My Life as a Quant (21 page)

Ten years later I found myself on Stan's side. Similarly but with more restraint, I deplored the ambitious maneuverers I often interviewed who tried to use the Quantitative Strategies Group I ran as a conduit into trading. “I'm willing to program for a while in order to pay my dues,” was their standard refrain, and it put an end to all interest on my part. I was sympathetic to their aspirations, but I had my own responsibilities and needed people eager to do the hard detailed analytical work. This meant programming, too.

Nowadays, transitions are easier and many PhD's do succeed at getting positions directly on the “business side,” especially in smaller banks and hedge funds. But then as now, quants who yearn to move to the business side often begin to scorn their former academic skills. Stan must have well understood this self-loathing, this need to become someone else, though he certainly never seemed to want to be a trader himself. The
Forbes
reporter asked him what academic degree he had. “A PhD,” Stan retorted, “But don't tell the people I work for—they'll knock a half million off my pay!”

I agonized over Stan's offer during a skiing vacation with Eva—pregnant with our daughter Sonya—and Josh in frigid New Hampshire in late 1983. Always the obsessive, weighing every alternative, I called everyone in the world who knew someone who had worked at Goldman. I pondered repetitively. Still, I couldn't make up my mind.

Through my friend Don Weingarten at IBM I tracked down a scientist from Watson Labs who had worked for Stan for one year, developing an in-house computer language for manipulating financial time series. The people I had met at Goldman told me he had found Wall Street intolerable after a life in research.

“What was it like there?” I asked him.

“Don't go,” he said tartly, trying to give me a feel for the environment, “Unless you're willing to be yelled at all day for something you didn't do.”

Those few words settled it. I knew I would not accept the job. I needed to suffer a little more before I was ready to move.

I slogged through another year at the Business Analysis center, collaborating on a more advanced version of HEQS. Then, a year later, I resumed the job search with a few more interviews. Once, I visited Jeff Borror's IT group at First Boston and, on the day of the 1985 hurricane scare in New York, they offered me about the same money as I had made at Bell Labs. I declined. Another time, a headhunter sent me to a small midtown medical software company. When I showed up in my AT&T business uniform—the navy blazer, gray flannel pants, white shirt, and knitted tie—the door to the office suite was opened by a bunch of barefoot, twenty-something guys in shorts and T-shirts. They gave me a written test on C programming; I recall being asked to write a routine to merge two files, and also to disentangle a sequence of intentionally confusing nested
#define
C macros. I wasn't interested in working there.

By mid-1985 I was ready. I called the headhunting firm that had originally introduced me to Ravi Dattatreya and Goldman and said I was now interested in the job I had turned down eighteen months earlier. I got little response—the headhunter seemed to have lost his touch with personnel at Goldman, and was unable to get me an interview. Time passed. I grew impatient and one day I called Ravi in FSG directly. He remembered me and brought me in, and after a full day there I was once again offered a job to work on software. During the year-and-a-half that had passed, Stan had unfortunately left Goldman for Bear Stearns. This time I didn't agonize too much and accepted the job.

In November 1985 I left the Labs for Goldman. At my farewell party Larry and Mark gave warm speeches. My closest friends there, they were both secretly on their way out, too. Larry, like me close to 40 years old, was enviably in the midst of deciding that, after having been a physicist and an AT&T business analyst, he would rather be a medical doctor, and was taking biology courses in the evenings and preparing for the MCAT. A year or two later he was at Mt. Sinai Medical School, and is now a psychiatrist at Columbia doing brain research using positron emission tomography, a true combination of physics and medicine. Mark, less than a year later, took a job working for Bob Kopprasch, head of options research in the Salomon Brothers BPA group.

During the party, my first and most
sympatisch
supervisor, Ron, took me aside for a chat. He told me that I already knew more about computers than I would ever need to know on Wall Street. It seemed self-evident then, but the world was changing fast. When I got there, Wall St. was indeed backward, mainframed and undistributed. My Labs training in UNIX software development would keep me a little ahead of run-of-the-mill IT day laborers on Wall Street for only four or five years.

1
Years later, I noticed that he was almost always the only author on his reports, thanking the people that worked for him in the acknowledgements, but never including them on the author byline. After my experiences in physics, I never again worried too much about sharing credit with collaborators. It almost never did you harm.

The Financial Strategies Group at Goldman, Sachs & Co.

Learning options theory

Becoming a quant

Interacting with traders

A new cast of characters

Is that
still
the same cold you had when you started here?” a woman in the crowded elevator asked brusquely as she peered down her dry nose at my wet one some time in January 1986.

The first day for new hires at Goldman is always a Monday, and mine had been December 2, 1985. That morning I attended orientation, listened to a brief description of the firm, chose my health and life insurance plans, got ID'd and fingerprinted, and, finally, used my free lunch voucher to eat in the cafeteria. Goldman had a mere 5,000 employees back then, and there was still an intimate feel to the place.

I had left Bell Labs in late November, with only a few days to decompress between jobs. The next day the weather turned suddenly chilly. On Friday, November 29, I went for a run on Riverside Drive, wondering how I was going to keep fit when I could no longer exercise at lunch-time. By Sunday morning I was coughing and feverish. Too embarrassed to postpone my first day at work, I went in for orientation.

For the next few months I simply couldn't get enough rest to recover. Our family had been stressed the prior two years: My father-in-law had died suddenly in early 1984, my father in 1985, and now, only a few months later, my mother-in-law was in the final stages of pancreatic cancer. While Eva tended to her mother and tried to still do some laboratory research, I looked after Sonya on evenings and weekends, bearing the brunt of a typical two-year-old's appreciable will. Nights consisted of short bursts of sleep interrupted by multiple bottle fillings and subsequent diaper changes. We were all weary and upset.

Six weeks later, the dreaded early-morning phone call came from an unexpectedly different source. My sister's husband in South Africa had suddenly died. The call about Eva's mother came several weeks later.

Over several weeks my “flu” transformed itself into a semipermanent cough and runny nose that prevented me from sleeping. One chilly winter's Friday night I drove Joshua and Sonya to spend the weekend with some friends in East Hampton so that Eva could have uninterrupted time with her mother. By 10 P.M. on Sunday night I was at the end of my tether. What I needed most was some rest and sleep. I abandoned our apartment at midnight, left a message at work that I was sick, and checked into the Empire Hotel at Lincoln Center. Then, I swallowed a Valium to ensure that I would sleep, and spent the next day and a half in bed. Miraculously, that minuscule amount of rest did the trick. My nose stopped dripping, my perpetual cough deliquesced, and by Wednesday morning I was back at work and more or less normal, at least as far as my physical health was concerned.

Despite my illness, I had been learning a vast amount rapidly. My supervisor, Ravi Dattatreya, had a combative sharp-edged personality, and he liked to take pointy little verbal jabs at people, feigning to see how they would respond. But he was an excellent mentor: He had deep intuition about financial theory, and he understood how the trading and sales businesses worked.

When I arrived at Goldman's Financial Strategies Group I had thought I was going to work on the software engineering aspects of bond trading. But in 1985 the over-the-counter bond options business was burgeoning, and Ravi managed their needs. Within a few days of my arrival he introduced me to the bond options desk and assigned me to work on its valuation models. One Friday he gave me a copy of the famous Cox-Ross-Rubinstein paper on the binomial options model and told me to read it over the weekend. It was an immensely lucky opportunity; I got to work on an interesting and very relevant unsolved business problem which eventually made my reputation, and, just as importantly, I developed a relationship with traders who had a genuine and practical interest in theories and models.

The bond options desk was run by Peter Freund, a thirtyish lawyer-turned-options trader. Peter had a deep and serviceable grasp of the subtleties of options theory, but was busy much of the time writing business memos to Bob Rubin, then the head of the Fixed Income Division and soon to become cohead of Goldman itself, before eventually moving to Washington as Clinton's Secretary of the Treasury. I enjoyed talking to Peter, who sometimes placed his outstretched hand on my shoulder as we spoke, resting the full weight of his arm there for a few minutes. It was my first experience of the tactile side of being an employee. At Bell Labs supervisors didn't touch you much; at Goldman senior people tended to squeeze your arm or shoulder in a paternal and yet patronizing sort of way, even if they were much younger than you.

The most striking and voluble member of Peter's group was David Garbasz, a former graduate student in geology who had become a trader of bond options. He befriended me late that December at the annual Christmas party thrown by the Fixed Income Division at the South Street Seaport. It was a fabulously raucous party, given during the heyday of Wall Street ostentation. Giant shrimp and bite-size chunks of steak lay heaped on tables; we danced to a rock band composed of a bunch of traders; it was quite unlike the child-friendly, brownies-on-paper-plates affairs at Bell Labs to which I was accustomed. David took me under his wing and, on that freezing December night, though I was still half sick, he dragged me to accompany him on the walk home. We strode several miles from the Seaport north through the Lower East Side, able to see our breath in the midnight darkness, until I eventually found a cab to take me the rest of the way home. David was hot-headed and lively, attractively impetuous and passionately sharp-tongued and cynical, always writing letters to Bob Rubin in which he reportedly threatened to leave for greener pastures because of some dissatisfaction or other. I was a novice, and he took it on himself to teach me the ropes. We had many lunches together at the Fledermaus, the “in” place on South Street for the traders.

Another member of Peter's team was Jacob Goldfield, a young, scruffily unshaven, lanky Harvard law graduate who never even bothered to take the Law Boards, but immediately joined Goldman as a trader. David took Jacob under his wing, too, and we got along well; Jacob had studied physics as an undergraduate, and was a hands-on programmer who wrote many of his own trading tools. A very smart kid, Jacob soon radiated a quietly impressive mystique, and colleagues began to defer to him. He held his cards close to chest, absorbing information from everyone while emitting none himself, the embodiment of a perfect trading temperament. Unlike the 1980s masters-of-the-universe-style traders, Jacob never used foul language.

Everyone knew that Jacob was a protégé of Bob Rubin and had his ear. It was clear he was on a fast track.

The problem Ravi had given me in my first week at Goldman was rooted in the economics of the early 1980s. Goldman was on the “sell side”: We sold securities to the money managers, insurance companies, pension, and mutual funds that comprised the “buy side.” Many of these buy-side firms bought bonds issued by corporations, municipalities, or governments, and in the late 1970s and early 1980s, when interest rates were high, they could earn a pretty good rate of return. As interest rates in the mid-1980s slid steadily downwards from the inflationary peak of the Carter era, bonds paid lower interest, and these funds could no longer earn the high yields of the past few years. In an attempt to pump up their declining returns, many funds began to generate revenue by the periodic selling of short-dated call options against the specific bonds they held. There was substantial money to be made supplying these calls, and so, in early 1986, the Treasury bond options business at Goldman accelerated.

Loosely speaking, a call option on a bond is a contractual bet between two parties that the bond price will end up above a prespecified strike price on some definite future expiration date. The buyer of the option pays the seller a premium, the cost of the call option. At expiration, if the bond price is above the strike price, so that the buyer has won the bet, he exercises his option to receive the payoff. The payoff of the call option is the difference between the market price of the bond at expiration and the strike price. Effectively, the seller of the call option has to deliver the bond to the owner of the call option in exchange for a payment equal to the strike price, even though it is worth more. The bottom line is that the owner of the call option receives any upside the bond brings over and above the strike price, but suffers none of the downside.

The buy-side funds sold call options to Goldman to generate extra income. A mutual fund owning a $100 Treasury bond could sell a one-year call option on that bond, struck at $100, for about $2, thus boosting their yield. As long as the $100 bond itself didn't appreciate, this $2 was pure gravy. But with options, every reward has a corresponding punishment. Suppose interest rates were to fall and the bond price were to rise to $106 a year later. In this case, Goldman would exercise its call option and force the fund to accept a price of only $100 for the $106 bond, leaving the fund with a gain of $2 from the option premium offset by a $6 loss on the bond—a net loss of $4. In essence, the fund was taking a bet against interest rates declining.