The New Market Wizards: Conversations with America's Top Traders (47 page)

For me, there’s a kind of magic around here, and I don’t know if that’s something you can pick up on or not. Without that, CRT would have been a much smaller trading company. We could have stayed in business and made a profit, but nothing compared to where we are now.

When you say “without that,” are you referring to the interrelationship among the people?

JOE: Yes, that’s the stuff that makes it a blast to come to work. I consider myself to be unbelievably lucky to be able to come to work in this kind of environment, with these people.

Joe Ritchie provides living proof that creative thinking can be more powerful than complex analysis. Although he has no formal mathematical training, by just thinking about how options should work Ritchie was able to develop an option pricing model that, judging by CRT’s performance, must be better than all the academically derived models commonly in use.

The type of trading done by CRT has little direct relevance to individual traders. The primary lessons to be drawn from this interview, I believe, are not related to trading, but rather management. The enormous financial success and widespread employee loyalty enjoyed by CRT is no doubt a consequence of Joe Ritchie’s managerial philosophy: Share the responsibility and share the profits. This policy makes so much common sense that you wonder why more companies don’t use it. Corporate America, are you listening?

GETTING THE EDGE

B
lair Hull came to trading by way of the blackjack tables. This is not as strange as it may sound, since there are actually very strong parallels between the two activities. The point is not that success in trading is akin to luck in gambling, but rather that consistent winning in both is a matter of strategy and discipline, not luck. Luck plays a role only over the short term, where its potential adverse impact must be neutralized by money management controls.

After the casinos caught on to Hull’s blackjack team, he sought another avenue for applying probability theory to making money. He found the same general principles could be employed to profit from the mispricings that occurred in the option markets. Hull started with $25,000 in late 1976 and by the start of 1979 had multiplied his stake twentyfold. He continued to score consistent profits in the subsequent years, averaging roughly 100 percent per year (excluding those years in which he took sabbaticals).

In 1985, he launched Hull Trading Company to allow for a more widespread application of his trading strategies. The company, which began with a skeleton staff of five, expanded rapidly, reaching nearly one hundred employees by mid-1991. If the growth in personnel can be described as arithmetic, the expansion of computers was geometric. HTC has an entire floor in its office building devoted to its computer equipment. Another option trader with the Chicago Board Options Exchange (CBOE) who maintains an office in the same building quipped, “The building had to put in another bank of air conditioners on the roof because Hull’s computers were sucking up all the cool air.”

Hull’s company employs complex strategies, trading a broad range of interrelated option markets against each other in order to profit from temporary mispricings, while simultaneously keeping the firm’s net risk exposure to minimal levels. HTC is a market maker on a wide variety of exchanges, including the CBOE, the Chicago Mercantile Exchange, the American Stock Exchange, the New York Stock Exchange, and various foreign exchanges. They account for over 10 percent of the total trading volume in a number of options in which they make markets.

All the positions taken by the company’s traders, who now number twenty-five, are constantly monitored in real time. Strategies are continuously revised to take into account changes in both market prices and positions held by the firm, with a real-time lag of only two seconds. Now you get the idea why Hull needs all those computers.

A graph of HTC’s trading profits looks like a simulation in one of those trading system ads, except in this case the results are real. A starting stake of $1 million in 1985 grew to $90 million by mid-1991, after expenses. (The gross trading profits during this period were substantially greater, approximating $137 million.) The really remarkable achievement is the firm’s apparent modest risk level despite these very substantial gains. Since its inception, HTC has been profitable in fifty-eight of six-nine months (after expenses), with only five of these months registering a net trading loss (before expense allocations).

This interview was conducted in a conference room at HTC. I found Hull relatively relaxed and open in discussing his career. I particularly liked his candidness in talking about his blackjack-playing experiences.

How did you first get involved in the markets?

My interest probably dates back to when my grandfather charted stocks. I didn’t really understand what he was doing, but the idea of having capital working for you was appealing. The desire to learn about the financial markets led me to business school at Santa Clara University. After graduating, I got a job as a security analyst at Blair and Company. Exactly three months after I started, the West Coast research department was eliminated during the bear market of 1969.

Did you learn anything about the markets during your brief stint there?

I learned what financial analysts do. I learned about Graham and Dodd and fundamental analysis.
^*
I thought that approach was too subjective; it couldn’t be quantified or systematized. So I didn’t want to have anything to do with it.

After your job was eliminated, did you get another position as an analyst?

No, I got a job selling time on large computers. However, that position was essentially a marketing slot, and I was interested in doing analytical work. After about a year, I left to take a job in operations research for Kaiser Cement. At that time, I got interested in playing blackjack by reading a book called
Beat the Dealer
by Ed Thorp. From 1971 to 1975, I went to the Nevada casinos regularly.

Did you live in Nevada at the time?

No, I lived in California. But I would take a blackjack trip every chance I got. I probably spent about five days a month in Nevada during that time. In a sense, I owe everything that I have to the state of Nevada. It not only provided me with my original trading stake, but the betting experience taught me a lot of things that allowed me to become a successful trader.

Would it be fair to say that Thorp’s book was in some way responsible for your success as a trader?

The book certainly taught me about the methodology of blackjack. Without this knowledge, I don’t think I would be in the trading business today.

What was the basic strategy espoused by the book?

In the basic rules of blackjack, the house has a small edge. However, if a lot of small cards have been dealt—that is, the deck is rich in tens and aces—then the odds can shift in favor of the bettor by, say, 1 to 2 percent. [Tens refer to the point value of the cards and include all picture cards as well as tens.]
^*

Would it be a matter of keeping track of the cards and placing very small bets, or not betting at all, whenever there were a relatively large amount of aces and tens out?

Right. That’s essentially what I did. I would place maybe five bets an hour using that method.

My image of a blackjack table is where you sit down and are continually dealt hands. From a practical standpoint, how do you bet so selectively without it appearing awkward?

My strategy was to play only the hands that had an advantage. I stood back and did what was called back-counting. You can get away with that if you’re betting small amounts of money.

Were you immediately successful using this technique?

Actually, in my first attempt, I made only about fifty bets and ended up with a net loss. At that point, I got a little more involved in calculating how many bets it would take to make sure that I would be a winner over the long run.

In other words, the reason fifty bets didn’t work was that fifty was too small a number and still left the odds of winning too close to even.

Right. I knew that if I kept on playing with the edge in my favor, eventually I would come out ahead. Following Thorp’s advice, I started with a base of $120 and placed bets between $1 and $4. After two years, I was ahead about $10,000.

Around this time, I became friends with another blackjack player who told me about a team of players that were doing quite well. He said, “This is a very secretive team, so I can’t give you the leader’s name. But I will give him your name, and maybe he’ll contact you.” A couple of months later, my friend was killed in an automobile accident, and I assumed that put an end to any chances of contacting the team. About a year later, the organizer of the team called me.

Actually, during the interim, I had tried to put together my own team. However, I wasn’t too successful in recruiting qualified members. For example, one time we were supposed to meet at the Sahara in Las Vegas, which is the city we always played in. One of the players, however, knew only of a Sahara in Lake Tahoe. So that’s where he went. All weekend long, we couldn’t figure out where he was.

Why were you interested in a team approach instead of continuing to play solo?

Whether you’re playing blackjack or trading, your profitability depends on your edge and how many times you get to apply that edge. The team approach provides two advantages. First, assume that over a weekend of playing, the odds of my coming out ahead are two out of three. By combining banks with another person, the total number of trading days would be doubled and, as a result, the probability of winning would rise to three out of four. The more players you combine, the better your chances of a successful outcome.

In other words, if you have the edge, by greatly increasing the number of bets, the probability of success approaches certainty. It sounds as if you had created a minicasino within a casino, with the casino taking the sucker bets. What is the other advantage you referred to?

The team approach allows you to increase the maximum bet size. Theoretically. the largest bet you can make should be one-fiftieth of your capital. If you have $1,000, that means your biggest bet should be $20. If five players with $1,000 apiece combine, however, the maximum bet size increases to $100.

Are you saying that each person could determine his maximum bet size based on the combined capital base of all the players without any increase in his individual risk?

That’s correct.

Did the team accept you as a member?

The team had a series of tests that one had to take in order to become a member. I thought I was a very skillful player, but I actually failed the test. I had to increase my skills in order to become a member of the team.

What were your shortcomings?

They were in all areas. I had some flaws in basic strategy. I didn’t count the cards fast enough. I didn’t estimate decks accurately.

How do you estimate the size of the deck?

The casinos typically used four-deck shoes. The team used eight different deck sizes in one-half deck increments. You would practice identifying these different stacks, until you could tell them apart from across the room.

What method did the team use to count cards?

They used a method called the Revere Advance Point Count: twos, threes, and sixes were assigned a value of two, fours a value of three, fives a value of four, sevens a value of one, eights a value of zero, nines a value of minus two, and tens a value of minus three; aces were kept as a separate count. The higher the count—that is, the more high cards remaining undealt—the more favorable the odds for the players.

Did the count have to be standardized by the number of cards remaining?

Yes. The true count is the raw count divided by the number of decks remaining. So if the raw count is ten and there are two decks remaining, the true count is plus five. If there is only one-half deck remaining, the true count is plus twenty. What you’re really concerned about is the density of high cards in the undealt deck.

Did Thorp use a similar approach in his book?

Thorp started out with a ten and non-ten count. In his second book, he revised that so twos through sixes had a value of plus one, sevens, eights, and nines were neutral, and tens and aces had a value of minus one.

So, in essence, the Revere Advance Point Count was a more sophisticated version of the Thorp approach.

Right. There’s a basic trade-off between accuracy and difficulty in keeping the count. Even the Revere approach doesn’t represent the optimal solution based on probability theory. But if you use a more complicated (and presumably more accurate) counting method, you would be prone to making more errors.

Was it hard to keep track of the true count?

It takes a lot of practice, and you need to have discipline. Also, you develop mental shortcuts. For example, if you see a five and a ten together, you automatically associate the combination as plus because the five has a value of plus four and the ten a value of minus three.

How much time did you actually spend in honing these skills?

Initially, it takes a lot of time. After a while, it’s a matter of practicing a couple of hours before each trip.

Were you shocked when you flunked the team’s test? What happened afterwards?

There was a battery of ten different tests. I was informed as to the lack of my knowledge. I practiced for about a month and retook the test successfully.

How many people were part of the team at the time?