Fortune's Formula (19 page)

J
AMES
“J
AY
” R
EGAN
was one of the relatively few finance professionals who read Thorp and Kassouf’s book and appreciated its importance. In 1969 Regan contacted Thorp and asked if he could meet him.

Regan, a decade younger than Thorp, was a Dartmouth philosophy major turned stockbroker. Regan had worked for three brokerages, most recently the Philadelphia firm of Butcher & Sherrerd. He decided he was bored with merely executing orders. At the meeting, Regan told Thorp that he intended to start an investment partnership. He had a list of four names of potential partners. By coincidence, all four lived on the West Coast. Thorp was one of the candidates. Regan held the list carefully, like a hand of cards.

When Regan got up to use the bathroom, he left the list on the table. Thorp turned the list around and read it. It was Thorp, Kassouf, and two other names. Thorp believed that Kassouf wouldn’t be interested and concluded that Regan was almost certain to choose him. This prediction was correct.

Regan was a natural promoter and extrovert. “He was going to do the things I didn’t want to do,” Thorp explained, “which were: interface with brokers, accounting, run around Wall Street getting information, that sort of thing. What I wanted to do was think—work out theories and try to put them into action. We were actually happy being separate because we had different styles and very different personalities.”

“Being separate” was one of the oddest parts of the arrangement. Thorp did not want to give up his UC Irvine post or California. It was agreed from the outset that it would be a bicoastal partnership, connected by a wire—phone and data lines. Thorp and a staff would do the math in California. They would transmit trade instructions to Regan and staff on the East Coast. The East Coast branch would handle the business end of things, including most of the recruitment of investors.

Thorp had come from the working class, and most of his friends were mathematicians. With the possible exception of Claude Shannon, mathematicians did not have piles of money sitting around. Regan came from a comfortable East Coast background. Through family, Dartmouth, and his brokerage career, he knew wealthy people. He also had a practical sense of the markets that Thorp still lacked. Regan was, like Kimmel had been in the casinos, someone who knew the ropes.

Thorp and Regan offered a “hedge fund.” That term goes back to 1949. Alfred Winslow Jones, a sociologist and former
Fortune
magazine writer, started a “hedged fund.” The final
d
in
hedged
was later dropped.

When Jones liked a stock, he would borrow money to buy more of it. The leverage increased his profits and risk. To counter the risk, Jones sold short stocks that he felt were overpriced. This was “hedging” the fund’s bets. Jones called the leverage and short-selling “speculative tools used for conservative ends.”

By 1968 there were about two hundred hedge funds competing for the finite pool of wealthy investors. Many who became well-known managers had started hedge funds, among them George Soros, Warren Buffett, and Michael Steinhardt. In the process, the term “hedge fund” drifted from its original meaning. Not all hedge funds hedge. The distinction between a hedge fund and a plain old mutual fund is now partly regulatory and partly socioeconomic. Mutual funds, the investments of the U.S. middle class, are heavily regulated and generally cannot sell short or use leverage. Hedge funds are restricted to the wealthy and institutions. Regulators give hedge fund managers much more latitude on the hopeful theory that their wealthy investors can look out for themselves.

Hedge fund investors are thumbing their nose at the efficient market hypothesis. A typical hedge charges its investors 20 percent of profits (as did Thorp and Regan). Today, funds often tack on an extra 1 percent (or more) of asset value each year for expenses. Investors would not pay that unless they believed the hedge fund would beat the market, net of the high fees. It might seem it would be easy to determine whether hedge funds live up to this somewhat incredible promise. It’s not. Unlike mutual funds, hedge funds are not required to make performance figures public. About all that economists have established is that the public database for hedge funds, known as TASS, is rife with survivor bias. The funds that report their returns to TASS do so voluntarily.

Thorp and Regan called their new hedge fund partnership Convertible Hedge Associates. “Convertible” referred to convertible bonds, a new type of opportunity Thorp had discovered. They began recruiting investors.

The dean of UC Irvine’s graduate school, Ralph Gerard, happened to be a relative of legendary value investor Benjamin Graham. Gerard was then looking for a place to put his money because his current manager was closing down his partnership. Before committing any money to Thorp, Gerard wanted his money manager to meet Thorp and size him up.

The manager was Warren Buffett. Thorp and wife met Buffett and wife for a night of bridge at the Buffetts’ home in Emerald Bay, a community a little down the coast from Irvine. Thorp was impressed with Buffett’s breadth of interests. They hit it off when Buffett mentioned nontransitive dice, an interest of Thorp’s. These are a mathematical curiosity, a type of “trick” dice that confound most people’s ideas about probability.

At the end of the evening, Ed told Vivian he believed that Buffett would one day be the richest man in America. Buffett’s verdict on Thorp was also positive. Gerard, who had done quite well with Buffett, decided to invest with Thorp.

Regan went to the courthouse and looked up the names of people who were already partners in hedge funds. He did a lot of cold calling and got some leads. One was two wealthy brothers, Charles and Bob Evans. Charles had made a fortune selling women’s slacks. His brother, Bob, was an actor who became head of production at Paramount Studios. Thorp and Regan met the Evans brothers in New York. The Evanses were intrigued by the story of Thorp’s success at blackjack.

Bob Evans knew something of that milieu. One of his first coups as studio head was to buy the rights to Mario Puzo’s Mafia saga,
The Godfather
. Puzo’s life was alarmingly close to his art. He told Evans that he owed the mob $10,000 in gambling debts and they were about to break his arms if he didn’t come up with the money. Evans paid Puzo $12,500 to write the screenplay.

Both Evans brothers invested in the fund. At one meeting at Bob Evans’s house in Beverly Hills, Evans lounged in the pool while Thorp, dressed in stiff business clothes, followed him around and tried to explain his investment results from the side. Evans tossed out a string of questions and seemed to approve of Thorp’s answers. Every time they met after that, Evans would ask nearly the same list of questions and Thorp would supply nearly the same answers.

The money began rolling in. Thorp and Regan got a major corporate pension fund account and raised money from Dick Salomon, the chairman of Lanvin-Charles of the Ritz, and Don Kouri, president of Reynolds Foods. By November 1969, Convertible Hedge Associates was in business.

T
HE FUND’S
W
EST
C
OAST OFFICE
became the conceptual antipodes of the efficient market school at MIT and Chicago. As Thorp recalls those days, “The question wasn’t ‘Is the market efficient?’ but rather ‘How inefficient is the market?’ and ‘How can we exploit this?’”

The fund’s namesake was convertible bonds. Like any other bonds, these are loans paying a fixed rate of interest. A convertible bond is special because it gives the holder the right to convert the loan into shares of the issuing company’s stock. This feature becomes valuable when the stock rises greatly over the term of the bond. A convertible bond is essentially a bond with a “bonus” stock option attached.

It is easy to figure out what a regular bond ought to sell for. That depends on the current interest rate and the issuer’s credit. It was the “stock option” part of a convertible bond that threw people. Evaluating that was still guesswork.

Unknown to the academic community, Thorp had just about solved that problem. By 1967 Thorp had devised a version of what are now called the Black-Scholes pricing formulas for options. The value of an option depends, obviously, on the strike price, the stock’s current price, and the time to expiration. It also depends strongly on the volatility of the stock’s price. The more volatile the stock, the more likely it is that the stock price will rise enough to make the option valuable. Of course, it’s also possible that the stock will go down. In that case, you can’t lose any more than you paid for the option. Therefore, greater volatility means the option should be worth more.

The pricing formulas were complicated enough that a computer was vital to use them. Computer-savvy Thorp had a real edge over most options traders of the time. Thorp was thereby able to find mispriced convertible bonds and hedge the deals with the underlying stock.

Thorp was successful from the start. In the few last weeks of 1969, the fund posted a 3.20 percent gain. In 1970, the first full year, the fund returned 13.04 percent, after the hefty fees had been subtracted. The S&P 500 returned only 3.22 percent that year. In 1971 the fund earned 26.66 percent, nearly double the S&P performance.

The fund was prospering enough to hire new people. While the Princeton office hired a typical mix of Wall Street people, the Newport Beach office recruited largely from the math and physical sciences departments at UC Irvine. In 1973 Thorp hired Steve Mizusawa, a former physics and computer science major. Mizusawa was quiet, self-effacing, and hardworking. He slept only five hours a day (a one-hour nap around 5 p.m. and four hours from 1 to 5 a.m.). This came in handy when trading on the New York, London, and Tokyo exchanges.

As the fund grew, the salaries increased exponentially. Thorp told another UC Irvine hire, David Gelbaum, that he could probably increase his salary fivefold in five years. After this came to pass, Gelbaum asked about the future. Thorp told him he thought he could expect another fivefold increase in five years. “But I don’t think I’ll be able to do that again.”

In 1972 the fund’s computer model determined that the warrants of Resorts International were incredibly underpriced. The company was building a casino in Atlantic City, and its stock had dropped to about $8 a share. The warrants had a strike price of $40. Since the chance of the stock rising above $40 was a long shot, the warrants were deemed to be just about worthless: 27 cents, to be exact.

Thorp’s model computed that the warrants ought to be worth about $4. This was due to the stock’s history of high volatility. Thorp bought all the warrants he could—about 10,800. The warrants cost him about $2,900. Thorp simultaneously sold short 800 shares of Resorts International stock as a hedge.

The stock slumped to $1.50 a share. Thorp took advantage of the low prices to buy the 800 shares he’d already sold at the $8 price. The shares cost Thorp about $1,200, for which he received $6,400, a $5,200 profit.

The stock’s plunge also depressed the price of the warrants. But the $5,200 gain covered the price of the warrants and left Thorp $2,300 ahead.

And Thorp still had the warrants. Six years later, in 1978, things were looking up for Resorts International. Its stock price had risen to $15. That was still a long way from the $40 strike price. People offered Thorp as much as $4 per warrant. That was nearly 15 times what he’d paid. Thorp checked his computer model and concluded that the warrants should have then been worth almost $8. They were still underpriced.

Thorp turned the offers down and bought
more
warrants, selling short the stock again.

By the mid-1980s, Thorp sold his warrants for $100 each. That was 370 times what he paid. It amounted to about an 80 percent annual return over the decade, not counting the profit on the common-stock short sale.

An irony of the deal was that Resorts International was the defendant in a lawsuit on the legality of card-counting in blackjack. The newly opened casino had barred card-counter Ken Uston and his team of Czech “shuffle-trackers.”

In trades like this, the size of Thorp’s investment was limited by the market itself, rather than concerns about overbetting. The optimal position was “all you can get”—in this case, a mere $2,900 worth of warrants. This was typical. In practice, Thorp’s use of the Kelly philosophy rarely required elaborate calculations. He could make a quick estimate to confirm that a position size was well under the Kelly limit. It usually was, in which case no more exact calculations were necessary.

The Kelly formula says to bet all you’ve got on a “sure thing.” In the real world, nothing is quite a sure thing. There were a few cases where Thorp had a virtual “sure thing” trade in readily available securities. On occasion, Thorp committed as much as 30 percent of the fund’s assets to a single trade. In the most extreme case, Thorp invested 150 percent of the fund’s assets in a single “sure thing” deal. That was everything the fund had and half again as much borrowed money.

Thorp said that the real test of these aggressive positions is “whether you can sleep at night.” He scaled back his position sizes when it bothered him too much.

The card-counter must worry about the invisible eye in the ceiling. The successful trader must worry about other people copying his trades. Had others known of Thorp’s success and then learned of his intention to buy Resorts International warrants, for instance, they might have bought up the warrants before Thorp did.

One risk in keeping trades confidential is the broker executing those trades. Some traders prefer to establish a strong relationship with a single broker who can be trusted not to divulge anything. Others attempt to spread trades among many brokers. They might place an order to sell short warrants with one broker, and an order to buy the stock with another. No broker sees the full trade.

Thorp and Regan decided it made more sense to use a single broker. Powerful brokers have leeway in helping favored customers. They can make sure trades are executed quickly and offer attractive rates. A broker can also pass on information ranging from research reports to rumors. The important thing was that the broker be someone of unquestioned honesty and discretion. Regan found someone who seemed just about perfect. His name was Michael Milken.