What a year, eh? COMDEX 2004 has been "postponed" and Microsoft's giving away cash to shareholders. Call me paranoid, but I'm starting to wonder whether we're at the end of an era.
Ever since 1979, COMDEX (short for Computer Dealers' Exposition) has been the one-stop-shopping place for all the latest and greatest stuff in the small computer business. Some of the most significant new items in the industry (e.g., Turbo Pascal, the Palm Pilot) made their debuts there. But in late June, the show's current owner, MediaLive International, announced (basically) that it couldn't see the point in running COMDEX this year. Now, if that didn't make you sit down and catch your breath, then you haven't been in the business very long. I mean, canceling COMDEX! For geeks like me, that's like being 8 years old and hearing that they're canceling Christmas.
I attended my first COMDEX in 1987. Back then, there were two events: one that took place around April and one that took place around November, known in the United States as "Spring COMDEX" and "Fall COMDEX," respectively. Fall COMDEX was the "real" one, the big one. Spring Comdex alternated between Chicago and Atlanta—I recall a snowstorm during one Chicago show that caused the attendees to dub it "Winter COMDEX"—but the big announcements and appearances always happened at Fall COMDEX, which took place in Las Vegas. The show grew nearly every year. It got so that if you wanted a hotel room during COMDEX week, you needed to reserve the room almost a year in advance and pay for it up front (no refunds allowed). Getting from one part of town to another during that week was nightmarish, and every attendee had a story about waiting more than an hour for a bus from the convention center to a hotel or about how much their feet hurt from walking for what seemed like miles around the convention floor. (I always figured that I could substantially increase my annual income by setting up a Rockport shoe concession at Fall COMDEX, but I never got around to it.) COMDEX was also the great leveler of the PC business. I met more of the industry's big names on those mid-November days than at any other time during the year. What other show came with a guaranteed appearance by Bill Gates?
So, what happened? The more I tried to figure out the answer to that question, the more I began to believe that it all makes sense in light of Microsoft's announcement earlier this year that it intends to give some of its cash to shareholders and use some of it to buy back Microsoft stock.
The very profitable Microsoft has no debt and a ton of cash, and things have been that way since the company's start. Up to this year, the company has reinvested all that cash back into the business, fueling pushes into new areas. (Did you notice how well I avoided the phrases "world" and "domination" in those sentences? Pretty impressive, eh?) This "cash as fuel for expansion" strategy has worked to the company's advantage over the years. Microsoft didn't have the first or the best GUI, but it kept funding version after version of Windows until it wound up with something acceptable. I'd argue that that was Windows 95, which grew out of Windows for Workgroups 3.11, Windows 3.1, Windows 3.0, Windows 2.1, Windows 2.0, Windows 1.0, Windows 386, and OS/2 1.x. Not many firms could afford that many rough drafts, but Microsoft could—and as a result, today's Windows GUI is pretty good. The same story applies to networking: Microsoft's early entries (i.e., the IBM PC Network Support Program, Artisoft LANtastic, and LAN Manager) cost it a lot of time and money, but thanks to that cash cushion, the company had that time and money. Those early products eventually led to Windows Server 2003, a high-quality networking platform. (Microsoft's been trying to use the same ploy to take over the mobile-computing world with Windows CE and PocketPC. So far, these platforms haven't overtaken Palm OS, but it could happen.)
But wait. Now Microsoft has announced that it intends to spend less of cash on implacable and indefatigable development of new category-killers and redistribute the cash instead. Why would it do that? One reason might be, of course, that the Powers That Be at Microsoft want to buoy up the company's stock price. Or, conversely, they might think that the stock price is too high. I'm grossly oversimplifying here, but many investors feel that one way to pick a good stock is by looking at its profit-to-earnings ratio (P/E). A stock with a P/E of 15 or lower is a usually thought a good deal; a stock with a higher P/E is considered to be overpriced … unless the investor expects that the company can grow at an unusually high rate. In other words, if a company is just a run-of-the-mill-successful company, don't buy the stock if the P/E is over 15. But if the company is capable of growing at, say, 100 percent a year, even a P/E of 70 might be justifiable. Microsoft's P/E has always been high (it's currently running at around 40), but that's never hurt the company's share price because people believe that Microsoft can continue to create new markets and grow. But does Microsoft still believe that? It's surely one explanation for the company's new cash strategy.
So I'm prompted to wonder: Is the computer business no longer a growth industry? We've talked about limits to growth for ages, but it's always been just talk. Do the COMDEX folks and the Microsofties know something we don't know? I sure hope not. It'd be pretty boring to work in a "maintenance industry." Let's hope that biological computing, search engines, man/machine interfaces and natural language processing come up with something exciting soon.