(FORTUNE Magazine) – On the afternoon of Monday, Oct. 19, most of Goldman Sachs' 189 partners are to gather at the firm's New York headquarters for a hallowed ritual--the biennial awarding of that ultimate badge of Wall Street achievement: a Goldman Sachs partnership. What will make this year's meeting perhaps the oddest in the firm's 129-year history is this: The meeting was never supposed to happen. As of two months ago, no one was ever to make partner at Goldman Sachs again--because, as you'll remember, Goldman was supposed to go public right about now, meaning that the 50 to 60 men and women who are about to become members of this exclusive club would instead merely have become highly paid employees of another new corporation.

But of course Goldman's IPO, okayed by the partners in June, was scuttled, officially "withdrawn," in late September. Instead of a hallmark of the roaring '90s, which some had predicted it would be, the IPO became a high-profile victim of the financial firestorm that has swept across the globe. Goldman put off the deal because once the stock market crashed (shares of the big Wall Street houses crashed even harder), the numbers didn't work anymore. The deal would have valued Goldman at $28 billion (a number cited in a letter to the firm from Goldman limited partner and former co-CEO John C. Whitehead, which FORTUNE obtained) if, as once envisioned, Goldman sold stock at around four times book value. Post-September, Goldman stock might have brought a slight premium to book, valuing the firm at a mingy $7 billion, which would have left it far short of the swag the IPO was supposed to spread throughout the firm to its partners, limited partners (mostly retired partners with capital still in the firm), and other employees.

According to the firm's co-CEOs, Jon Corzine (rhymes with "sign") and Henry "Hank" Paulson, putting the IPO off "was a pragmatic decision." Says the bearded Corzine, with his trademark Cheshire Cat grin: "At the appropriate time we will revisit the matter. Now we have to move forward." But contrary to the blase nature of that statement, it is clear, based on interviews with Goldman partners, employees, customers, and Wall Street sources, that debating--and then icing--the IPO was a wrenching experience that has bruised the firm. It has spotlighted, and in some cases revealed, weaknesses both in Goldman's capital structure and in its mix of businesses. It created tensions between the firm's general and limited partners. And it heightened rumors already circulating on Wall Street about the rivalry between Goldman's investment bankers and traders in general, and Corzine and Paulson in particular.

One of Goldman's official reasons for going public was to "match our capital structure to our mission." That may sound like boilerplate, and it is, in part. Whitehead, for one, wrote, "I don't find anyone who denies that the decision of many of the partners, particularly the younger men, was based more on the dazzling amounts to be deposited in their capital accounts than on what they felt would be good for the future of Goldman Sachs." Still, capital is a valid issue. Goldman, in fact, has a smaller equity base than either Morgan Stanley or Merrill Lynch ($6.6 billion at midyear vs. $13.8 billion and $11.7 billion, respectively). Yet both current co-CEO Paulson and ex-CEO Whitehead say that having less capital merely forces the firm to make better decisions. Competitors don't think Goldman suffers from a lack of capital. Says Donaldson Lufkin & Jenrette CEO Joe Roby: "I wish I could find a business where Goldman is capital constrained."

In fact, the problem is not so much the amount of capital Goldman has as it is the structure and stability of that capital. When partners "go limited" they can withdraw capital; if large numbers of partners go limited, large amounts of capital can go with them. That's what made 1994 so scary for the firm. Turmoil in the bond markets depressed Goldman's profits, and partners fled in droves. (It may not be a coincidence that Corzine, a chief proponent of the IPO, became CEO in 1994.) The restrictions are more onerous now. When partners go limited, they can take only a fraction of their capital and can withdraw the rest over a minimum of five years. But capital can still flee.

Another problem is the cost of Goldman's capital. Two outside institutions, Japan's Sumitomo Bank and Hawaii's Bishop Estate, have invested substantial amounts in Goldman--$1.5 billion as of mid-1998, 23% of total capital. Because a significant chunk of Goldman's pretax profit goes to these outsiders, it is "harder to grow capital via retained earnings, particularly in rough markets," says Robert Upton, a vice president at Duff & Phelps Credit Rating. Equity raised in a public offering at a multiple of book value would have been a cheaper way for Goldman to build its equity base.

Even the minimal financial disclosure that Goldman made in preparation for its IPO revealed that its mix of businesses is less than ideal. Some of it, though, showed that Goldman is even more formidable that anyone had guessed. In the Street's most lucrative business, mergers and acquisitions, Goldman consistently out-earns its major rivals. Last year Morgan, Merrill, and Goldman did roughly the same dollar volume of deals, but Goldman earned 50% more than Merrill and 30% more than Morgan. But what's worrisome is that Goldman has a more volatile revenue mix than its peers because it relies more on its trading operations. In the first two quarters of 1998, Goldman got 43% of its revenues from trading, vs. 23% for Merrill, and around 28% for Morgan. Also, at midyear, Goldman's total assets were over 36 times the firm's equity, indicating higher leverage than its peers. One major customer likens the firm to a steel mill: "At 100% capacity, they're making tons of money; at 60% capacity, they're losing tons of money." Such facts have led some Wall Streeters to compare Goldman to a hedge fund--not a complimentary moniker in these times. For that reason, some think it's likely Goldman would have had to price its stock at a discount to Morgan and Merrill had the IPO proceeded, and that its stock would have continued to trade at a discount. "Goldman would have traded more like Lehman," sniffs one (perhaps overly harsh) competitor.

That reliance on trading was an issue in 1994, and it is an issue today. Although Goldman's preliminary prospectus takes great pains to point out how much the firm's risk management has improved since 1994, it may not have been improved enough to handle the shock waves that have reverberated through the markets over the past few months. Evidently few investment banks had planned for such a worst-case scenario.

Goldman, along with the rest of Wall Street, is reflecting some of that worst-case scenario in its recent performance numbers, and there's no doubt that the firm's reliance on trading is playing a huge part in the pendulum-like swing of Goldman's profits. Until this summer, Goldman's numbers had been nothing short of phenomenal. Net revenues climbed from $4.5 billion in 1995 to $7.4 billion last year. Pretax profits (because Goldman is a partnership, taxes are paid by the partners) climbed from $1.4 billion to $3 billion. For the first two quarters of this year, revenues climbed 50%, and the firm was on pace to exceed $4 billion in earnings. But not anymore. In the quarter ended Aug. 28, pretax earnings dropped 27% from the previous quarter. September was even worse; there's speculation the firm perhaps lost as much as $900 million. True? "Absolutely not," says Paulson.

Of course, Corzine and Paulson don't have to fess up at all, because Goldman is still private, and it has ways to obscure the magnitude of any losses. Some limited partners saw this as a powerful reason for staying private. "Are you ready to lose the flexibility we now have in reporting up and down earnings?" asked Whitehead and former CEO John Weinberg in another letter to Corzine and Paulson. In any event, Goldman's 1998 earnings will fall far short of the $4.5 billion some were expecting as recently as this summer.

Another weakness in Goldman's mix is its asset-management business, which provides a thin cushion for riding out market vicissitudes. While asset management is growing rapidly, it is a smaller and less profitable business at Goldman than at its rivals. At the end of 1997, Goldman had about one-third the assets under management that Merrill Lynch did; in 1997, Goldman's $458 million in asset-management revenues were not much more than Morgan Stanley's $400 million or so in asset-management net income.

Asset management may be less sexy and less lucrative than swashbuckling multibillion-dollar investment banking deals, but it generates fee-based income that is in theory steadier than trading profits. Goldman was a late entrant to the asset-management game, charged low fees to get business, and suffered from both a lack of direction and personnel turnover. At the end of 1995, Goldman had just $52 billion under management, 40% of which was parked in low-fee money market funds. Asset management at Goldman was "the ugly stepchild," says Andrew Guillette, a consultant at Cerulli Associates.

That perception is changing, both inside and outside the firm. In the past 2 1/2 years, assets have more than tripled, to $165 billion, only 20% of which is in money market funds. More important, the firm's attitude has changed. Goldman's asset management arm employs over 1,000 people today, vs. around 250 a few years ago. Not only has Goldman brought in outside honchos with the promise that they'll become partners (which is rarely done at Goldman), but stars from other areas within Goldman have moved to asset management as well. "Goldman is a powerhouse in the making," according to Financial Research Corp. Still, outsiders criticize Goldman for being better at sales and marketing than investing, and Goldman has a long way to go to achieve powerhouse status. Without a publicly traded stock to use as currency for acquisitions, the road may be tougher.

The move to go public exacerbated, and brought out into the open, divisions within the firm, and it dragged Goldman's business through the press as never before. One point of friction was between the firm's general partners--really its controlling owners--and the limited partners. Although Paulson and Corzine like to use the word "overwhelming" when describing the support of Goldmanites for the IPO, perhaps that's an overstatement. The process left some of Goldman's current partners and 108 limited partners unconvinced, angry, or both.

From the limited partners came the most dissension, stemming from a deeply held belief in the sanctity of the partnership and ire over the terms of the deal. Initially, they were to have received a mere 25% premium over the book value of their equity, while the general partners were expecting to get premiums approaching 300%. "If these figures are accurate," Whitehead wrote in June, "I feel rather shocked and offended that any of you would have felt that that was fair." He warned "you will have a major problem on your hands if you stick to your present plans." The offering, which fortune obtained, was restructured to give the limited partners several choices as to how their capital would be treated, one of which was that they would get a 55% premium to their current equity in stock. Still, had the IPO happened at a 300% premium, current partners would have collected the lion's share. Ironically, what helped scuttle the deal is that when the market fell, that 300% premium dwindled so far that the limiteds might have made out about the same as--or even better than--the generals.

Even when the deal was on, there was an abundance of rumors surrounding Paulson and Corzine. The word was that Paulson, who was chief operating officer at year-end 1997, had played his cards masterfully, saying "no" to an IPO but finally trading a "yes" (and the "yes" of a key cadre of investment bankers) for the co-CEO position. There's even gossip that Corzine, the cheerleader for the IPO, is under fire for its failure.

It's clear that Paulson is far more powerful than he was a year ago. But it seems misguided to blame Corzine, and it is doubtful in market conditions this miserable that anyone has the time to think about anything but making money. "We're closing ranks," says one general partner.

Goldman's attempted IPO also brought to the fore issues of "fairness" (a word Corzine and Paulson used repeatedly) and compensation in the firm's middle ranks. Goldman doesn't necessarily pay top-dollar salaries to its all-important junior executives--the ones who do the grunt work--holding out instead the brass ring of partnership and its potential for eight-figure incomes. With the IPO, Corzine in particular pledged to push greater rewards down the corporate ladder. But with the offering shelved, Corzine and other senior partners must now rejustify the partnership system. That helps explain why Goldman is making over a dozen more partners this year than in 1996. Even so, hundreds more midlevel employees may be wishing that Goldman had somehow done the IPO and asking themselves if Goldman is still the only place they ever want to work. "We're getting calls [asking about jobs] from Goldman people that we never used to get before," says a senior Morgan Stanley banker.

Of course, there is more than a little Schadenfreude there. Goldman Sachs is the firm the other guys love to hate, in large part because Goldman is so damn good. So why is this firm different from any other Wall Street firm? Much of what sets Goldman apart is a concept often talked about on Wall Street, yet that few firms actually possess, and that is culture. From the moment new hires walk into 85 Broad Street at the bottom of Manhattan, it is impressed upon them how superior Goldman is and how inferior is the competition. They are warned that they must never disgrace the firm or appear in the press. And they work 14-, 16-, 18-hour days, ostensibly for that elusive prize of a partnership that is only slightly more achievable than an all-state high school basketball player making the NBA.

The big questions remain. The first: Did Goldman miss its golden opportunity by not getting its IPO done, or did it have a lucky escape? Probably the latter. Had it sold shares last spring, its stock would now sell substantially below the offering price--a humiliating beginning for the Street's premier firm. In fact, many partners feel the firm dodged a bullet.

The other question: Will Goldman go public anytime in the near future? Corzine and Paulson still hew to the party line. But plenty of others, from longtime Wall Streeters to current and former partners, think it unlikely. Not only is there a whole new group of partners--some 25% will be new after the Monday meeting--who would need to be cajoled, but after this IPO's failure those who voted "yes" last time may reconsider. And don't forget that it was the rich multiple that securities firms were selling for that made Goldman's IPO possible. In the wake of the latest round of excessive speculation on Wall Street, it may take a fresh generation of investors to once again believe that these stocks should sell for four times book value.