(FORTUNE Magazine) – From the moment when Travelers Group and Citicorp announced last spring that they would merge to form Citigroup and do their thing under co-CEOs, the buzz said that the chairman from Citi, John S. Reed, 59, would leave in relatively short order and that his Travelers companion, Sanford I. Weill, 65, would stay on, and on, and on. That scenario may still play out: Reed made it plain in a late November interview with FORTUNE that he doesn't want to be a long-termer. But he will not leave soon, because Weill won't let him. "There is no way," Weill says, mentally twirling a lasso, "that John's getting out of here, no way he's leaving me alone."

He doesn't say "alone with this mess," but the thought hangs in the air. In nine months, this merger--$73 billion in size and second only to Exxon-Mobil in the megadeal year just ended--has proved for the zillionth time that big corporate marriages can be exhaustingly difficult to pull off. This strange honeymoon has also shown, though, that one part of a business union can go desperately bad while others progress with the speed and elan laid out in the original plan.

That is the situation today at Citigroup: of its two major parts, one--the global corporate business--is a case study in trouble. The other, the global consumer business, is moving ahead under two executives who get along--yes, mergers are always about people--and have a clear sense of direction. They are building Citi's brand and beginning to cross-sell.

But there is drama developing in that business also (see box): The tribe from Travelers, long used to a lean, efficient way of operating, is determinedly imposing its style on a Citi culture that has traditionally been bored by the details of day-to-day managing. That could bring on interesting clashes. But the efficiency drive is otherwise upbeat news for profits. If world markets are relatively calm next year--though that is a huge "if" for this company--the expense cuts under way should by themselves allow Citigroup to report good earnings in 1999.

That will still leave the world remembering all that went wrong in 1998. Check out the stock price, as, in fact, Sandy Weill does maybe 100 times a day. At the beginning the panache of this deal pushed the stocks of the two companies to a combined market value of $165 billion. In mid-December, though, Citigroup--officially born on Oct. 8--had a market value of only $108 billion.

A market revolt against financial stocks helped that happen, but the clincher was shocks at Citi. One was earnings. Not counting a restructuring charge that was always in the plan, profits for 1998 were originally expected to rise and to be nicely in keeping with the growth records that both companies had brought to the party. Instead, the good gains of the first half were erased by train wrecks in the third quarter: Both Travelers and Citicorp suffered large losses when Russia repudiated its debt, and Travelers got knocked around still more by huge fixed-income trading losses at Salomon Smith Barney. Leaving aside the restructuring charge--around $1.3 billion before taxes--Citigroup's earnings for the year are likely to be down by more than 15%.

Except for the special charge, these things would probably have happened to Travelers and Citi had they not joined, right? Fair enough, but other problems are directly attributable to the merger--most especially to Citigroup's built-in institutional fondness for divided authority, which works when it works but requires exceedingly careful casting.

It was this management reflex that contributed to the most shocking event in the new company's short life, the firing of James Dimon, 42. Dimon, brainy, confident, and sometimes arrogant, was both Travelers' president before the merger and Sandy Weill's joined-at-the-hip comrade in founding and building the company; he is also a talent whom the intellectual Reed immediately liked. Admired by Wall Street analysts for his candor and mastery of figures and facts, Dimon was named president of Citigroup in May. He was the heir apparent.

Consequently his ejection rocked just about everybody who followed the company and also bombed Citigroup's stock. Five years back, a financial executive named Christopher Steffen was called "the $2 billion man" because his arrival at Eastman Kodak added that much to its market value, and his rapid exit erased that much. A yawn for Chris Steffen. Jamie Dimon's sudden departure from Citigroup on Nov. 1 caused its market value to drop by $11 billion in two weeks, a decline of more than 11% (see chart).

By the end of November, the stock had crawled back from this dive. But it is not clear that the management confusion in the corporate business--one reason, though only one, for Dimon's ouster--was cured in that month or will be in many another. This business is complex, taking in Citibank's commercial banking and trading and the entirety of Salomon Smith Barney. Unless Citi's commercial bankers and Salomon's investment bankers begin to work together to produce incremental revenues, the full potential of this merger cannot be realized.

On paper the partnership looks good for all, because it promises each side access to new customers. Citi has relationships with thousands of companies around the world, whereas Salomon deals with relatively few. Managing to connect with more customers should help Salomon sell its underwriting and M&A services--products that Citi hasn't carried in its line. Citi deals instead in such big-banking items as loans and cash-management services, and typically sells these to mid-level executives, mainly treasurers and assistant treasurers. Adding Salomon's investment banking products to the mix should allow the Citibankers to deal more often with the executive suite--that is, with the CFOs and CEOs who normally contract for these products--and nicely broaden their relationship with the customer. Even if a partnership sounds sweet, it's nevertheless a challenge to organize. Each side has relationships that it thinks of itself as "owning." So how does the other side fit in? Who gets cut if there are overlaps? How do you handle the sticky fact that investment bankers normally think themselves superior to commercial bankers--and have only to look at their paychecks to reassure themselves that they are? Indeed, how do you reconcile the broad differences in pay between the two?

Jamie Dimon was a part of the management team that worked for seven months, from April through October, at answering these questions. But notice those words "a part." The corporate business didn't have just two bosses--which is de rigueur at Citigroup--it had three: one of them Dimon, the others Deryck Maughan, 51, head of the old Salomon Brothers, and Victor Menezes, 49, of Citicorp. This tri-headed monster was terminally dysfunctional. Maughan himself says wryly, "I don't believe that three co-heads reporting to two co-heads is an ideal situation. Five people in a room is a lot of people."

So Weill and Reed have now killed off one head, Dimon; moved Maughan to a different spot on the organization chart; and from the executive barn trotted in a new two-headed creature to run the corporate business--one head belonging to Menezes, the other to Michael Carpenter of Travelers. Menezes, a native of India, starred in a variety of Citi international jobs over the years and then, until last May, served as the company's chief financial officer. Carpenter ran Travelers Life & Annuity before the merger and did an excellent job. But in another sense he has been brought back from the dead. He was head of Kidder Peabody in 1994, when it was flattened by bond losses that came out of Joseph Jett's trading. To have Carpenter back running an operation that has just itself been flattened by losses is irony of a high sort.

Carpenter and Menezes may--repeat, may--bring order to the corporate business. But, at the least, the upheaval there has delayed the day when the merger, in Reed's words, can be judged as "effectively in place and motoring along." Reed has promised Weill he will stay until that moment is reached, which Reed says should not be "forever" in the future. "We are either going to get it done or not...." And from his seat alongside, Weill mutters, "But it would be a miracle if it's accomplished in one year."

Dimon's fall from grace, which has Shakespearean overtones, is indubitably connected to the problems in Citigroup's corporate business but rooted as well in pre-merger events. The fact is, the close relationship that formed between Dimon and Weill in 1982, when Weill was still at American Express and Dimon was a newly minted MBA hired as his assistant, began to sour several years ago, though this was not widely known. Maybe, as some people think, Dimon started getting too much press, particularly in the mid-1990s when he added the running of Smith Barney to his duties as president. Weill, in any case, started making big decisions without consulting Dimon, his supposed confidant. In one move, Weill asked Frank Newman to become a vice chairman of Travelers; instead Newman became CEO of Bankers Trust. In another, Weill hired TIAA-CREF's Thomas W. Jones in August 1997 to run Travelers' asset management business, which Weill pulled out of Smith Barney and made a separate division reporting directly to him.

Well before that, however, Dimon had run into the famous nepotistic problem of Weill's daughter, Jessica Weill Bibliowicz, now 39, who headed Smith Barney's mutual funds. Jessica, smart and well liked, wanted a bigger job--head of asset management--but Dimon said no, because he thought she wasn't ready for the responsibility. Jessica also wanted a seat on Smith Barney's executive committee, and he didn't give her that either. A friend of both Dimon's and Weill's remembers saying to the younger man, "You're crazy not to have done this." Dimon didn't, so he has said, because he regarded Jessica as junior to others who themselves hadn't been named to the committee.

Then, in June 1997, Jessica got an offer from John A. Levin & Co., a New York investment advisory firm, to become its president, and she took it. There is no question that her father hated her departure from Travelers and that he blamed Dimon. If Shakespeare were handling this plot, we might now have a poetic way to say that nepotism seldom is beneficial in large, publicly owned corporations. Lacking verse, we will submit what Hollywood producer Jack Warner is supposed to have said to Albert Einstein: "I have a theory of relatives too. I don't hire them."

Amid the mounting dissent between Weill and Dimon, Travelers bought Salomon Inc. in September 1997 for $9 billion after whirlwind negotiations between Weill and Deryck Maughan, who besides being business friends served together on the Carnegie Hall board. The deal fed Weill's long-standing appetite for acquisitions. But Weill also saw Salomon's international operations as giving Travelers a global reach that he thought essential. Unfortunately, the deal had barely closed when Asia began falling apart, which left Weill realizing he'd strayed into an unfamiliar "scary world." It might have helped if Salomon's international operations had been totally solid. But as Weill quickly and rather belatedly discovered, they were weak--"hemorrhaging," says one executive from Travelers. That left Weill realizing that he might draw embarrassing criticism for having made this big merger.

Meanwhile, who was in charge of the new operation called Salomon Smith Barney? Not the going-in CEO, Dimon, but rather Dimon and Maughan, as co-CEOs. This sharing of power, mandated by Weill, greatly rankled Dimon and got his relationship with Maughan off to a bad start that never really improved. This writer talked to Dimon in late 1997 and speculated that the co-CEO arrangement wouldn't survive: "I'm guessing that things will be different a year from now, though I don't know how they will be different." Answered Dimon soberly: "Yes, they will be different." Neither party to this conversation had a clue as to just how extraordinary the changes would be.

And of course the most extraordinary development was the agreement by Travelers in April to merge with Citi. For Dimon, the merger delivered a body blow in that Weill and Reed decided that only the two of them, and no other insiders, would be on Citigroup's board. This decision de-robed important executives on both sides of the fence, but surely none despised his loss more than Dimon, who thought his exclusion terribly unjust. "My God," he said to a friend, "I helped build Travelers." Salve was then applied: Weill and Dimon's new pal, Reed, made him president of Citigroup. But it was largely an empty title, since the only executive set to report to Dimon was the company's new chief financial officer, Heidi Miller, a Travelers colleague.

In the meantime Dimon also took his place as one-third of the executive beast charged with integrating the corporate businesses of Travelers and Citi. Without a doubt, it was a strange scene that this awkward body lumbered into. The integration of Smith Barney and Salomon was only partly complete and was in fact proceeding with all sorts of bumps and grinds. A big part of what Salomon had brought in the door was its well-known bond-arbitrage business, which had a history of volatile profits. Weill, shelving his usual risk-averse thinking, initially pronounced himself willing to accept the lurches, maintaining that the business' generally juicy returns would make its occasional bad results tolerable.

A sidelight to the story spinning out in 1998 concerns Weill's son, Marc, 42, who was chief investment officer of Travelers before the merger and who now holds the same title at Citigroup. Last spring Sandy Weill proposed that Marc add the supervision of the bond-arb department to his other duties. Dimon resisted the idea, and the boss of the department flatly said no. Talk about luck: Had Marc Weill made this move, he would have been overseeing the part of Salomon Smith Barney that was about to crater.

The trouble was preceded by the firm's statement in early July that the U.S. arb business was running losses and turning riskier--and would be closed down. That meant Weill's risk-averse nature was reasserting itself. But no one then focused on Salomon's lending and derivatives exposure to a hedge-fund customer named Long-Term Capital--"we simply missed it," Dimon said in October--or recognized that the firm had made fixed-income bets that all but replicated LTC's and would go just as bad if interest rates diverged from their normal patterns. When Russia devalued the ruble in August and repudiated its debt, and when a flight to quality ensued, a divergence occurred of exactly that type, and of an extreme character. In the third quarter, Salomon Smith Barney suffered pretax losses of $1.3 billion on principal transactions, with most of the damage attributable to the bond-arb desk and to trades made with hedge funds that had taken a Russian fling.

Wall Street's big banks and brokers went on then to bail out Long-Term Capital, with Salomon anteing up $300 million. Citi, however, was salubriously among the missing, because it had virtually no exposure to LTC. Hold the praise, though. Citi has a vaunted program called Windows on Risk, which is supposed to flash warnings. But Windows didn't provide a clear view of Russia, and in the third quarter Citi swallowed pretax losses there of $384 million. Here's language you've heard before: "We simply missed it," says Reed. He adds that one of Citicorp's board members, John Deutch, a former director of the CIA, later drawled out, "Stoo-pid!" But Reed, who habitually spills out whatever is on his mind, notes that Deutch didn't warn him in advance.

While these disasters were reeling out (and chewing up executive time), the three-headed management team was coming to grips with the questions of how the corporate businesses of Salomon and Citi should be integrated. Some decisions were easy. Salomon had the powerhouse fixed-income business, so fold in Citi's; do the opposite in foreign exchange, where Citi was dominant. Then the issues got harder: How to address, for example, a big overlap in derivatives and simultaneously decide whose employees should be let go?

"Once the knife came out--once the scalpel of efficiency appeared--I think we started to slow down," Maughan says. "It was taking so long to go through the process--the five of us--particularly when it came to the cutting of expenses."

Nonetheless, Maughan does not think--nor does FORTUNE, from its broader reporting--that the slowed pace of decision-making was the root cause of the radical management changes that soon took place in the corporate business. The problem was the lack of an undisputed boss or, in this land of "co's," even two bosses. By early fall the tri-heads had themselves grown so frustrated with their ineffectiveness that they went to Reed and Weill and pleaded for their number to be reduced to one or two. But the co-CEOs gagged on the decision, possibly because they knew that neither Dimon nor Maughan would work for the other. Reed and Weill shuffled some duties but otherwise let the triumvirate live.

In the end it took a large October management meeting at the Greenbrier resort in West Virginia to get things off the dime. The 140 Citigroup executives there broke into small groups to come up with recommendations for moving the merger along, and each group emerged with a version of this plea: "Do something about the management of the corporate business." Within a week--it happened on a Sunday--Reed and Weill called Maughan in and asked him to take the title of vice chairman of Citigroup and apply his well-known marketing skills to the winning of corporate customers. Then they summoned Dimon and asked for his resignation. He tendered it, so this will not go down in the books as a firing. But it was a firing.

Dimon has told friends at Salomon Smith Barney that he was stunned, mainly because he had never believed that any kind of problems--with Weill or with management structure or whatever--could outweigh his value to the company. He has also said that Reed once questioned whether Dimon was really doing as much as he could to smooth an obviously difficult management situation. In his derision for the structure in place, Dimon probably wasn't. The charge is nevertheless funny, because Reed himself earned an enduring reputation for being bluntly undiplomatic during the years when he was building Citi's consumer business and laying the base for his becoming CEO.

It would be priceless to know what Weill and Reed said to each other about Dimon in that week between Greenbrier and in-the-soup Sunday, but they are not recounting that conversation. In November, though, they did tell FORTUNE a few things that slightly pierced the fog. Said Reed, chuckling morbidly about needing to ax the "crazy triumvirate" and come up with a new management arrangement: "Jamie got himself positioned so he couldn't do it." How so? "Internally. The way he was behaving; the way he was acting. But we don't want to get into this."

Then Weill chimed in: "He needs to learn how to reach out to a broader range of people. It was just hard to have that happen." Back to Reed: "If I had known the first day of the merger what I know now"--he did not expand on that--"I think we could have used Jamie differently, and in so doing maybe not run into the problem. But you know what? Mergers are hard. In a merger, the skills of getting along with people are much more important than just your personal professional skills."

Reed made it clear that the problem was certainly not a breakdown in the rapport between him and Dimon. He said he'd just told his wife, Cindy, that he was going to ask Dimon to have dinner. "I like Jamie," he says. "He's an impressive person, and I really feel sorry...." But not sorry enough, obviously, because he concurred in the firing. It wouldn't have happened had not he and Weill agreed.

With the triumvirate a historical artifact and with the management of the corporate business breathtakingly simplified--whew, just two co-heads reporting to two co-heads--Carpenter and Menezes have moved in on their priorities. The first of these, says Carpenter, who took the lead in FORTUNE's interview with the pair, is to get the risk levels of the business under control so that "we don't have a repeat of what happened in the third quarter." Just where do the risk levels stand? That seems to be a murky subject, even to Sandy Weill. Asked in November about the bond-arb positions, Weill said they'd been "$80 billion to $100 billion" at the beginning of 1998 and that he hoped soon to see them at one-quarter of that. At that date in November, he said, they were maybe $40 billion. But then he had second thoughts about the figure. "Check that out," he told his public relations director: "Maybe they'll tell you." The PR director came back a month later with word that no figure would be supplied or confirmed because none was "meaningful."

The second priority, said Carpenter, is to foster the integration of Salomon Smith Barney and Citibank "in a productive way," and the third is for him and Menezes to get to know the people and the businesses they've been put in charge of. That last priority suggests how much of a learning curve these men have to climb: Menezes hasn't had a job in corporate banking for nearly ten years and doesn't know investment banking at all; Carpenter hasn't been in investment banking since Kidder Peabody and has never been a commercial banker. No problem, says Menezes: He and Carpenter have the advantage of coming in, he says, with "no preconceptions" and with a "fresh eye."

For the most part, the integration called for in the second priority seems to be moving ahead well, under plans laid out by the old triumvirate and now being carried forward by the new dyad. Internationally, Salomon cannot possibly set up shop in all of Citi's emerging markets, so the idea is to have regional investment banking offices that will feed out their services. In the U.S., Citi's bankers will continue to manage most corporate relationships, calling in Salomon's investment bankers as experts. About 60 large companies, however, have been assigned special status and will be regularly called on--so the blueprint says--by a two-person team made up of a Citi banker and a Salomon banker. This is where the turf battles could erupt. Maughan, from his new observatory post, says that he believes most joint-calling arrangements are jelling, except in a few instances where "cultural stereotypes"--the arrogance of investment bankers, the stay-off-my-block attitudes of commercial bankers--are standing in the way. And these, he says, are just management issues that have to be dealt with.

Another of these is derivatives. In December, nine months into this merger, the two sides were still arguing over how to combine two derivatives operations that are both moneymakers but that happen to operate in hard-to-conform ways. Carpenter describes this as a "high-class problem." Nonetheless, it is clear he and Menezes are getting migraines trying to solve it. "Running separately," Carpenter says, "the two operations are obviously bumping into each other," and that's not bearable. It's a good thing that there weren't many overlaps in this merger (other than chairmen, of course), because derivatives proves how hard they'd have been to handle.

January's big problem will be bonuses. They are due to be paid on the 28th, and the organization sits in suspense, wondering how many defections may occur once this loot has been ladled out. Two headhunters who specialize in financial services, Emanuel Monogenis of Heidrick & Struggles and Roderick Gow of LAI Ward Howell, said recently that the number of Citigroup people calling to say they are considering leaving has jumped. Both men say they've seen "waves" in the calls: They came mostly from Citi people right after the merger was announced, but lately--most especially "since Dimon"--they've come from the Salomon camp.

Morale has not been helped by a speech that John Reed, ever candid and truthful, made in late October to the Consumer Bankers Association and that in effect blasted just about everybody in Citigroup's corporate business. Reed remembered that he had first reacted to Weill's merger overtures with the thought, "Who wants to be part of Salomon Brothers' trading and all that stuff?"--which, considering third-quarter results, wasn't such a dumb thought. He said he'd had someone look at 186 years of Citi profits and had concluded that the corporate side of the business probably had not made any money in that time. (If you find this hard to believe, think of the billions of dollars of losses the company took on emerging-markets debt in the 1980s.) Said Reed: "I really wish that my job were to run only the consumer side of the business, because I think it is a fantastic business. But I have been destined--you know, God punishes us all--to be constantly surrounded by the other part of the business, which inevitably gets us into trouble."

Yes, says Sandy Weill, he can well imagine that the people on the corporate side of Citigroup might have found those remarks hard to take. "He's not going to say that again," Weill adds with a grin. "John now loves the corporate side of the business, and he's never going to say anything like that, ever again."

Sure. Well, maybe he won't if this corporate thing that he and Weill have basted together gets organized and begins making good money. That thought raises another, related to Reed's answer when he is asked by financial reporters what companies Citigroup should be compared with. There are really none that make sense as a comparison, he says: "We're in a class by ourselves." Okay, guys, prove it.

REPORTER ASSOCIATE Jeremy Kahn