CHAPTER 5

Petopia.com, San Francisco, 2000

As 1999 rolled on, the NASDAQ touched new highs. Investors demanded a piece of the Next Big Thing, leading stock prices of many internet companies to surge on the first day of trading. The media hyped those price gains, considering them signs of a successful offering, even as they handed near-instant profits to anyone lucky enough to get stock in the IPO.

The frenzy granted even more power to the Wall Street investment banks underwriting the offerings. The process involved helping company executives to shape the company’s story, price the stock being sold and choose the investors to receive the shares.

Largely hidden from the investing public, the final one of those three roles, the allocation process, was one of the most important steps in the entire IPO process. Startups hired bankers to find buyers for their shares at attractive prices, leaving banks empowered to distribute shares as they pleased. Executives may have had some favorite investors, but most of the allocations were based on suggestions made by bankers.

It was the same dual agency conflict of interest that had bothered Jobs back in 1980. Hired by companies to sell their stock, banks also had a duty to the investor clients buying the shares. What made it particularly problematic on Wall Street was the mismatch between the frequency with which investment banks dealt with the two sides of their customer base. Investors—mutual funds, hedge funds, and high-net-worth individuals, predominantly—often dealt with the bank’s trading desks day after day, with the best clients paying millions of dollars in trading commissions. Corporate issuers, on the other hand, had needs that were much more episodic—they usually underwent one IPO only, and maybe hired the bank for a bond sale, follow-on equity offering, or merger advice in the years after.

The financial incentives encouraged bankers to give preferential treatment to investors over company issuers.

As the millennium ended, investment bankers had come to realize that IPO shares were like currency they could use to reward their best clients, or prospective clients—especially if the shares were discounted in order to deliver immediate gains.

Their lock on the process gave banks the means to engage in a practice that became known as spinning. Bankers took some portion of the shares in an IPO they underwrote and allocated them to corporate executives they had identified as potential clients. Though not illegal at the time, the practice was thought to encourage corporate executives to favor those banks that had handed them shares in previous deals. Even Hambrecht & Quist was the subject of an inquiry into the practice. Hambrecht later said it took place after he’d risen to chairman and given up the CEO’s job.

In other cases, banks allocated desirable shares to investors in return for the promise of later trading commissions, sometimes calculated as a percentage of the gains made by the investor when the IPO shares surged. As banks worked for issuers to sell shares and raise corporate funds, they were handing cheap shares to favored clients in order to win future business. Critics were left to wonder, rightfully, if the banks were deliberately underpricing the IPOs.

With technology fees fueling investment bank coffers, some of the industry’s biggest players looked to cash in. In May, Wall Street’s last major investment banking partnership came to an end when Goldman Sachs sold shares to the public. Unlike the IPOs for many of its clients, Goldman’s shares rose just 33 percent on their first day of trading.

Few bankers personified the excesses of the era more than Frank Quattrone. By this time, his technology group, which included investment bankers, research analysts including Lise Buyer, and private bankers, set up retail brokerage accounts for corporate executives and seeded them with hot IPO shares. During Quattrone’s time at Credit Suisse, the bank set up more than two hundred brokerage accounts for technology executives and filled them with a few hundred shares each of the most popular IPOs, regulators later alleged. As the stock appreciated in the days and weeks following, the executives found themselves sitting on tidy sums due to Quattrone’s generosity. Known as “Friends of Frank,” the executives would often select Credit Suisse for future investment banking assignments, netting Quattrone’s group millions in fees. If they didn’t, according to the allegations, Credit Suisse would later cut the share of IPOs that they awarded to those executives.

The practice was beginning to bother Buyer, who explained how the system worked in a July 1999 email. The “only unpleasant part of all this,” she wrote to someone outside the bank, was that allocations were increasingly going to friends and family of the investment bank. “So it becomes something of a ‘you scratch my back, I will scratch yours.’”

Credit Suisse also allocated shares to institutional investors in return for trading commissions. Beginning in April 1999 and continuing into 2000, Credit Suisse bankers allocated IPO shares to over one hundred customers, largely hedge funds, willing to send between 33 percent and 65 percent of their profits back to the bank, the SEC later alleged. The bank paid $100 million to settle the allegations.

The link between allocations and trading commissions had devolved into “almost a straightforward financial transaction,” Hambrecht recalled. If you gave an investor $1 million in first-day profits from allocations of an underpriced IPO, that investor would repay the favor with $500,000 of commissions within a week. “That was pretty hard to avoid,” Hambrecht later said. “That was the primary reason why I finally felt that the only way we could change the system was to do [the IPO] with a new firm that was not dependent on the institutional brokerage business.”

Goldman Sachs was also getting into the game. On September 17, 1999, Goldman’s David Dechman sent an email to some colleagues asking them to consider how the firm could better use the IPOs they were doing to win business. “The hot deals are obviously a currency, which can be used to please institutions, please high net worth individuals, acquire new customers (perhaps for GS.com), help ECM as per the memo, etc.,” he wrote, referring to the bank’s equity capital markets business. “How should we allocate between the various Firm businesses to maximize value to GS?”

A senior Goldman Sachs executive, Robert Steel, who was co-head of equity sales, later wrote an email to a senior executive at Putnam Investments, a large Boston-based asset manager that was then one of the firm’s biggest clients: “It is my view that we should be rewarded with additional secondary business for offering access to capital markets product.”

The same month that Goldman’s Dechman sent his email, Chase Manhattan, the second-largest U.S. bank, pledged to purchase Hambrecht & Quist for $1.35 billion to expand its footprint in securities underwriting. One of the architects of the deal was James B. Lee Jr., a Williams College classmate of McCarthy’s who was a senior banker at Chase Manhattan. The transaction made H&Q the last San Francisco–based investment bank to give up its independence. Citigroup Inc. had bought Travelers Corp. the prior year, ushering in a wave of consolidation that witnessed the largest U.S. commercial banks making inroads into the Silicon Valley tech scene. NationsBank Corp. and BankAmerica Corp. had already purchased Montgomery Securities and Robertson Stephens, respectively.

The system of underwriting IPOs had broken down in other places too. The opinion of research analysts and the tailwind they could provide in selling a deal became so important that senior bankers leaned on them to write favorable reports.

But analysts had to walk a fine line. The NYSE and the National Association of Securities Dealers rules required analysts to have a “reasonable basis” for their recommendations and to write reports that gave a balanced picture of a stock’s risk and return without exaggerated or unwarranted claims. In November 1999, Citigroup analyst Jack Grubman upgraded AT&T from a longtime rating of “neutral” to a “buy” after Sanford Weill, the co-CEO of Citigroup and a board member at AT&T, asked Grubman to take a “fresh look” at the company. In return, Grubman allegedly asked Weill for help getting his kids into a prestigious nursery school. Just a few months later, in February 2000, AT&T named Citigroup subsidiary Salomon Smith Barney one of the lead managers on the IPO of its wireless business, paying SSB $63 million in fees for its work. The SEC later charged that Grubman had upgraded AT&T to curry favor with Weill and get his kids into the nursery school. Grubman got a lifetime ban from the securities industry and a $15 million fine, which Citigroup settled with regulators. Weill avoided a penalty.

“Grubman was the linchpin for SSB’s investment banking efforts in the telecom sector,” the SEC later wrote in its complaint. “He was the preeminent telecom analyst in the industry, and telecom was of critical importance to SSB. His approval and favorable view were important for SSB to obtain investment banking business from telecom companies in his sector.”

At one point, one of Credit Suisse’s bankers asked Buyer to take a look at a company called Petopia.com and prepare an IPO pitch. The request was surprising to Buyer, who felt that she should be able to maintain her independence and endorse only those companies that she honestly felt had a strong future ahead of them. “He didn’t ask me if I was interested in covering it—I was told to, because CFSB wanted the banking business,” she said. Buyer later refused to cover another company the bankers wanted to take public.

Bankers got even more bold when they engaged in laddering, the practice of allocating underpriced IPO shares with the implicit understanding that investors would buy more shares in the aftermarket at successively higher prices. When securities officials later investigated the practice, more than three hundred companies ultimately were forced to settle lawsuits over it. Why the companies? Securities law holds them to a higher form of liability for misstatements in the securities filings, with the understanding that they have the best idea of what’s said in those filings. The law holds other parties to the transaction, such as underwriters, to a lesser standard that requires them to simply prove they conducted a reasonable look into a claim, or did due diligence.

By November 1999, WR Hambrecht + Co. was busy propping up its first OpenIPO transaction. A public filing said Hambrecht and his firm had purchased as many as 6,000 shares a day, most of the 6,400 average trading volume, giving them a 13 percent stake.

When Christmas rolled around, fifty IPOs had been priced in December alone, with an average first-day trading gain of close to 100 percent.

On Friday, March 10, 2000, the NASDAQ Composite Index peaked at 5,048.62. The crash began the following week, just as McCarthy was preparing to file documents for Netflix’s upcoming IPO. Instead, Netflix shelved its plans. The company would lose almost $58 million in 2000, remaining in business by virtue of a cash cushion McCarthy had raised months before to make the company look good to IPO investors.

On March 20, a Barron’s story highlighted the fact that many of the most sought-after internet companies were burning through their cash reserves. The carnage quickly piled up. In the week of April 10, at least sixteen IPOs were withdrawn, including that of internet browser AltaVista. Many of the transactions that went through did so at prices below their offering ranges, including those of Packard BioScience, PEC Solutions, and Embarcadero Technologies. Still, as many as 376 companies were in registration to sell shares totaling $50 billion.

In May, Lise Buyer joined the venture capital industry, becoming a general partner at Palo Alto–based Technology Partners. She arrived in time to help Technology Partners deploy its seventh fund.

In November 2000, with the markets still showing a lukewarm reception to internet companies, Aristotle International Inc. withdrew its planned IPO slated for Hambrecht’s OpenIPO system. The NASDAQ Composite Index fell 39 percent for the year.

In May 2002, McCarthy tried again at Netflix. This time he was successful, raising $83 million by selling 5.5 million shares for $15 apiece. The stock closed the first day of trading at $16.75, an increase of 12 percent. The IPO valued the company, with 600,000 subscribers, at $330 million.

That month, the public began to get some answers to the question of what lay behind the dot-com boom and bust. Eliot Spitzer, an ambitious New York State attorney general who by then had earned the moniker “Sheriff of Wall Street” for his efforts to clean up the financial industry, reached a settlement with Merrill Lynch over charges that it offered “tainted” analyst research.

Merrill Lynch’s Henry Blodget, the analyst who had placed a $400 price target on Amazon.com’s shares when he was at Oppenheimer, agreed to be banned from the industry after the SEC charged him with securities fraud for allegedly promoting shares of investment banking clients even as he privately doubted the companies’ prospects.

The following year, Quattrone’s empire crumbled. The National Association of Securities Dealers alleged in a civil case that Friends of Frank accounts were used to win investment banking business, violating rules against gifts and gratuities. The NASD investigation led the SEC and the Department of Justice to open their own probes.

The DOJ criminally charged Quattrone with obstructing justice over an email he forwarded from a colleague that reminded Credit Suisse employees to “clean up” their files pursuant to the firm’s document retention policy. Authorities claimed that Quattrone knew that Credit Suisse was being investigated over an IPO it had underwritten and that his email amounted to encouragement to employees to destroy evidence.

In April 2003, Quattrone turned himself in to U.S. authorities to face the obstruction of justice charges, becoming the first Wall Street executive to be charged in connection with the investigation into the allocation of IPO shares to favored clients. His case became a sensation, captivating the attention of the public when it went to trial. Quattrone testified that he was simply following the firm’s document retention policy and knew next to nothing about the investigation into IPOs. He was convicted at trial and sentenced to eighteen months in prison. NASD imposed a lifetime ban from the industry. (Both punishments were later dropped after Quattrone successfully appealed his conviction and the Justice Department agreed to a deferred prosecution agreement under which all charges were dismissed in August 2007, and Quattrone was free to return to the financial industry.)

Analysts were caught up in the investigations as well. Spitzer joined with the SEC, NASD, and the New York Stock Exchange in announcing final terms of the Global Analyst Research Settlement, an agreement intended to eliminate conflicts of interest between investment banking and securities research. At more than one firm, analysts had been paid based on their participation in the investment banking business and the fees the firms brought in from it.

Ten of Wall Street’s largest underwriters paid $1.4 billion and agreed to a series of structural reforms to prevent analysts from being influenced by investment bankers or company executives. The underwriters agreed to physically separate research from investment banking, set up separate reporting lines and budgets, remove investment bankers from having a role in choosing which companies analysts covered, discontinue the practice of paying analysts based on investment banking fees, and ban analysts from participating in pitch meetings or roadshows.

A Frontline documentary about the worst of the dot-com crash that aired around that time explored alternative approaches to the flawed IPO process. Both Buyer and Hambrecht, who had continued using his OpenIPO system with some success, were interviewed about how a Dutch auction might be an answer to the excesses of the past. “Dutch auctions make a great deal of sense theoretically, right?” Buyer said. “Everybody pays what they’re willing to pay, and that’s how you parse out the stocks. But it didn’t play well in an environment where people wanted that marketing news, right? There is no headline that says, you know, ‘XYZ Corp Trades Up 90 Percent on Its First Day’ if there’s a Dutch auction.”

And yet that’s exactly what Google would try.