(FORTUNE Magazine) – Last year during an MSNBC program filmed at Stanford University, Amazon.com CEO Jeff Bezos was asked a question by a shareholder in the audience: "Can you tell me what it is exactly that I own?" In one form or another, that question has been on the minds of Amazon observers ever since the company began morphing from a simple online bookstore to a money-losing purveyor of everything from CDs and Palms to power tools and waffle irons. Was Amazon an Internet company? An online super-retailer? Something else altogether? Bezos' answer was surprising. "What you own," he explained to the shareholder, with his trademark grin and in that pitch of fevered enthusiasm, "is what we like to think of as an incubator for e-commerce companies that can start companies at lower costs and more quickly than any other company in the world."

Whoa! That was news to many in the audience, given that Bezos had only a few weeks earlier begun referring to Amazon as an "incubator." He didn't elaborate at Stanford, but later made clear that what he was referring to was a spate of deals Amazon had just signed with dot-coms like drugstore.com, living.com, Audible, and online car-sales company Greenlight.com. Each was going to pay Amazon big fees for the opportunity to be showcased on Amazon's highly trafficked Website. Unlike the launch of a new Amazon product category, which requires spending lots of money to buy products and build physical infrastructure, these deals were clean. They would make Amazon's business scalable and efficient, just as e-commerce in general was supposed to work. And the fees from all those dot-coms would drive Amazon to profitability, a goal Bezos had repeatedly told investors was a sure thing.

Cut to late 2001. Bezos no longer calls Amazon.com an incubator. He's done with that now--he's busy writing a new script. The big news now is that during the past eight months he and his team in Seattle have signed a flurry of deals with brick-and-mortar retailers like Target, Circuit City, and Borders. Amazon will run all or part of their e-commerce operations. It will sell the retailer's products on Amazon.com, and in some cases it will warehouse products, distribute orders, and run the partner's Website. Unlike the dot-coms, these new partners won't go out of business in a year, and they will pay Amazon in cash instead of rapidly falling stock. Bezos would like us to believe that this time it's different, that Amazon's new "commerce platform" will put the company firmly on the road to profitability--and restore to it a market cap worthy of an Internet superstar. The question anyone who owns or cares about Amazon's stock needs to ask is simple: Should we believe him?

The answer is probably, No. Whichever of Bezos' stories you read, the numbers don't look good.

Start with the original tale, that Amazon was the bookseller to beat all booksellers. The division that sells books (along with music and video) has been a fabulous growth story--until recently. In the third quarter of last year Amazon sold $400 million worth of books, music, and videos; this year in the same quarter it sold only $351 million. The decline is a big problem, because that is Amazon's only profitable division. The problem is even bigger because books and music and video were supposed to be the foundation of the first sequel to the original story, which Bezos was telling in 1998 and 1999. As he opened new "stores"--tools and hardware, health and beauty, toys and games, and electronics--he explained that Amazon would round up huge sales by selling everything people would want to buy online. Purchase a set of mixing bowls when you order a cookbook was the thinking. Wall Street analysts gave the sequel rave reviews, and projected in early 2000 that by the end of 2001 the company would have $5 billion in annual sales. But 2001 sales will actually come in at around $3 billion, and even with all its new stores, Amazon's sales growth has dropped precipitously (see chart). Just as ominous is the fact that Robertson Stephens analyst Lauren Cooks Levitan estimates that quarterly revenue per customer is down from $31 in early 2000 to $18 in the most recent period.

That brings us to the incubator story, another tale that Wall Street swallowed eagerly. "Amazon could sign 25 Greenlight-sized deals over the next two to three years," Merrill Lynch analyst Henry Blodget wrote in January 2000. "At $75 million apiece, these would generate an incremental $375 million of revenue a year for Amazon." Blodget added that all of these deals could contribute $300 million of operating profit.

That didn't happen, and neither did the prediction of Goldman Sachs' Anthony Noto that "if Amazon signed two or three more of these deals, they could break even in 2001." Today, no fewer than six of the companies in Amazon's "incubator" have gone out of business. Most of the fees Amazon was supposed to collect never materialized because the "money" was to be handed to Amazon in the form of stock. (This detail wasn't exactly spelled out in company press releases and is now an issue in an informal inquiry by the SEC and several shareholder lawsuits.) The net effect of the incubator deals? Losses and write-offs.

Which brings us back to Bezos' new tale of brick-and-mortar partnerships with the likes of Target. In recent conversations with FORTUNE, Bezos seemed a bit more circumspect and defensive than in the past. He nonetheless hinted (and his top executives positively swear) that these deals will put Amazon on the "path to profitability," and perhaps even fulfill its unwavering desire to be, as Morgan Stanley analyst and Amazon die-hard Mary Meeker famously put it, "one of the great companies of our age."

Once again, let's turn to the numbers. There aren't a lot of hard data, since Amazon won't predict how much revenue each deal will bring in. So we have to use estimates, and we'll be generous. Let's take Target, which is, with $38 billion in annual sales, certainly one of the larger companies Amazon could sign up. Safa Rashtchy at Piper Jaffray has spent several days looking at the deal, and he estimates that Target is doing anywhere from $50 million to $100 million in sales online each year. Amazon will get a high single-digit percentage of that revenue, plus a per-item charge for warehousing and distributing Target's goods, and a fixed fee for letting Target use Amazon's technology to run Target.com. Given all that, he figures Amazon's cut of Target's online sales could reach 20%. If we take the high end and assume $100 million in sales, Amazon rakes in $20 million. That's not pocket change, but on its own it's not going to get Amazon closer to that $5 billion-a-year figure it was supposed to have reached by now. So what about the cumulative effect of lots of deals? Mark Britto, the senior vice president in charge of those deals, says that Amazon intends to do "hundreds or thousands" of deals with retailers, manufacturers, and catalog companies, and that several dozen are in the works now.

Let's assume that Britto is a master dealmaker, and that Amazon lands another 15 Target-sized deals in the next year. That would mean $300 million in revenue and, assuming operating margins of 20% (which is on the high end of Amazon CFO Warren Jensen's estimate of 10% to 20% on such partnerships), an additional $60 million in operating profits. Is this now a world-changing number? Alas, no. Some of Wall Street's more optimistic analysts do think that would give Amazon, at long last, operating profitability in 2002. But if you were to factor in interest payments from the company's heavy debt load and other expenses, Amazon would still be at least $50 million in the hole.

Neither Bezos nor Jensen wanted to comment on these figures. That may be because looking too closely at the numbers raises scary questions. For instance, if Target.com were a great business, why would the company fork over 20% of revenues to Amazon? Kmart e-commerce chief Richard Blunck outsourced Kmart's Website, bluelight.com, to Global Sports, which is Amazon's main competitor in this business. But he says that the only reason he decided to hand it over is that Kmart's Web sales aren't high enough to justify the company's spending big money online. If Kmart's Web sales ever reach an exhilarating level, Blunck says that the company will likely terminate its relationship with Global Sports and do e-commerce on its own.

Now, let's be clear about something: despite all of that, the brick-and-mortar deals do make sense for Amazon. They leverage the considerable investments Amazon has made in e-commerce technology and in its distribution centers, which are still only half full. And they're a useful antidote to the slowdown in Amazon's retail business. The problem is that these are incremental improvements, and if the company is ever going to be worth, say, $15 billion again (less than half its highest valuation), it needs a whole lot more. Amazon executives know that, and hate comparisons to traditional retailers. Asked if running the Websites of Target, Best Buy, and Borders is kind of like operating a mall, Amazon dealmaker Harrison Miller responded, "I think we're something more interesting than that." Investors disagree. At its current market cap of $2.6 billion, you could even argue that Amazon is slightly overvalued, compared with Best Buy and Costco. They have comparable price-to-sales ratios, but they make money. Even Amazon bulls aren't expecting its new partnerships, by themselves, to enable the company to turn the corner. None of them, not even Meeker or Blodget, raised their ratings on the stock or changed their estimates for revenue and losses in response to the signing of the brick-and-mortar deals.

Bezos' Amazon is a land of eternal hope. The CEO still avers that Amazon can eventually generate operating margins of 10%, despite the fact that it is still miles from that goal. And it is unlikely that Amazon will run out of cash soon, because Bezos' extraordinary talent for convincing people to invest in Amazon should never be underestimated--no one expected AOL to hand over $100 million to Amazon in the form of an equity investment this past July.

So yes, there is a chance that someday Amazon will grow up to be a company with real profits. But it will never be the high-growth, wildly profitable, super-efficient company of Internet lore. The only place that company lives is in the history books, and in the powerful imagination of Jeff Bezos.

FEEDBACK: mwarner@fortunemail.com