(FORTUNE Magazine) – They did not think of themselves as Masters of the Universe, at least not yet. They were just a bunch of startup folks, gathered in a startup office, listening to that startup pep talk that everybody else in Silicon Valley was hearing back then. And like every startup pep talk, it included a statement of principles. There were ten in all, the first of which admonished everyone there to "act ethically."

Their boss, Frank Quattrone, hammered home the point that if the people in that room did not follow that prescription, they would not, and could not, succeed. The goal of the group--most of them, like Quattrone, defectors from Morgan Stanley--was to build an investment banking powerhouse. The business cards said that they all now worked for Deutsche Morgan Grenfell, the investment banking arm of the giant Deutsche Bank, but everybody in the room--indeed, everyone in Silicon Valley--knew that it would have been a lot more accurate to call the enterprise Quattrone & Co.

In the years following that 1996 meeting, Quattrone & Co. would become almost as closely associated with Netmania as Drexel Burnham Lambert had been with the junk-bond-fueled takeover battles of the '80s. It was, in fact, Quattrone himself who, as a Morgan banker, had brought Netscape public, sowing the seeds of the Internet frenzy. At Deutsche Morgan Grenfell and later at Credit Suisse First Boston, Quattrone would become the most visible--and powerful--investment banker in Silicon Valley. The Internet stock run-up of the late 1990s would make many of the people in that room, particularly Quattrone and his top lieutenants, George Boutros and William Brady, extremely rich. And it would sorely test the limits of that early terse prescription.

By any estimation CSFB is now in very hot water. It is the subject of investigations by the National Association of Securities Dealers, the Securities and Exchange Commission, and the U.S. Attorney's office in New York. The firm has been notified by the NASD that six East Coast sales and trading officials may face charges for taking inflated commissions--essentially kickbacks--in exchange for doling out hot tech IPO shares in 1999 and 2000. In response to the investigations, CSFB has fired three members of an elite private client services (PCS) brokerage force that catered to rich tech insiders--the venture capitalists and entrepreneurs known in Silicon Valley as Friends of Frank. And it has seen CEO Allen Wheat sacked by CSFB's parent company, the Credit Suisse Group. Though the firm claims that Wheat's firing is unrelated to the investigation, few in Silicon Valley or on Wall Street give that much credence.

In the harsh glare of the investigations, CSFB has begun stripping away elements of Quattrone's pervasive control, eliminating his authority over the private client services group as well as the 60-member technology research team that Quattrone both supervised and paid. In a July telephone conference, journalists and analysts pressed Credit Suisse Group CEO Lukas Muhlemann and Wheat's replacement, John Mack--Quattrone's former boss at Morgan Stanley--with pointed questions about CSFB's "compliance structure" and Quattrone's future role at the firm.

Through all this, CSFB executives maintain that the firm has done nothing but follow accepted industry practices. They note that at least half a dozen Wall Street firms, including Morgan Stanley and Goldman Sachs, have been subpoenaed in the IPO probes. Quattrone, the executives add, has not been named as a target of any investigation and was "not responsible" for overseeing brokerage accounts, commissions, or IPO allocations. CSFB (whose spokesperson said in a statement that it is cooperating with investigators but would not comment on specific points in this story) insists that he did nothing wrong, even though the brokers that it has fired were part of a select group that Quattrone supervised--and indeed had created to cater to the personal finances of his most treasured clients. Nonetheless, there is little question that Quattrone, who declined repeated requests for interviews, is in the sights of the prosecutors and regulators circling the firm.

In trying to understand Frank Quattrone and his organization, FORTUNE conducted its own four-month investigation, which included more than 100 interviews with current and former employees of CSFB, investment bankers from around the industry, and executives and financiers in Silicon Valley. In discussions with both Quattrone's fans and detractors, the name Michael Milken comes up over and over. There are obvious parallels: As the dominant investment bankers of their day, Milken and Quattrone came to personify the giddy excesses of their respective eras--the former, the go-go 1980s; the latter, the wildly speculative Internet stock bubble. Quattrone's fans emphasize that, like Milken, he is a maverick who broke the mold of the traditional investment banker and argue that he is being pursued less for any real misdeeds than for his success. His detractors say that--again like Milken--he amassed enormous power by ignoring the normal checks and balances of the financial world.

Eventually, of course, Milken pleaded guilty to six felonies and spent two years in jail. For all the headlines the CSFB investigations have generated, it is far too early to say whether the same fate awaits Quattrone. The law is murky; not all the facts are in; potential key witnesses have not yet come forward. Indeed, it isn't hard to find people who say that the actions taken by Quattrone & Co., even cast in the harshest light, simply do not cross into illegality. It's not too early to say this, though: Rarely has any advice sounded so hollow as that uttered by Quattrone back in 1996--before it all got crazy.

Profiles of Quattrone, 45, invariably note that South Philadelphia, where he was raised, was the setting for the movie Rocky. It is an image Quattrone relishes. He has always presented himself as a scrappy, street-smart outsider and likes to snicker about Wall Street bankers with "suspenders and slicked-back hair." That picture, however, is a bit misleading. By 1995, the year he took Netscape public, Quattrone--who has an undergraduate degree from Wharton and an MBA from Stanford--had already spent 17 years at Morgan Stanley, certainly enough time even for a kid from South Philly to become acculturated to Wall Street.

Quattrone came to Silicon Valley in the early 1980s, a time when the big Wall Street firms regarded tech banking as just an interesting niche business. But Quattrone believed that the tiny, struggling companies that then made up the tech galaxy were destined to become fast-growing giants. Building relationships with them was certain to pay off, he argued, even if it took a long time. All that, of course, seems obvious in hindsight. But at the time it was revolutionary. "He was the first guy in the New York investment banking world to take Silicon Valley seriously," says well-known tech investor Roger McNamee. "He was the first guy at a major firm who bet his career on Silicon Valley."

As an institution, Morgan Stanley was a huge draw; entrepreneurs and venture capitalists practically drooled at the prospect of being courted by so prestigious a firm. But Quattrone often operated as if the franchise he was developing wasn't Morgan's but his own. When he hosted tailgate parties at Stanford football games for VC buddies, a former colleague says, other bankers who worked with him at Morgan sometimes found themselves cut out.

Quattrone built his own intensely loyal following in the Valley by cultivating an anti-banker persona. He was rumpled and studiously casual--friends called him "Frankie." He grew a bushy mustache, wore ugly sweaters (they became a Quattrone trademark), carried his pitch books around in a gym bag, and organized karaoke sessions at client outings.

But that shtick conceals a banker who even bitter rivals concede is enormously talented. His grasp of technology, his work ethic, and especially his attention to detail were extraordinary. "He'd call up a second-year associate and ask if the confidentiality letters had gone out to the B list," recalls a former colleague (who, like most people FORTUNE spoke with about Quattrone, would talk only if promised anonymity). "Then he would call to the syndicate desk person doing reservations and ask, 'Did you remember that the management team wants to be in Denver for the weekend?'"

By the mid-1990s, Quattrone's bet had paid off--at least financially. His salary and bonus had reached $6 million. Yet Quattrone constantly complained that he and his group were understaffed, underfunded, and underappreciated. "What ticked Frank off," says one former Morgan colleague, is that while he "was getting enormous stature in the Valley, it was not translating into stature in New York. Everything was a struggle."

In 1995, Quattrone presented his superiors in New York--including then-Morgan president John Mack--with what they regarded as an "outrageous'' proposal. He and his top lieutenants wanted to be paid not the usual salary and bonus but a percentage of a tech "revenue pool." As defined by Quattrone, it was to include not only the group's underwriting fees but also a portion of the entire firm's trading commissions for all tech stocks his bankers had underwritten. Quattrone also wanted to oversee and help set compensation for Morgan's tech analysts--a breach of the long-established separation of investment banking and research. "It's part of his controlling nature," says one of Quattrone's former Morgan superiors. "He didn't want analysts writing negative reports on clients without his having a chance to not just argue the point but dispose of the point. He wanted the final say." Morgan's top executives wouldn't even consider his proposal. They deemed it dangerous, wrong-headed--and totally at odds with the notion, taken seriously at Morgan Stanley, of a single, integrated firm.

Quattrone's response came soon enough. On Easter Sunday 1996, he announced that he was jumping to Deutsche Morgan Grenfell--and taking his top deputies, Brady and Boutros, with him. The move had been orchestrated by Carter McClelland, Quattrone's mentor at Morgan Stanley. McClelland had left Morgan in late 1994 and become head of DMG's fledgling investment banking operations in the U.S. Eager to establish DMG as a major presence in booming Silicon Valley, McClelland gave Quattrone everything he had wanted from Morgan Stanley--and more. In fact, says McClelland, DMG hadn't really hired Quattrone; the German firm and the tech banker had entered into a "joint venture."

DMG was bankrolling what amounted to a firm within a firm. Quattrone and his top deputies would all get multimillion-dollar contracts. But the real windfall would come from profit participation. After Quattrone's group covered a narrowly defined set of direct costs--salaries, bonuses, and office expenses--it would split all tech revenues fifty-fifty with the bank. That pool would include tech underwriting and M&A fees as well as a piece of the commissions from all tech stocks the firm traded worldwide. The package was so rich that other top bankers McClelland tried to hire asked whether they, too, could "get a Frankie."

In everything but name, the group that DMG let Quattrone create was Quattrone & Co. It had its own budget, PR staff, and marketing team. It had its own cadre of high-net-worth brokers to deal with the personal finances of Silicon Valley executives important to Quattrone, who had long felt that Morgan Stanley's brokers didn't give his clients the kind of service they deserved. Quattrone even gave himself the title of CEO. "All of us at DMG Tech felt like we were working for Frank, not some bank in Frankfurt," says David Britts, who headed Quattrone's data-communications banking group for three years. Quattrone, who had always cast himself to his Silicon Valley clients as one of them, now pronounced himself a full-fledged entrepreneur, leading the industry's hottest "startup.'' The following spring he landed his first big deal, winning a hard-fought contest to underwrite Amazon's IPO.

This was the group that, two years later, CSFB would inherit. In early 1998, in the aftermath of the Asian crisis, the Germans announced huge layoffs and abandoned plans to build out their U.S. investment banking arm. Beyond that, the bank seems to have had specific issues regarding Quattrone's operations. McClelland soon resigned and was replaced by an executive with serious misgivings about Quattrone's extraordinary contract. Part of the issue was the money. According to a former top DMG executive, in 1997 the tech group had produced $157 million for Quattrone's team and the bank to split--including $21 million that Quattrone pocketed himself. Some Deutsche Bank officials, however, argue that when the real costs of the tech group were added up--costs that were charged to other parts of the bank--Deutsche Bank actually lost money on Quattrone.

Control was also a point of contention. Even after gaining all that he'd asked for at Morgan Stanley, Quattrone demanded more. He often spoke of wanting all the disparate parts of an investment bank "singing from the same hymnbook"--his own. One DMG executive says that he would call "at all times of the day and night" to argue for direct control over IPO allocations. After leaving DMG, Quattrone told a Silicon Valley reporter, "We had control over about 90% of our needs, but we didn't have control over sales and trading. In the end, that was an issue."

In July 1998, Quattrone, Brady, and Boutros announced on a firmwide voice mail that they were moving to Credit Suisse First Boston. The three were contractually barred from contacting their 150 colleagues, but CSFB executives weren't--and did. Within days, virtually the entire tech group had signed on to follow Quattrone. The defection was so complete that CSFB negotiated to take over for Quattrone's group the lease of a new two-story building in Palo Alto that DMG had nearly completed. The signs were simply changed in front of the building.

Quattrone & Co. would remain Quattrone & Co., only under a different flag. That suited CSFB much better than it had Deutsche Bank. First Boston was known for its risk-taking culture. It also had a history of letting key rainmakers create the kind of autonomous "firm within a firm" that Quattrone desired.

There's no question that the CSFB Technology Group brought in a great deal of money. Both in 1999 and 2000, CSFB managed more Internet and technology IPOs than any other firm. Credit Suisse Group CFO Philip Ryan told Bloomberg Markets magazine that Quattrone's group generated 12% to 15% of CSFB's total revenue--which last year reached $12 billion. Quattrone himself may have earned as much as $100 million, according to a close associate. On the surface, CSFB's bid to buy a technology franchise had clearly paid off. But to anybody who looked past the numbers, there was--or should have been--a lot that was troubling about how Quattrone & Co. did business.

For one thing, in Quattrone's shop, research was expected to serve the bankers' interests. The Internet craze had led analysts at every investment bank to issue glowing reports on Internet companies that were little more than an idea and some PowerPoint slides--a process that Bill Burnham, a former CSFB Internet analyst, calls "the competitive devaluation of underwriting standards." But nowhere did the wall between research and banking fall so completely as in Quattrone's group--both at DMG and at CSFB.

While some analysts insist that Quattrone believed in honest research, others say he tried to bully them. "I'll have you out of here Monday morning if you say that," one former DMG managing director recalls Quattrone telling an analyst who wanted to issue a less than flattering report about a client. "Do you want to work in this firm? Do you want to be a team player? When it comes time for bonus review, all this will be remembered." The managing director says Quattrone would even demand that he raise his earnings estimates. "I don't think you understand the business," he recalls Quattrone telling him. "I've been doing this for 25 years, and I think this company's going to make a shitload of money."

At CSFB, one former employee recalls that on at least two occasions Quattrone's bankers agreed to underwrite deals without first seeking an analyst's endorsement of the company. It's a notable oversight, since even in the middle of the Net frenzy, bankers stuck to the convention of at least pretending that the firms whose stock they were selling had been vetted by their research department. When the analyst expressed doubt that the companies were worthy of positive research reports, the bankers said they would simply assign them to someone else.

The merging of research and banking at Quattrone & Co. into one seamless operation clearly compromised the firm's research. There is, however, nothing illegal about that. And if bankers at competing investment houses did not explicitly have Quattrone's power to order analysts to rewrite reports, their research was hardly less conflicted. But there were other practices at DMG and CSFB that edged further into the realm of the unsavory.

During the IPO frenzy of 1999 and 2000, shares in IPOs were routinely priced below their real market value. The ability to "get into" hot technology offerings--to buy shares at a low IPO price and see them double or triple in value in the first day of trading--became a sign of status in Silicon Valley, as well as a lucrative source of pocket money for executives and venture capitalists. IPO shares "were like party invitations," recalls one dot-com CEO. "People felt snubbed if they didn't get them."

It is customary for underwriters to offer blocks of IPO shares to their top brokerage clients as a reward for generating lots of commissions. But in the 1990s firms also began offering shares to potential clients who weren't big brokerage customers but who might, say, want to hire bankers for millions of dollars of M&A advice. They did this by setting up brokerage accounts specifically for hot IPOs. Under those arrangements, VCs and entrepreneurs made a moderate deposit (perhaps $250,000) and signed over discretionary authority to the brokers whose firms were seeking their favor. Typically, IPO shares would be flipped for a quick--and riskless--windfall. "The stock would go into the hands of venture capitalists and the managements of companies that were going to go public next," notes a Silicon Valley fund manager. "This was the closest thing to free money that there was. It may not be all that different from a briefcase filled with unmarked tens and 20s."

This controversial practice, called spinning, drew embarrassing press in a 1997 Wall Street Journal article. Not only did it smell like an out-and-out bribe, but it also appeared to violate SEC rules. Among those denouncing the practice in the Journal article was Quattrone. "At its extreme," he told the paper, "an IPO is priced Wednesday. Thursday morning you call 25 venture capitalists and say, 'By the way, XYZ just went public at 15; it's now trading at 30. You just sold the allocation at 29 1/2. I hope you're happy.' That to me is smarmy."

In fact, but for a nicety or two, Quattrone's private client services group embraced precisely that practice. Indeed, there were several aspects of First Boston's PCS group that set it apart even from those of other firms. First, the head of the PCS group, John Schmidt, did not report exclusively to CSFB's sales side--as was common at most firms; he also reported to Quattrone himself. Schmidt, a distinguished-looking 54-year-old, had run the San Francisco high-net-worth office of Lehman Brothers, had served on the NASD board of governors, and was Deputy Assistant Secretary of the U.S. Treasury during the Carter Administration. His stature is itself an indication of the importance that Quattrone attached to the group.

Second, many of the accounts set up by Schmidt's group had only one purpose: to allow Friends of Frank to flip IPOs. More than a dozen influential Silicon Valley figures confirmed to FORTUNE that they had PCS accounts with CSFB--variously known as Friends of Frank accounts or simply as QAs, as in "Quattrone accounts." Some described having two brokerage accounts: one for their normal business and a second dedicated exclusively to flipping IPOs. One prominent venture capitalist in Silicon Valley says he opened a discretionary account at CSFB with $250,000. Six months later it was worth $1.5 million.

A tech insider recalls getting a phone call from a PCS broker named Mike Grunwald the day he joined a firm where he could steer business to CSFB. "Welcome to the team!" he recalls Grunwald telling him. According to the insider, CSFB opened up his discretionary IPO account after he faxed in a form and sent over an initial deposit of just $10,000. "Every IPO [Grunwald] bought and sold in my account before I knew it happened. One day you get a confirm, the next day you get another confirm, and you made 22 grand."

However unseemly the practice seems now, handing out cheap shares in hot public offerings was so pervasive during the Net bubble that it hardly raised eyebrows. Even today many of the recipients defend spinning. Indeed, two Silicon Valley CEOs, who asked that their names not be used, said that because several competing investment banks were offering them cheap IPO shares, they could not have been influenced when choosing between them.

More to the point, an SEC investigation prompted by the 1997 Wall Street Journal story ultimately petered out with no charges. Which brings us to the crux of the issue now facing CSFB: Did the behavior at the firm--both inside and outside Quattrone's group--cross the boundary from the merely dubious into the illegal, and if so, who at the investment bank knew about it?

In mid-1999 the SEC received an anonymous e-mail from a hedge fund investor who complained that CSFB was demanding payments in exchange for allocations of hot IPOs. According to a story in the Financial Times, CSFB uncovered the identity of the e-mail writer and persuaded him to withdraw the complaint. But the genie was out of the bottle, and the SEC launched an investigation. Since then, the NASD has also launched a probe and has sent notices to at least six CSFB sales and trading employees, five in New York and one in Boston, advising them that it may charge both them and the firm with violating its rules. A task force at the U.S. Attorney's office in Manhattan, led by the prosecutor who convicted the World Trade Center terrorists, is now conducting a criminal grand-jury investigation.

The investigations center on two charges. The first is that underwriters manipulated the IPO market, adding to a feeding frenzy for new issues by getting institutional investors to bid up the price with commitments to buy shares as soon as new stocks started trading. The second and potentially more explosive charge is the one prompted by that initial anonymous letter: that underwriters who doled out shares of cheap IPO stock asked for and received kickbacks in the form of inflated commissions. Regulators believe these arrangements run afoul of an assortment of laws and regulations barring fraud, extortion, undisclosed underwriter compensation, and market manipulation.

To understand the charges, it helps to know a bit of the history of the initial public offering. In the pre-Internet era, IPOs were considered successful if the shares rose 10% to 20% on the first day of trading. Thus, getting IPO shares was a perk that institutions competed for, and Wall Street firms generally allocated hot IPOs to big investors in accordance with their own financial interests. That usually meant that underwriters would look at how much the customer paid the firm in trading commissions. Huge customers like Fidelity and Janus historically got the biggest allocations. Smaller ones got much less.

Over the years institutional trading commissions have plunged to pennies--5 cents is typical for NYSE-listed stocks. There is a tradition on Wall Street of customers rewarding good service by giving a firm a bigger percentage of their business--or by telling the traders who took their orders to add a little extra commission, perhaps as much as 5 cents. "Put a dime on it,'' an institution might say. On Wall Street, that is known as giving the firm a "kiss."

During the the tech bubble, the IPO stakes got much higher. Investment banks had convinced their dot-com clients that getting an explosive first-day opening was more important than raising as much money as they could. IPOs routinely rose 200%, not 20%; instead of raising money for the issuer, the IPO gave big profits to the institutions allowed to buy shares. Both underwriters and the funds clamoring for big allocations knew that getting a big block of shares could mean a multimillion-dollar windfall.

Anew arrangement evolved: certain clients--for the most part smaller hedge funds or day-trading operations--began buying their way to the front of the line. Unable to direct massive trading volumes to a single firm, they "repaid" the underwriter for the juicy profits from a flippable IPO by quickly buying and selling a block of shares in a heavily traded company such as IBM or AT&T, then paying a commission on those trades of as much as $1 a share--a very big sloppy kiss.

At CSFB, it is clear that salesmen and traders who dealt with institutions were routinely pressured to increase commissions from beneficiaries of the firm's hot IPOs. The point man on the effort was apparently a CSFB veteran, George Coleman, who ran the sales-trading department, which took orders from institutional buyers, including many small hedge funds. Coleman, one of the New York employees who has been told by the NASD that he might face charges, regularly distributed to the sales force a list of clients' IPO allocations for the year. It included a calculation of how much money each client would have made by selling the stock at the close of the first day's trading, after one week, and so on. "The conversation would be, 'Look, we just handed this guy $500,000 in paper profits, and he's done a million dollars' worth of commissions with us. Is that right?'" says one salesperson. (Coleman declined to comment.) Says another: "You knew what they wanted us to use it for--'You made all this money from us; do more business with us.'"

Four former CSFB salespersons told FORTUNE that they knew of clients paying "superlarge" commissions. On hearing about the $1-a-share commissions, one recalls worrying that the arrangement would squeeze out clients who weren't playing the game. "I remember thinking, 'My guys who are doing regular business are going to get booted by these guys who are just buying their way into deals.'"

Then there's the trading in the aftermarkets, which has also come under the scrutiny of investigators. Though barred by SEC rules from accepting aftermarket orders until shares begin trading, underwriters freely acknowledge that they routinely sought informal "indications" of aftermarket interest from institutions seeking IPO allocations.

It seems clear, however, that in exchange for cheap IPO shares, underwriters asked institutional investors for advance commitments to buy blocks of stock at much higher prices as soon as trading began--a violation of SEC "tie-in" rules. Former CSFB salespeople say this practice played a role in determining IPO allocations. One recalls telling some clients what commitments to make to maximize their IPO allocations. "I don't think about it in terms of supporting the deal or artificially supporting the IPO," he told FORTUNE. "I only think about it in one term: How am I going to get my account the best allocation possible? Am I violating the 'tie-in' rules? I suppose I am. But at the time all I cared about was keeping my clients happy."

What was Quattrone's role in all this? Of course it was his IPOs that put CSFB in the game in the first place. But there is no evidence at this point that Quattrone masterminded any firmwide scheme to squeeze customers for kickbacks on their IPO profits. The practices now under investigation seem to have evolved, in the greenhouse of a superheated market, from questionable behavior that existed on Wall Street for years.

Yet by CSFB's estimation, the most suspect behavior at the firm took place under Quattrone's nose, among his handpicked PCS brokers. As it turns out, the tech brokers led by John Schmidt worked not just with wealthy VCs and entrepreneurs but with 15 to 20 small institutions too--mostly the sort of small hedge funds that would have needed special arrangements to get big IPO allocations. And it is allegations of squeezing those funds for big commissions, defense lawyers tell FORTUNE, that got Schmidt and two of his deputies fired.

The fired brokers are Schmidt, who reported to both Quattrone and a New York executive named Andy Benjamin, one of the six CSFB employees who face possible sanction by the NASD; Schmidt's top deputy, Mike Grunwald, an aggressive 35-year-old who had worked at Morgan Stanley and Lehman Brothers; and Scott Bushley, who joined the PCS team at DMG in 1998, when he was 25. As lawyers familiar with the case describe it, Quattrone's San Francisco-based PCS group embraced the highly lucrative practice of seeking supersized commissions from IPO customers--the practice employed in New York and Boston--and took steps to institutionalize it. Within Quattrone's PCS group, defense lawyers say, there were even stated percentage goals--a chunk of clients' profits from flipping hot IPOs--that the brokers were seeking to recover through outsized commissions. According to one source, these financial arrangements were mentioned in at least one e-mail.

Precisely who originated the practice is unclear. Schmidt is widely described as a hands-off manager. Originally his top deputies were three men--Dave Leyrer, Will Weathersby, and Rob Horning--who had joined the tech group after running their own money-management firm. That trio remained until the fall of 1999, when they left CSFB to return to running their own firm full-time. Their replacement was Grunwald, a hard-charging broker who was close friends with Bill Brady. Grunwald was hired to work with wealthy individuals--the Friends of Frank--and also to build the group's corporate cash-management business. When he arrived, day-to-day management of the hedge fund relationships was in the hands of Bushley and Scott Brown, another young broker, who left CSFB early last year. (All the PCS brokers declined to comment for this article.)

It is clear that Grunwald, upon his arrival, began pushing everyone in the group to make more money. A big chunk of the brokers' pay depended on both their individual performance and the profitability of the PCS team. At the time, the hedge fund business was becoming a major profit center. Among the lawyers for the various brokers who have been fired, there is a fair bit of finger pointing about who initiated the practice of what amounted to selling IPO allocations to clients. Grunwald's lawyer, for example, insists the arrangement was in place before his client arrived.

But defense lawyers involved in the case insist that the huge commissions could not have gone--indeed, did not go--unnoticed by the PCS group's supervisors, including Quattrone. "This was a firmwide, indeed an industrywide practice," says Schmidt's attorney, Richard Marmaro, a partner with Proskauer Rose. "John believed it was totally appropriate. At all times, John's chain of command knew of everything he did. To fire him for that now is absurd and outrageous." Says another defense lawyer: "The company knew damn well what was going on. It's impossible to look at the books and records that show what the hedge funds do tradingwise and not understand what's going on."

All three fired brokers, through their lawyers, have publicly denied any wrongdoing--not because they did not charge large commissions but because they insist there was nothing illegal about the practice. As for Quattrone, Schmidt's lawyer says his client spoke to Frank, his direct superior, regularly.

Throughout his career, Quattrone appears to have been deeply interested in who got IPO shares--and to have had a hand in doling them out. At DMG, in fact, his involvement in IPO allocations was an ongoing source of conflict. "There were several instances where he wanted to do something that I thought was extraordinarily generous to a particular investor," says an executive who took part in the allocation process. "It was totally out of whack with the pattern the investor had with the firm. They didn't trade with us.'' Adds the former DMG executive: "To the extent that he could, Frank wanted to influence the allocations."

Was the situation different at CSFB? First Boston maintains that allocation decisions were in the hands of the bank's New York-based syndicate desk, and that Quattrone had no influence over the process. But former CSFB employees based on both coasts say that while Quattrone might not have had formal authority, in practice allocation of technology IPOs was largely in the hands of Andrew Fisher--a CSFB equity capital markets expert who in 1999 actually moved to Palo Alto at Quattrone's behest to join the tech group. Indeed, Fisher's bio on the tech group's Website notes that "his responsibilities include managing the origination and execution of equity transactions for all technology clients."

On some deals Quattrone influenced allocations directly. "There was a group of accounts that were really Frank's accounts," says one former CSFB institutional salesperson. She points to the example of Technology Crossover Ventures, a venture capital fund whose managing partner, Jay Hoag, has had a long association with Quattrone--and that reportedly received a huge allocation in the superhot December 1999 VA Linux IPO. "TCV might get 50,000 shares on a deal,'' the salesperson explains, "but they couldn't justify that in terms of straight trading business.... It was always understood that this was an account Frank had a relationship with." (Hoag declined to comment.) Another salesperson says Fisher (who was unavailable for comment) sometimes invoked Quattrone's name in tinkering with allocations. "He'd say, 'This Friend of Frank should be nudged up.'"

None of this, of course, proves that Quattrone knew who was being charged what. It's certainly possible that Quattrone could have been involved in decisions about IPO allocations without being aware of any alleged kickback arrangements. But is it likely he was totally in the dark? Says Marmaro: "The notion that John Schmidt is out there on a boat in the middle of San Francisco Bay without any vertical accountability is absolutely false." Schmidt, says Marmaro, is being set up to serve as a "scapegoat" for higher-ups.

Not long ago Quattrone contacted Tony Ridder, CEO of the Knight Ridder newspaper chain, whom he's known for years. He was upset to hear that the San Jose Mercury News, Silicon Valley's hometown paper, was planning a major story about his troubles--and he angrily told Ridder that there was no story there. The Wall Street Journal, which has closely charted the progress of the investigation, was out to get him, Quattrone groused. Why did Ridder's paper have to join the witch hunt? "He thinks everybody's after him," Ridder explains. Despite the complaints, the story ran.

While Quattrone has maintained an uncharacteristic public silence during the eight months since the Journal broke word of the IPO investigations, it is clear that he is seething about his entanglement in the affair. During the spring he sent an e-mail to "CSFB Technology Group Clients and Supporters" complaining about the "inaccurate and misleading content" in the media coverage and warning that more stories--"including some that are likely to be very critical of me personally"--were almost certainly forthcoming.

Many of Quattrone's Silicon Valley allies share his outrage. As they see it, he is being made the scapegoat for the excesses of the tech bubble. They insist Quattrone is being targeted by regulators for practices that were widespread--and perfectly proper. "Sounds like the free market in operation," Intuit Chairman Scott Cook says of the kickback allegations. "You pay more, you get more. If you pay a higher commission...it would seem reasonable that they would give you better service." Prominent venture capitalist Dick Kramlich, co-founder of New Enterprise Associates, also calls the investigations swirling around Quattrone a "witch hunt." Says Kramlich: "There are a lot of people who want to bring him down."

Indeed, CSFB's defense centers on the argument that providing larger allocations for outsized commission payments doesn't violate any law. Columbia Law School professor John Coffee, a leading expert on securities law, says underwriters may indeed have illegally manipulated the IPO market. But he thinks going after IPO kickbacks is bogus, however well that might play to a jury. "To prosecutors, manipulation is complicated," he says. "Kickbacks are what they deal with every day with garbage-hauling companies." Echoing Quattrone's many Silicon Valley supporters, Coffee argues that if hedge funds paid inflated commissions, they were not victims but willing participants in the scheme--and that the SEC and prosecutors are mistaken in focusing on the "morality" of the side payments rather than the inefficiencies of the IPO allocation game.

The complexity of the issues ensures that the more serious investigations--by the SEC and the U.S. Attorney's office--will drag on for months. But it is hard to believe that Quattrone will remain at CSFB long enough to see them resolved. For starters, there is the possibility that investigators will uncover evidence that Quattrone was aware of the kickbacks that prompted the firm to fire the three PCS brokers who worked for him. The comments by Schmidt's lawyer certainly do not bode well. If Schmidt and the other fired PCS brokers come under government pressure, they will certainly have every incentive to cut a deal with the feds to try to implicate Quattrone.

Even if Quattrone manages to dodge any bullets from the three ongoing investigations--not to mention scores of civil suits--it is hard to view his position at CSFB as any better than tenuous. John Mack, his new and former boss, has made it clear that he won't tolerate fiefdoms. While decommissioning Quattrone's army of tech bankers would be exceedingly expensive and difficult, a huge, highly paid tech group seems a lot less indispensable in the summer of 2001 than it did two years ago.

"The seeds of his troubles," says a former colleague at Morgan Stanley, "were sown with the deal he structured, which had all the perverse incentives you could come up with if you were consciously trying to conjure up every perverse incentive. There was so much power and so much autonomy that there were none of the checks and balances you really do need."

"It'll be hard to find the violations, since everyone's a willing player," adds a prominent New York securities lawyer. "It's a tough case. Nobody's seen anything like it." But as for how Quattrone found himself in the middle of it, says the lawyer, he has only himself to blame. "You can trace Milken ten years ago," he says, "and it's like they learned nothing."

FEEDBACK: pelkind@fortunemail.com; mgimein@fortunemail.com

REPORTER ASSOCIATES Julia Boorstin, Doris Burke