hat if Microsoft had a bad quarter? I just spent a day at Microsoft, on its amazing campus in Redmond, Washington, getting my regularly scheduled brainwashing. I've been doing this for at least six years now, spending one or two days reviewing Microsoft's products, talking to product managers, and getting refreshed on the state of the world according to Microsoft.

This visit was pretty typical. I reviewed Microsoft's new efforts in operating systems, Internet tools, and online service and content. I discovered that Microsoft is close to having its own rapid-transit system--close, I say, because the only thing em ployees seem to complain about is the quality of the company's fleet of shuttle buses. I learned that Microsoft is moving just as fast and aggressively as it ever has, despite the fact that it is now a truly giant company.

Since Microsoft is king of the hill, the game among analysts is to be the first to find the hole in Microsoft's strategy. It's an extremely frustrating game, of course, because Microsoft has been on such an incredible winning streak. Last quarter, for example, the company once again delivered outstanding results, earning $741 million on $2.68 billion in sales. It's hard to remember the last time Microsoft made a really stupid mistake.

In fact, many analysts have given up trying to find the hole in the strategy. Instead, they operate on the assumption that Microsoft is invulnerable. It always grows. It always gains market share. It always makes more money. As long as the computer ind ustry is growing, there seems to be zero chance that Microsoft can have a bad quarter.

But as I was walked through the Microsoft campus, the thought occurred to me: What if something finally did go wrong--this year? Floods, plagues, death and dismemberment--that's what. A Microsoft downturn would be just as big as the federal government' s announcing that we were facing a severe recession. The computer industry would be in serious trouble. The world might wonder whether the technology boom that has fueled so much growth was coming to an end. Wall Street would get really unhappy. Venture c apitalists would wonder whether manna would stop falling from heaven.

If nothing else, all those happy nerds in Redmond would begin to experience the first twinges of self-doubt and uncertainty.

So the question is: Is there a credible scenario in which Microsoft might have a bad quarter? I believe there is, but I wouldn't want to bet on it. Essentially, four things have to happen in the right order to make Microsoft look bad. These things can' t really line up before next fall, so fiscal 1997, ending in June, is safe. The bad quarter probably wouldn't arrive until the second quarter of fiscal 1998, ending in December.

Here's what would have to happen:

Office 97 must fail to meet expectations. Microsoft Office represents a significant portion of Microsoft's total revenue, as much as 30% last year. The company needs a lot of people to upgrade from Office 95, but there isn't a compelling reason to do so. The new product pretty much does the same stuff as the old one. It works more intelligently and is more fully integrated with the World Wide Web. But Microsoft made one terrible mistake: Documents created in the Word and Excel components of the new Office can't be read by owners of old versions. That will create real headaches in offices where only some users upgrade. That pain, combined with the lack of a compelling reason to upgrade, will slow the growth of Office 97.

The price of application software must fall dramatically. This happened once before, around 1990. Ironically, it was Microsoft that made prices crash, by introducing the idea of bundling individual applications into a suite. The original Office reduced the price of each application from about $300 to about $100. That change forced an industrywide reduction in the price of applications.

Another repricing of software is now under way, spurred by the World Wide Web. On the Web, software is given away at first; then companies try to develop a business by selling upgrades and support. The net effect is that people are beginning to think t hat individual pieces of software should cost less than $40. This shift may well force Microsoft to reduce its prices.

The company must miss its shipping deadline for new versions of Windows. Upgrades of both Windows 95 and Windows NT are supposed to come out later this year. Microsoft has already demonstrated both upgrades at industry conferences, showing that much of the work is already complete. But Microsoft has a history of shipping software three to six months late. I think most corporate buyers are expecting Windows 97 this spring--they'll probably get it sometime late this fall. If Microsoft is that late , companies might delay PC purchases, which could flatten growth for the whole industry, and flatten Microsoft revenue growth as well.

People must become less enthusiastic about computers. Microsoft has no effective competition for PC operating systems or applications. Its businesses in software tools, server systems, and applications are booming. So even if it runs into troubl e with Office and Windows 95, there needs to be a bigger event in the world to get the company into real trouble. The only event that could give the company trouble is if people really slowed down purchasing computers.

So there you have it: Cross your fingers on both hands, stand on one leg, and drink a glass of water upside down, and Microsoft will have a bad quarter. I guess even this curmudgeon has to admit defeat.

STEWART ALSOP is a partner with New Enterprise Associates, a venture capital firm. Neither he nor his partnership has financial interests in the companies mentioned. Alsop can be reached at stewart_alsop@fortunemail.com