(FORTUNE Magazine) – You may think you've heard all the crazy-quilt news there could possibly be about Priceline.com: weakness in its core business of selling airline tickets; the demise of its grocery- and gasoline-selling licensee, Priceline WebHouse; Delta Air Lines' urge to bail out of its Priceline stock; enough customer complaints to get Priceline pilloried on television's 48 Hours.

Well, wait, because we have another bulletin. It's a combination of news (which we'll get to soon) and, perhaps more important, insight. The fact is that Priceline may always jolt the market, because it is so much the product of its unpredictable founder, largest shareholder, and vice chairman, Jay S. Walker. Furthermore, it's not just Priceline that's in Walker's kitchen. He has cooked up what lawyers generically call a corporate "plate of spaghetti." Priceline and WebHouse are strands in this mixture, and so are other Walker companies. Like real spaghetti, this dish is hard to handle. If it weren't, we wouldn't have so much strange news about Priceline.

One example of strangeness--and the path to the news we have to report--is Walker's much ballyhooed sale of Priceline stock in August and September to three big investors. First, he sold John Malone's Liberty Media and Paul Allen's Vulcan Ventures a total of eight million shares of Priceline at $23.75 a share. Liberty's part of the deal was roughly two-thirds of the stock, for a cost of $126 million. Vulcan had the rest, $64 million. The eight million shares were about 16% of Walker's direct holdings and the first he'd ever sold. A bit later, on Sept. 11, Walker offloaded another $50 million in stock: two million shares, at $25, to Saudi Arabia's Prince Alwaleed. Walker also handed over some vigorish, giving all the buyers options to buy equity in Walker Digital, a closely held think tank he controls--and yet another strand in the spaghetti.

Then, within 16 days of the prince's purchase, Priceline announced revenue shortfalls. Days later Walker closed WebHouse for lack of cash. And, bam, Priceline stock was selling for $5.

After that, the loudest sound was silence from the three buyers. They didn't rage or even grouse--the dogs didn't bark, as the saying goes--and it was puzzling why they didn't. Sure, the transactions had a twist to them: They were forward contracts in which the buyers couldn't take delivery of their stock for at least one year, and therefore couldn't sell the stock until then even if they wanted to. But the fact remains that the buyers paid $240 million up-front for stock that, within an Internet minute, fell to about $50 million. Yes, these guys are big boys who ought to be able to take their licks. And yes, Malone and Allen have backed Walker for a couple of years and have made money. But human nature being what it is, you'd think they'd be angry, really angry, about their transactions.

If they were, you couldn't tell it from their public comments. Malone and Allen didn't address the subject at all. The prince spoke calmly of the market's "overreacting" and expressed his faith in Priceline's long-term prospects. Discussing WebHouse's closing on CNBC, Walker dismissed any notion that the three buyers had been tricked. "I've talked to them," he said, "and they are confident I didn't mislead them, because I didn't."

Okay, let FORTUNE deduce why the dogs weren't barking. First, the big buyers were surely embarrassed at having made what so quickly took on the look of a stupid deal--and chose to be quiet about it. Second--here's our news bulletin--FORTUNE has learned that behind the scenes, Malone's Liberty Media and Allen's Vulcan Ventures made a side agreement with each other that shuffled the risks of their Priceline stock in ways they probably didn't care to make public. When pressed by FORTUNE for information, they proffered only a little. Priceline was even less inclined to give out details. "We're not supposed to talk about that," said an insider.

Here's our understanding of the side agreement: Malone's Liberty had more of a tolerance for Priceline risk than Allen's Vulcan. Therefore, Liberty agreed that if Vulcan has lost money on its Priceline stock at some settling-up date in the next two years--we're not sure precisely when--Liberty will absorb any loss up to $12.8 million (20% of the $64 million Vulcan put into Priceline in August). Conversely, if the stock has risen above the purchase price by the settlement date, Vulcan will pay the first $12.8 million of its profit to Liberty. Meantime, if Liberty has to fork over money to protect Vulcan, Liberty will get some of the Walker Digital options that Vulcan now owns. Liberty, in fact, seems to have been much more interested in those options than Vulcan was.

Oddly, of the three buyers, John Malone is at least temporarily the big loser--not a spot he is accustomed to. He's riding the caboose because he not only bought more Priceline stock ($126 million) than Vulcan ($64 million) and the prince ($50 million), but because he then contracted to take on $12.8 million of Vulcan's risk. With the stock down 75% from where everybody bought, Malone is looking at a major flop. Of course, Priceline could be way up by the time the contracts end.

The prince has got his 75% loss too, and as far as we know--he declined to comment to us--none of it is covered by downside-protection agreements. We have a suggestion for the prince: Check in with that guy who prevents embarrassment by providing protection. John Malone's business phone is 720-875-5400.

Vulcan is looking like the smartest of the three investors, if anyone in these star-crossed deals can be called smart. But Vulcan's president, William Savoy, has us confused. He told FORTUNE that arrangements he had made covered all but a "very small" part of his Priceline loss. That may be the case: We are by no means sure we know every side agreement that exists between the principals in these transactions.

But from what we know about Vulcan's $64 million purchase, that company is hardly sitting in the catbird seat. Right now it has about $48 million in losses, with John Malone potentially covering only $12.8 million. The remaining $35 million of loss does not sound "very small." FORTUNE could not get Savoy to talk about that. In fact, as we began asking questions about who had loss protection for what, we could not get a full story from anyone.

Strangely, the fellow who appears to have ducked this particular bullet is Walker. When he first made plans to sell his Priceline stock, he considered offering buyers some protection against loss. Then he backed away from the idea, which, considering what has happened to the stock, was certainly a good economic decision.

But in another way, Walker went before a firing squad. The one crystal-clear point here is why Walker sold $240 million of stock in the first place: He desperately needed the capital to bankroll WebHouse. You get the picture. Walker raised the cash, put significant amounts of it into WebHouse--and then realized that even this couldn't keep WebHouse afloat.

Whatever else they do, Walker's deals with the big investors illuminate the complexity of his world. Good-looking and extremely articulate, Walker, 45, has been an entrepreneur all his life--progressing from lemonade stands in Queens to subscription hawking at Cornell to one new venture after another as an adult. He failed sometimes but scrambled expertly and stayed liquid. Then came the Internet, a world in which creativity and innovation were kings, and where entrepreneurs could spin endless dreams because nobody, including those who spun, knew how to distinguish hogwash from brilliance.

Walker emerged with a collection of moneyed backers and an empire of intertwined companies in Connecticut. Priceline, born in 1997 and poised to take in more than $1 billion in revenues this year, is the only one public. Walker is also the largest shareholder in Walker Digital and in the mortally wounded WebHouse. He is, as well, the co-founder and a significant stockholder in a direct marketer of magazine subscriptions, Synapse Group (once called NewSub), whose major investors include Time Inc., FORTUNE's parent. Time Inc., which owns 23% of Synapse, bought most of that from Walker himself last summer when he was corralling money to put into WebHouse.

All the Walker companies--we are now forking into the spaghetti--do business with each other. Walker Digital has licensed Priceline certain patents, and Priceline in turn was sublicensing some of them to WebHouse. Priceline buys research and development expertise from Walker Digital, and Walker Digital buys management help from Priceline. Synapse extracts subscriptions from Priceline's Website. Royalties, fees, and warrants whisk back and forth, and the mind reels at keeping it all straight.

Earlier this year, apparently aware that the spaghetti could get tangled, Priceline set up an independent board committee to monitor all the relationships. Priceline's last proxy statement lists the committee's chairman as Nicholas J. Nicholas Jr., former CEO of Time Warner and a significant Priceline shareholder. But Nicholas pointed out at the first meeting that he also had a small investment in Walker Digital and probably didn't belong on the committee. The new chairman is a man who hardly needs Priceline duties to fill his time: Paul Allaire, just returned as chairman and CEO of ailing Xerox (see page 141).

Another subject that taxes the mind is sorting out who owns what in the Walker establishment. Many of the ground-floor backers have anted up for more than one Walker enterprise. Among them, besides Nicholas (who has an investment in Synapse as well), are Paul Allen; Malone's Liberty Media; a Connecticut venture capital firm called General Atlantic Partners (which is very big in the Walker companies); Richard Braddock, once president of Citicorp and now chairman of Priceline; Marshall Loeb, former managing editor of FORTUNE; New York's Allen & Co. (no relation to Paul); and Nancy Peretsman, one of Allen & Co.'s key investment bankers. Paul Allen and Liberty Media had put money in both Priceline and WebHouse. So the money they paid for Priceline stock in August went into helping them not lose money in WebHouse--except that the effort didn't work.

As the re-upping of investors suggests, and as entrepreneurial success so often requires, Walker is a supersalesman, able through his ideas, his mastery of words, and the strength of his beliefs to command a loyal following. At Allen & Co.'s Sun Valley bash for business leaders in July, he gave a presentation that ranged over 200 years of American innovation and left many listeners powerfully impressed by his intellect and style (although he came across to at least one listener as maybe too smooth, too close to a "promoter").

But he is a believer--nobody ever said he wasn't that--and in July he had plenty of good news to think about. Priceline had just finished a quarter in which it realized only a small loss and seemed on track soon to move into the black. Synapse had just secured the Time Inc. investment and, with this clout behind it, was preparing for an IPO. The registration statement Synapse filed with the SEC later in July visualized a market value of at least $425 million.

WebHouse in July was a mixture of good news and bad, but to an entrepreneur with faith, the balance surely would have seemed positive. The good news was that swarms of Internet households were using WebHouse to make discount bids for groceries and gasoline. The bad news was that Walker had not lined up a base of manufacturers that, in return for incremental sales, would absorb the customers' discounts. This flaw left WebHouse itself paying the bill. In effect, it was selling customers dollar bills for 90 cents.

Still, the construct in Walker's head was saying that if he just hung in and maybe redesigned his system a little, he would win over the manufacturers and get past the losses. All he needed was the time that more capital would give him. As he spoke at Sun Valley, he had $50 million in hand (the amount Time Inc. had paid him for Synapse stock), and he was negotiating with Liberty Media and Vulcan for more.

That the capital burden was falling on Walker and WebHouse, and not on Priceline, was the result of an imperial decision made in the fall of 1999. Though Priceline had built its growth on airline tickets, the company had always known it would have to broaden its line of merchandise to be really successful. Besides expanding on its own, Priceline figured it could leverage its Website by doing joint ventures and alliances with suppliers of goods and services, and by licensing its patents to affiliates. It may seem odd--though why should it in this exotic establishment?--that the first affiliate should be one set up by Walker himself. But that was the decision: He would personally form a new company to sell groceries.

Braddock says Priceline was busy then with other projects, especially technology upgrades, and gave almost no thought to selling groceries itself. Priceline, however, did keep a corporate opportunity if Walker's venture went well. It had, and still has, a warrant to buy 77% of WebHouse for $413 million, a figure that implies Walker thought the value of the business could grow to above $500 million.

Priceline WebHouse Club put its grocery business on Priceline's Website last November and later added gasoline. From the start, it was irrelevant that the products had the WebHouse label attached to them; consumers simply thought they were dealing with Priceline. This was highly significant. Priceline may not have taken on the business risk of selling groceries and gasoline, but it had put its brand name on the door. And as any business person who has made an alliance knows, that was a business risk as well.

That isn't how they saw it at Priceline--not for several months anyway. Braddock says Priceline executives gave WebHouse "virtually zero" surveillance, letting Walker run it. At Priceline board meetings, says one director, Walker's report about the new venture would be a kind of tag along, coming up last on the agenda. Even then, what the directors had their ears mainly tuned to were the glowing reports of how many customers--eventually three million--were using WebHouse. So there were big losses. Who cared? Wasn't that how Internet companies worked as they highballed down the road to prosperity?

Ultimately, though, it was two Priceline directors, neither of them investors in WebHouse, who raised the stop sign: Nancy Peretsman of Allen & Co. and William Ford, General Atlantic's representative on the board and a powerful force in the Walker empire. A Priceline insider says Walker considered those two people his "eyes on the capital markets." What they saw in late September was not only the large amount that Walker and his backers had sunk into WebHouse--$390 million in less than a year--but also the certainty that WebHouse would need more. Walker was figuring he'd raise that by selling additional Priceline stock. But the stock was down to about $15, and new sales at that low price--even if Walker could have made them--would have seemed embarrassing. A company doesn't want its founder selling stock at fire-sale prices. So Peretsman and Ford persuaded Walker to draw the curtain on WebHouse--and to do it very fast, while the company still had enough money to wind down its affairs.

The fact is that Walker also needed to move decisively to save the Priceline brand. This was not something he publicly acknowledged. He said instead that WebHouse faced the possibility of running out of cash, and that it had to stop doing business while it still had enough money to pay the parties to whom it had made legal or moral commitments. Chief among them were customers who had contracted to buy discounted groceries and gasoline and hadn't yet taken delivery. Walker knew the cost would be heavy.

Here's why. Imagine a customer (we know one like her) who had paid WebHouse, by way of her credit card, for ten gallons of gas at $1.67 a gallon but hadn't yet gone to the gas station to complete the purchase. After Walker shut WebHouse, this customer's card was credited for $1.82 per gallon, which WebHouse figured would cover her cost when she next bought ten gallons of gas the standard way. That 15-cent difference is roughly a 9% toll, and it suggests where WebHouse stood on all the undelivered sales it had made. Apply 9% to millions of gallons of gas and tons of much costlier groceries, and you are into big money.

If this gasoline customer had not received the savings she'd been promised, she wouldn't have blamed WebHouse--she barely knew it existed. She would have blamed Priceline, the name she knew. So while there is no question that the mere shutting of WebHouse tainted Priceline's reputation (in part because groceries and gasoline had suddenly been rubbed out as routes for expansion), you can make a case that if Walker had not made an exit while he still had cash for his promises, the Priceline brand would have been not just maimed but entirely destroyed.

We can count the cash cost of these problems, too, over the 11 short months of WebHouse's existence. If the $390 million put into the operation is entirely used up before WebHouse closes, the rate of spending will have been about $35 million a month--a burn rate conjuring up the Great Chicago Fire. Walker himself put up almost half the $390 million, including around $135 million he contributed after selling Liberty, Vulcan, and the prince their $240 million of stock. He says he is prepared to put another $15 million of his take into WebHouse to wind down its operations, and he also used $40 million to pay off a loan that he took out earlier this year to buy WebHouse stock. He implies that the rest of his proceeds, $50 million, will be used to pay taxes on the gains he realized from his Priceline sales.

To the Chicago Fire at WebHouse, add the conflagration at Priceline, whose stock was recently selling between $5 and $6 (down from a peak of $162 in April 1999). That's tough wreckage for insiders to behold. Still, the positives right now about Priceline include a management that's highly experienced (and getting more so every minute): Braddock; CEO Daniel Schulman, an ex-AT&T executive; and CFO and chief strategist Heidi Miller, who formerly was CFO at Citigroup. Miller joined Priceline only last March, after getting an infusion of Internet enthusiasm when she sat next to Walker at a dinner party.

The positives also include millions of customers--Priceline puts the number at 6.8 million--and a line of products that, besides airline tickets, includes hotel rooms, rental cars, new cars, long-distance minutes, and financial products such as mortgages. The company is also respectably stocked with cash and equivalents: $139 million at the end of the second quarter. But it has been depleting cash steadily; the amount spent in the second quarter was $11 million. CEO Schulman has already announced that a dismaying September, filled with bad news about airline-ticket sales, will make the third quarter a loser. (Actual results for the quarter will be reported Nov. 2, after this article has gone to press.) The fourth quarter, and others to come, are question marks, newly uncertain as regards ticket sales and tarred at least temporarily by the WebHouse debacle.

A Priceline insider notes that the company has never worried for a minute about running a tight ship, and that battening down the hatches could create significant savings. That's no doubt true, but the company also has new expenses to worry about. At least 15 law firms have filed class action suits against it for displaying optimism in its second-quarter earnings report, on July 24, and then careening into a period of unrelievedly bad news. The plaintiffs may have trouble proving Priceline knew before September that things were going sour, since it was only then that its airline-ticket business hit an air pocket. Nonetheless, when did logic ever hold the plaintiffs bar back?

In the meantime, Priceline continues to battle two parties that have challenged its patents (a California lawsuit is set for a hearing in early November). However, the main burden of paying for those fights is borne by the original patent holder, Walker Digital. Priceline does have to absorb the costs of still another suit, in which it is charging Microsoft's Expedia with patent infringement.

A successful challenge to its patents would cripple Priceline. But at the moment its toughest threat may be a new competitor, Hotwire, an Internet seller formed by the venture capital firm Texas Pacific Group and invested in by six big airlines: America West, American, Continental, Northwest, United, and US Airways. The presence of the airlines supports what many skeptics have prophesied about the Internet: Prove that a business works--as Priceline has provisionally done in getting consumers to bid for airline tickets--then brace yourself for supplier competition.

Though both Hotwire and Priceline merchandise the airlines' "distressed inventory"--seats primed to take off unfilled--Hotwire's approach is different from Priceline's. At Priceline a customer basically states the price he will pay for a given trip and rummages for an airline that will accept his terms. At Hotwire, the customer doesn't initiate the fare. Instead, he specifies his trip and receives the best price offered by the six airlines (a nucleus to which other carriers are likely to be added). The customer has 30 minutes to accept. Compared with Priceline's process, which can be tedious, Hotwire's is fast and easy.

However, customers will not necessarily get a lower price at Hotwire than at Priceline. In fact, consumers who get a big kick out of saving $20 on an airline ticket--and that pretty well defines the customers of both companies--may well try to game the system. They could conceivably do so by rushing from Hotwire to Priceline (or to Expedia or any other travel agent) to line up a better price before Hotwire's 30-minute window shuts.

Hotwire, based in San Francisco, formally started its service at the end of October. Braddock points out that it is just the latest of several well-publicized competitors that have drawn a bead on Priceline without knocking it from the leader's spot. But clearly no one at Priceline is taking Hotwire lightly. In the periodic visits that he makes to airlines, CEO Schulman has recently sought assurances that they will continue to send distressed inventory in the direction of Connecticut.

Could Priceline, distressed inventory itself, be a candidate for a merger? Insiders do not hoot at the question. But Priceline would be looking for a price significantly above the current level, and who might pay that isn't clear. The likely lure for a buyer would be Priceline's customer base. But it's worth remembering that Priceline has not yet proved that millions of customer relationships translate into profits.

Here's news from another part of Walker's empire: Synapse has delayed its plans to go public. The company's CEO, Michael Loeb (son of Marshall, the Walker backer and a Priceline director), says the market for IPOs isn't good enough now to support an offering, and besides, Synapse has no pressing need for funds. But you know it couldn't have helped, given Priceline's troubles and WebHouse's demise, that Walker's name is associated with Synapse.

Walker is an indefatigable entrepreneur, and it will take more than a few months of bad news to knock him down for the count. But today, in a balance-sheet sense, he's a current liability.

REPORTER ASSOCIATE Matthew Boyle

FEEDBACK: cloomis@fortunemail.com