4.13/The Lived Worlds/Class and the City

Wall Street as Workplace, 1989–2001

The Number You Left the Desk

That principle governed a floor that still ran, in the same years, on paper. The largest bond-trading operations in the world still ran on paper. Every trade in 1990 began as a slip written by hand on a pre-printed ticket, passed to a clerk, and entered into the record before the paper went into a wastebasket that overflowed by the close. The trading floor was where the future of the American economy was priced by the minute, and the used forms lay in drifts on the floor by six in the evening. The electronic systems that supplemented the tickets did not yet replace them; the paper trail was still the primary trail, and the men who moved the paper — the clerks and junior assistants at the margins of the desk — were the ones who knew, physically, what the desk had done all day.

The room itself was an acre of humanity. Michael Lewis’s Liar’s Poker, published in October 1989, described the Salomon Brothers floor of the mid-1980s as a place where forty hours of new information arrived every minute, worked by hundreds of people making faster decisions with larger sums than their predecessors had imagined.¹ By the end of the decade Salomon’s trading operations filled much of 7 World Trade Center, and the template held: long banks of desks — “positions” — facing the same direction or facing each other across a common aisle, each position carrying two to four cathode-ray monitors in green or amber phosphor. The number of screens on a man’s desk was itself a rank. So was the seat. Who had a window. Who faced a wall. Who could read the ticker without asking someone else to read it for him.

The noise was structural. On the exchange floors it was open-outcry, shouted, hands signaling over the roar. On the over-the-counter bond desks that Salomon dominated it was phone-based and no quieter: traders shouting prices into open lines, a squawk box broadcasting market color from the desk head, and the accumulated vocal effort of four hundred people all trying to be heard at once. The windows did not open. By mid-afternoon a large floor had breathed its own air many times — hot electronics, coffee from the arrivals who had been in since six, paper, the synthetic smell of institutional carpet. Managing directors watched it through the glass walls of offices that overlooked the floor, visible in both directions: the desk could see the MD, and the MD could see the desk.

The Analyst

The two-year analyst program was the decade’s most intensive professional initiation. The major investment banks — Goldman Sachs, Morgan Stanley, Merrill Lynch, Lehman Brothers, First Boston — recruited from the same dozen universities through a campus process that opened in September and closed in November with offers for the following June. Base pay began at roughly forty thousand dollars in the early 1990s and reached around sixty thousand by decade’s end; the base was only the frame, filled in each January by a bonus that ran from a fifth to more than half of it again.

The work product was the pitch book: a bound presentation, often a hundred to two hundred pages, built for a single client meeting. It carried a model — a spreadsheet analysis of the proposed transaction — a set of comparable deals and how they had been priced, and the bank’s argument for why the client should hire it. The night before a pitch, the analyst worked until the book was ready, which might be two in the morning or five or not at all, and then carried it into a meeting where it was sometimes barely opened. What the book demonstrated was not primarily its contents. It was that the analyst could produce two hundred finished pages under a deadline that assumed he would not sleep.

The conditions were the point of the program. Eighty to a hundred hours a week was the standard description, and live deal execution pushed past it. The tether in the early 1990s was the pager, clipped to the belt and surrendered a number for at every departure from the building, a discipline of continuous reach the program never wrote down and never had to. Attrition ran from a fifth to two-fifths of each class. Some left for graduate school or the buy side; some were pushed, for an error on a live transaction or an inability to produce under the load. The analyst who left early carried a specific stigma — not career-ending, since the hedge funds and the business schools were open, but legible inside the culture as a failure of stamina. It was not discussed directly. He was simply not there anymore.

The Number

The bonus was the organizing event of the calendar, and because it was paid in January, December was a month of political maneuvering unlike anything in the ordinary corporate year.³ Reviews were conducted in November. The analyst or associate sent his manager a “brag sheet” — deals closed, hours logged, client praise recorded — and the manager carried that material, plus his own read, up to the next level to argue for a specific figure. The process was frankly political: who you had worked for, who owed your manager a favor, whether the deals you had staffed were counted as wins.

The determination happened above all of that, where the firm set a total compensation pool for each rank and divided it. In a strong late-1990s year a managing director might take home well over a million dollars against a base a fraction that size; an analyst in his second year, a small fraction of the MD’s number.⁴ The figures were not published. They moved through an informal network of comparisons that reassembled them, in aggregate, within about seventy-two hours of the day the numbers were handed out.

Bonus day itself was a brief meeting with a direct supervisor, who conveyed the number aloud — five minutes for an analyst, perhaps ten for a senior associate. The supervisor watched the face. Neither disappointment nor delight was acceptable; the poker face was the professional standard, and the departure from it — the man who went pale, the one who said something a shade too warm — was noted and remembered. The phone calls followed within the hour. The numbers were not stated plainly. “Roughly in line with what I expected” meant average; “a good year for me” meant above; a silence followed by “I’m thinking about my options” meant a man who believed he had been wronged. The comparison happened anyway, in code, across the desks and between the firms.

The city registered the January bonuses in measurable ways. The luxury-car dealerships on the West Side reported their best week. The higher-end apartment listings — the two-bedroom on the Upper East Side, the loft in TriBeCa — moved fastest in January and February. The restaurants of Midtown and the Financial District filled with the specific dinners of men who had just learned what they were worth for the year.

Where the Street Drank

Harry’s at Hanover Square was the canonical after-work institution, opened in 1972 in the vaulted brick cellar beneath the India House, the 1854 brownstone built for the Hanover Bank and later home to the private India House club.⁵ Its founder, Harry Poulakakos, had worked in the city’s restaurants for years before taking the basement on Hanover Square. The room was low, wood-paneled, dim, sawdust reportedly underfoot in the old New York manner; the drink was scotch or bourbon on the rocks, beer acceptable, wine beside the point. Private phone lines ran from Harry’s to the surrounding brokerage houses, so a trader who needed to be reachable in the evening left the desk Harry’s number.

The function of the room was liminal. A meeting at Harry’s after five was not a business meeting and not a social one; it was where professional discretion relaxed a notch and general intelligence moved — who was buying what, which desk was having a bad week, which firm was losing people. The clientele was overwhelmingly male, drawn from the firms in the surrounding blocks, and it was not discussing its families. It was processing the day and competing for standing through the telling of it.

Delmonico’s, at 56 Beaver Street, served the other function. The power lunch there was the managing partner taking a corporate client, the senior banker entertaining a chief financial officer weighing a transaction. It was expense-account ground: white tablecloths, a Continental menu, a wine list that required knowledge, prices that made the point plain. You went when the occasion required the implicit statement that you were the kind of person who came here. The recession closed the dining room in 1993, and it stood dark until new owners reopened it in 1998.⁶ At the very end of the period the grandest downtown room reopened for the same class of evening — the banking hall of the former National City Bank at 55 Wall Street, its Corinthian columns and sixty-foot ceiling turned to a hotel ballroom and hired for receptions of a thousand and more.⁷ The genuinely uptown rooms — the King Cole Bar at the St. Regis, the Oak Bar at the Plaza — were for client entertainment near a client’s hotel, not for the informal drinking that ran downtown after ten.

Drexel: February 13, 1990

Drexel Burnham Lambert filed for Chapter 11 bankruptcy protection on February 13, 1990, the largest failure of a securities firm to that point in American history. Drexel had built the modern high-yield bond market and financed the leveraged-buyout wave of the 1980s — KKR’s takeover of RJR Nabisco, the corporate raiders who used Michael Milken’s junk bonds to put companies without blue-chip credit into play.⁸ At its mid-1980s height the firm generated more revenue than Goldman Sachs.

Several forces converged at the end of the decade. The junk market had collapsed, with many high-yield bonds trading at fifty cents on the dollar or less, and major issuers defaulting. Federal legislation in 1989 required savings-and-loan institutions to divest their junk-bond portfolios, removing the largest single class of buyers from the market.⁸ Drexel had already agreed, in 1988, to plead guilty to six felony counts and pay a $650 million fine — a settlement that drained its capital. And Milken, the firm’s revenue engine, had been indicted in 1989 on ninety-eight counts and was gone from the firm. On February 9, 1990, the Securities and Exchange Commission ordered Drexel to stop moving capital from its regulated broker-dealer to its parent. The firm was insolvent.⁸ The board met on the twelfth and concluded it could not survive; the filing came the next day.

The afternoon had the finality that financial collapses carry. Roughly five thousand employees were told to clear their desks. The trading floor went dark; the securities positions had to be unwound as counterparties called and the back office fell into chaos. The shock was not that the firm was in trouble — its difficulties had been public for months — but that something this large and this permanent-seeming was suddenly not there.

What followed was a reckoning about heroism. Milken had been presented, by himself and his admirers, as the meritocratic disruptor: the kid from Encino who had made it to Wharton and grasped that junk bonds could open capital to companies the establishment would not fund. On November 21, 1990, a federal judge sentenced him to ten years in prison and he agreed to pay $600 million in fines and penalties; he served under two.⁹ He had pleaded guilty that April to six felony counts — securities and tax violations arising in part from his dealings with the arbitrageur Ivan Boesky, not the racketeering he had been charged with.⁸ The cocktail-party line shifted from persecuted for making too much money to he knew exactly what he was doing, and both positions kept their adherents through the decade. The industry had elevated a man to the status of folk hero and then watched him plead guilty, and the more interesting question — what that elevation had said about the industry’s standards — was harder to sit with than either verdict.

Long-Term Capital: September 1998

Long-Term Capital Management was founded in 1994 by John Meriwether, formerly the head of bond arbitrage at Salomon Brothers, who had resigned in the Treasury scandal of 1991. His founding team included former Salomon traders and two academic economists, Robert Merton and Myron Scholes, who shared the Nobel Prize in economics in October 1997 — a year before the fund failed. The strategy used quantitative models to find small pricing gaps between related securities and then applied enormous leverage to make the small gaps pay. The model assumed that historical relationships would hold: that spreads which widened would, in time, converge.

They held for three years. The fund returned forty-three percent in 1995, forty-one percent in 1996, seventeen percent in 1997.¹⁰ By 1998 it controlled roughly $125 billion in assets on a few billion of capital — leverage of about twenty-five to one — with off-balance-sheet derivative positions whose notional value exceeded $1.25 trillion. On August 17, 1998, Russia defaulted on its domestic debt and devalued the ruble. The flight to quality that followed moved correlations in directions the model had ruled out: instead of converging, divergences widened across asset classes at once. The fund lost $4.6 billion in under four months — $1.9 billion of it, nearly half its capital, in August alone.¹⁰

On September 23, 1998, the Federal Reserve Bank of New York, under its president William McDonough, gathered the fund’s major counterparties at its offices at 33 Liberty Street. Fourteen institutions — Goldman Sachs, Merrill Lynch, Morgan Stanley, and the major European and American banks among them — put up $3.625 billion to take ninety percent of the fund and unwind its positions in order.¹⁰ The Fed lent none of its own money; it supplied the table and the urgency. The knowledge that the global financial system had come within days of a cascading failure sat, at first, with perhaps a couple of hundred people directly in the negotiations. The wider Street knew something serious was moving but not precisely what. The men at Harry’s in late September who did understand were drinking with the particular knowledge that they had watched the machinery of world finance nearly seize, and that most of the room did not know what it had just survived.

The Constraint No One Discussed

The compliance department existed in every major firm and belonged to none of the culture that Liar’s Poker had described. It was the institutional admission that the floor needed a constraint it would not impose on itself. Compliance reviewed customer communications for the misrepresentations and prohibited practices — front-running, churning, unauthorized trading — that the regulations forbade. It maintained the documentation of the Chinese wall between the banking side of the firm, which held material non-public information about the companies it advised, and the trading side, which dealt in those same securities. It kept the regulatory record on every registered employee — the Form U4 filed on hiring and the U5 on departure, each carrying the employment history and any disciplinary marks.

The compliance officer was at once a potential source and a potential obstruction. The department was one of the few corners of the firm where women reached senior levels — in part because the trading floor had been hostile to them, in part because compliance was seen as less prestigious and so more open to candidates from outside the usual line. She knew where the procedural bodies were buried: which traders had drawn written warnings, which desks had been reviewed, which supervisors had failed to supervise. She also had a professional duty to the firm, and under some circumstances to the regulators — not to a reporter.

The department’s characteristic artifact was the cover-your-ass memo. Asked to do something he was unsure of, the careful professional documented the concern and the answer he received — “I have discussed the above with [supervisor] and have been advised that [the action] is appropriate” — filed internally, not with any regulator, as proof that he had raised the question if the transaction later went wrong. The memo was the evidence that a culture running on oral instruction and handshakes had learned to make paper when paper was protective. The Salomon Brothers Treasury scandal of 1991 was the case of the paper not made. Paul Mozer, who ran the firm’s government-bond desk, had submitted unauthorized bids at Treasury auctions, using clients’ names to exceed the limit any single bidder was allowed. He told his supervisor, John Meriwether; Meriwether told the chairman, John Gutfreund; and the firm did not report the violation to the Treasury for months.¹¹ When it surfaced, Gutfreund, Meriwether, and the firm’s president, Thomas Strauss, resigned; Warren Buffett stepped in to run the firm on an interim basis, and Salomon paid $290 million to settle.¹² The infrastructure had existed. The culture had chosen not to use it.

It was a culture that had put its own name on a game. Lewis took his title from the floor’s own pastime — a bluffing contest played with the serial numbers on dollar bills — and from the most famous hand of it, when Gutfreund reportedly turned to Meriwether on the Salomon desk and proposed a single round: “one hand, one million dollars, no tears.”¹ The men who ran the firm competed at reading and concealing information for its own sake. Each of the two men who sat across that hand presided, before the decade was out, over a different failure of exactly that skill.