The New Market Wizards: Conversations with America's Top Traders (14 page)

That’s right, it doesn’t make any difference because there are so many different trading styles that you can always find one that will suit your personality.

Any specific advice for a losing trader?

Sometimes the reason people lose is that they’re not sufficiently selective. Upon analysis, a trader may find that if he only concentrates on the trades that do well and lets go of the other types of trades, he might actually be successful. However, if a trader analyzes his trades and still can’t make money, then he probably should try another endeavor.

What is the first rule of trading? I would argue that before anything else, the prospective trader must find the approach that he or she is comfortable with—that is, the approach that suits the trader’s personality. McKay cites this quality as the single most important element separating winners from losers. Each trader must select the appropriate market arena, choose between system trading and discretionary trading, fundamental and technical methods, position trading and spread trading, short-term and long-term horizons, aggressive and conservative approaches, and so on. For all of these opposing choices, one alternative will suit the trader’s personality, while the other will lead to internal conflict.

At this point, you might be thinking that the concept of selecting a trading methodology in sync with one’s personality doesn’t sound like much of an insight. “After all,” you might ask, “doesn’t every trader choose a method compatible with his or her personality?” Absolutely not! My own experience in this regard is detailed in the final section of this book.

In a more general sense, it is remarkably common for traders to adopt methods entirely unsuited to their personalities. There are traders who are good at system development but end up consistently overriding and interfering with their own systems, with disastrous results. There are traders who are naturally inclined toward developing long-term strategies but end up instead trading short term because of impatience or a compulsion to “do something.” There are naturally born floor traders with great intuitive skills who abandon their environment of expertise and become mediocre portfolio managers. And there are theoretically oriented individuals who develop intricate, low-risk arbitrage strategies but then decide to become position traders—an approach that may require a degree of risk acceptance far beyond their comfort levels in order to be applied successfully.

In all the above cases, individuals with a natural bent for one style of trading end up utilizing a diametrically opposite style, usually to fulfill some emotional need. In other words, the need to match personality and trading style may be a matter of common sense, but it is certainly not common. The importance of this concept, however, is highlighted by McKay’s assertion that virtually every successful trader he knew ended up with a trading style suited to his personality.

An essential element in McKay’s own trading approach is the drastic variation in position size. When he is doing well and therefore assumes his chances for success are greatest, McKay will trade very large. On the other hand, when he is doing poorly, he will shrink his trading size to minuscule levels. It is not uncommon for McKay to vary his trade size by more than a factor of 100:1. This approach serves not only to reduce risk during the losing periods but also to enhance profits during the winning periods. A trader who utilizes a constant-position-size approach gives up an important edge in much the same way as does a blackjack player who always bets the same amount regardless of the cards that have been previously dealt.

Risk control is another essential element in McKay’s approach, as indeed it is for most of the great traders. In addition to sharply reducing position size during losing streaks, as just discussed above, McKay also believes in immediately getting out of a position that has gone sour. In one of the few instances when he deviated from this self-proclaimed critical principle (the long Canadian dollar position discussed in the interview), an uncharacteristic two-day procrastination turned a $3.5 million loss into a $7 million loss.

Although McKay is predominantly a technical trader, fundamental analysis plays a critical role in defining his major trade strategies. His use of fundamentals, however, is somewhat unconventional. McKay doesn’t try to gauge whether the fundamentals are bullish or bearish, nor does he place any direct weight on whether the fundamental news is bullish or bearish. Rather, he focuses on the market’s response to fundamental news. For example, if the market is shrugging off a barrage of bearish news, McKay would view that as evidence of an impending bull move.

THE MATHEMATICIAN

W
illiam Eckhardt is one of the key figures in a famous financial tale, yet he is virtually unknown to the public. If elite traders were as familiar as leading individuals in other fields, one could picture Eckhardt appearing in one of those old American Express ads (which featured famous yet obscure names such as Barry Goldwater’s vice presidential running mate): “Do you know me? I was the partner of perhaps the best-known futures speculator of our time, Richard Dennis. I was the one who bet Dennis that trading skill could not be taught. The trading group known in the industry as the Turtles was an outgrowth of an experiment to resolve this wager.” At this point, the name WILLIAM ECKHARDT might be printed across the screen.

So
who is
William Eckhardt? He is a mathematician who just short of earning his Ph.D. took a detour into trading and never returned to academics (at least not officially). Eckhardt spent his early trading years on the floor. Not surprisingly, he eventually abandoned this reflexive trading arena for the more analytical approach of systems-based trading. For a decade, Eckhardt did very well with his own account, primarily based on the signals generated by the systems he developed but supplemented by his own market judgment. During the past five years, Eckhardt has also managed a handful of other accounts, his average return during this period has been 62 percent, ranging from a 7 percent loss in 1989 to a 234 percent gain in 1987. Since 1978, he has averaged better than 60 percent per year in his own trading, with 1989 the only losing year.

At the time of our interview, after a career of anonymity, Eckhardt was poised to expand his involvement in managed money to a broader audience. Why was Eckhardt now willing to emerge into the limelight by actively seeking public funds for management? Why not simply continue to trade his own account and those of a few friends and associates, as he had done all along? In an obvious reference to the Turtles [see next chapter], Eckhardt candidly admitted, “I got tired of seeing my students managing hundreds of millions while I was managing comparatively paltry amounts.” Obviously, Eckhardt felt it was time to collect the dues he had earned.

Trading system research is obviously something Eckhardt enjoys, and, of course, it is the way he earns his living, but his true passion may be scientific inquiry. Indeed, in a sense, trading and trading-related research is the means by which Eckhardt generates his own personal grants for the scientific projects that intrigue him. He is drawn to exploring some of the great paradoxes that continue to baffle scientists. Quantum mechanics has captured his interest because of the common-sense-defying Bell’s theorem, which demonstrates that measurements on distantly separated particle systems can determine one another in situations in which no possible influence can pass between the systems. Evolution is another area he studies, trying to find an answer to the riddle of sexual reproduction: Why did nature evolve sexual reproduction, wherein an organism passes on only half of its genes, whereas in asexual reproduction 100 percent of the genes are passed on? Perhaps his most intensive study is directed at understanding the concept of time. When I interviewed Eckhardt, he was working on a book about the nature of time (his basic premise is that the passage of time is an illusion).

Eckhardt brings many strengths to the art of trading system design: years of experience as a trader both on and off the floor, an obviously keen analytical mind, and rigorous mathematical training. This combination gives Eckhardt an edge over most other trading system designers.

How did you become partners with Richard Dennis?

Rich and I were friends in high school. We probably met because of a mutual interest in the markets, but the friendship was never about trading. Rich began trading when he was in college. I stayed in school, working toward a doctoral dissertation in mathematical logic. In 1974 I got bogged down for political reasons.

What do you mean by “bogged down”?

I was writing a doctoral dissertation on mathematical logic at the University of Chicago under a world-famous mathematician. Everything was going along fine until a new faculty member whose specialization happened to be mathematical logic joined the staff. Theoretically, I was his only student. Consequently, the supervisory role on my thesis was shifted from my existing advisor to this new faculty member, who then decided that he really wanted me to do a different thesis. As a result, after I had done all my course work, taken my exams, and finished three-quarters of my dissertation, my progress was stymied.

At the time, Rich suggested that I take a sabbatical to try trading on the floor. I did, and I never returned to school.

The shift from being a graduate student of mathematics to a floor trader sounds like a radical transition.

Yes, it was. Although I had maintained an interest in the nature of speculative prices, I have to admit that mathematical logic is a far cry from floor trading. If anything, I went into the pit with too many preconceptions of how markets work.

What kind of preconceptions?

I went in with the idea that I could apply the analytical techniques that I had picked up as a mathematician to the markets in a straightforward manner. I was wrong about that.

Did you try doing that?

Off-the-floor traders live or die by their ideas about the market or their systems. That’s not true of floor traders. As a pit trader, you only need to be able to gauge when a market is out of line by a tick, or a few ticks. Once you master that skill, you tend to survive, whether your underlying theory is sound or not. In fact, I know a lot of pit traders who subscribe to various bogus systems: moving averages, lunar cycles, and god only knows what. When they get signals from these systems, they essentially buy on the bid or sell on the offer. At the end of the month, they have a profit, which they always attribute to their system. Yet some of these systems are completely vacuous. Perhaps I did a variation of the same theme. I had ideas about speculating and trading, and I did well in the pit. But I’m not sure that I made any money from my ideas about the market.

What was the basis of your buying and selling decisions on the floor?

Basically, I would buy when weak hands were selling and sell when they were buying. In retrospect, I’m not sure that my strategy had anything to do with my success. If you assume that the true theoretical price is somewhere between the bid and the offer, then if you buy on the bid, you’re buying the market for a little less than it’s worth. Similarly, if you sell on the offer, you’re selling it for a little more than it’s worth. Consequently, on balance, my trades had a positive expected return, regardless of my strategy. That fact alone could very well have represented 100 percent of my success.

Is that, in fact, what you think?

I think that the execution edge was probably the primary reason for my success as a floor trader. The major factor that whittles down small customer accounts is not that the small traders are so inevitably wrong, but simply that they can’t beat their own transaction costs. By transaction costs I mean not only commissions but also the skid in placing an order. As a pit trader, I was on the other side of that skid.

As a former Ph.D. candidate in mathematics, did you miss the intellectual challenge in what you were doing?

Initially, yes. But I eventually got into serious research on prices, and that was as tough a problem as anything I ever came across in academia.

Were any of the areas you studied in mathematics applicable to developing trading systems?

Certainly—statistics. The analysis of commodity markets is prone to pitfalls in classical statistical inference, and if one uses these tools without having a good foundational understanding, it’s easy to get into trouble.

Most classical applications of statistics are based on the key assumption that the data distribution is normal, or some other known form. Classical statistics work well and allow you to draw precise conclusions if you’re correct in your assumption of the data distribution. However, if your distribution assumptions are even a little bit off, the error is enough to derail the delicate statistical estimators, and cruder, robust estimators will yield more accurate results. In general, the delicate tests that statisticians use to squeeze significance out of marginal data have no place in trading. We need blunt statistical instruments, robust techniques.

Could you define what you mean by “robust”?

A robust statistical estimator is one that is not perturbed much by mistaken assumptions about the nature of the distribution.

Why do you feel such techniques are more appropriate for trading system analysis?

Because I believe that price distributions are pathological.

In what way?

As one example, price distributions have more variance [a statistical measure of the variability in the data] than one would expect on the basis of normal distribution theory. Benoit Mandelbrot, the originator of the concept of fractional dimension, has conjectured that price change distributions actually have infinite variance. The sample variance [i.e., the implied variability in prices] just gets larger and larger as you add more data. If this is true, then most standard statistical techniques are invalid for price data applications.

I don’t understand. How can the variance be infinite?

A simple example can illustrate how a distribution can have an infinite mean. (By the way, a variance is a mean—it’s the mean of the squares of the deviations from another mean.) Consider a simple, one-dimensional random walk generated, say, by the tosses of a fair coin. We are interested in the average waiting time between successive equalizations of heads and tails—that is, the average number of tosses between successive ties in the totals for heads and tails. Typically, if we sample this process, we find that the waiting time between ties tends to be short. This is hardly surprising. Since we always start from a tie situation in measuring the waiting time, another tie is usually not far away. However, sometimes, either heads or tails gets far ahead, albeit rarely, and then we may have to wait an enormous amount of time for another tie, especially since additional tosses are just as likely to increase this discrepancy as to lessen it. Thus, our sample will tend to consist of a lot of relatively short waiting times and a few disquietingly large outliers.

What’s the average? Remarkably, this distribution has no average, or you can say the average is infinite. At any given stage, your sample average will be finite, of course, but as you gather more sample data, the average will creep up inexorably. If you draw enough sample data, you can make the average in your sample as large as you want.

In the coin toss example you just provided, computer simulations make it possible to generate huge data samples that allow you to conclude that the mean has no limit. But how can you definitively state that the variances of commodity price distributions are not finite? Isn’t the available data far too limited to draw such a conclusion?

There are statistical problems in determining whether the variance of price change is infinite. In some ways, these difficulties are similar to the problems in ascertaining whether we’re experiencing global warming. There are suggestive indications that we are, but it is difficult to distinguish the recent rise in temperature from random variation. Getting enough data to assure that price change variance is infinite could take a century.

What are the practical implications of the variance not being finite?

If the variance is not finite, it means that lurking somewhere out there are more extreme scenarios than you might imagine, certainly more extreme than would be implied by the assumption that prices conform to a normal distribution—an assumption that underlies most statistical applications. We witnessed one example in the one-day, 8,000-point drop in the S&P on October 19, 1987. Normal estimation theory would tell you that a one-day price move this large might happen a few times in a millennium. Here we saw it happen within a decade of the inauguration of the S&P contract. This example provides a perfect illustration of the fact that if market prices don’t have a finite variance, any classically derived estimate of risk will be significantly understated.

Besides implying that traders need to be more conservative in risk control than might be implied from straightforward statistical interpretations, are there other practical implications of using what you term a
robust
approach as opposed to methods that assume a normal probability distribution?

One important application concerns a situation in which you have several indicators for a certain market. The question is: How do you most effectively combine multiple indicators? Based on certain delicate statistical measures, one could assign weights to the various indicators. But this approach tends to be assumption-laden regarding the relationship among the various indicators.