:30 a.m. "He's almost too nice to kick the asses that need to be kicked," says John Thibault, a former Cisco executive and colleague of Chambers' at Wang Laboratories during the late 1980s.

If Chambers is worried about the tough decisions ahead, he's not letting on. "Last year was tremendously humbling. It knocked us on our tail. But if we execute right, our future is very, very bright," he says. "Cisco is clearly positioned to break away." Investors buying up Cisco stock at 95 times earnings are banking on it. But Chambers will need much more than bold accounting and long-shot bets on fledgling markets to deliver.

By John A. Byrne and Ben Elgin

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