The New Market Wizards: Conversations with America's Top Traders (21 page)

Over the long run, sure. But for the short term, single traders can definitely move the market. There’s no doubt about it. I move the market every single day. The idea is to do it in a smart way. For example, if the S&P is near yesterday’s high and I’m long a lot of contracts and want to get out, I may try to push prices through yesterday’s high to generate excitement and boost the trading volume. The increased volume will make it a lot easier for me to dump my position.

Why did you leave the floor?

I found that what I did best on the floor was position trading. I had experience developing computer models in college. Each day when I finished on the floor, I would go back to the office and develop trading models. Victor was kind enough to let me trade off some of these models, and I started making fairly consistent money. Since on the floor I could only directly trade one market, it was more efficient for me to trade off the floor.

What is the typical length of a trade generated by your models?

A holding period of about a day to a week.

My experience with trading models is that the ones that generate very short-term signals—for example, average trade lengths of one week or less—don’t beat transaction costs. I see you nodding your head, so you obviously know what I’m talking about. What is it about your models that’s different?

First, our models tend to be more statistically oriented. Second, we have lower transaction costs than virtually anyone in the business. Our round-trip commission is probably lower than 99 percent of the funds.

Why is that?

Because the combination of several hundred million dollars under management and frequent turnover means that we generate more trades than virtually anyone else. This large trading volume makes it possible for us to negotiate pretty low commission rates. Also, I believe we get some of the best executions of any off-the-floor trader. We use many different brokers in each pit, and there’s a constant selection process going on. If a broker isn’t good, we get rid of him. Conversely, if a broker does a good job for us, we give him more business. I have the manpower to call around. We can get ten people on the phone calling a given pit if we need to. We also carefully monitor our slippage [the difference between an estimated fair execution price and the actual execution price]. At the end of every single day, my staff gives me a summary sheet listing the slippage in each market.

How do you determine what the execution price should be when you’re entering an order?

Every time an order leaves a trader’s mouth, he looks at the last price on the screen. Say the price for bonds on the screen is 17 and a buy order is filled at 18, we record the slippage as -1

How do you know if it’s slippage or if the market really moved since the last screen print?

We make the simplifying assumption that over the long run it will work both ways—that is, the market will trade in the direction of your order as well as away from it. Therefore, sometimes the market will give you a more negative fill than it probably should, but at other times you’ll get a more positive one. After you’ve done thousands of trades, you get a pretty good idea of who the good brokers providing the best fills are, as well as what is a reasonable slippage factor in each trading pit.

Basically then, since your transaction costs [commissions as well as slippage] are very low, some marginal systems that wouldn’t work for a typical trader might be profitable for you.

Right. For example, if the average profit on a bond trading system is $40 per lot and I’m trading under $10 commission with a slippage factor of half a tick [$16], I’m going to be able to trade that system for a consistent profit. On the other hand, for someone paying $30 a round turn and averaging a slippage factor of one tick, that same system would be a money drain.

What portion of your profits would you estimate is a function of your control of transaction costs?

I would guess that we save about 6 percent per year in reduced slippage by carefully selecting and monitoring our brokers and another 6 percent by paying $10 per round turn instead of, say, $20.

Why did you eventually leave Victor’s organization?

Two reasons. First, I wanted to avoid the daily commute of one hour and forty-five minutes one way. Second, I felt that, over the long run, I could probably make more money on my own—although that certainly wasn’t true over the short run.

How do you define success in trading?

I sincerely believe that the person who has the best daily Sharpe ratio at the end of the year is the best trader. [The Sharpe ratio is a statistical performance measure that normalizes return by risk, with the variability of returns being used to measure risk. Thus, for example, assume Trader A and Trader B managed identical-sized funds and made all the same trades, but Trader A always entered orders for double the number of contracts as Trader B. In this case, Trader A would realize double the percentage return, but because risk would also double, the Sharpe ratio would be the same for both traders.
^*
Normally, the Sharpe ratio is measured using monthly data. Thus, only equity variability that occurs on a month-to-month basis would be considered. Trout is going a large step further by saying that, by his definition, trading performance should not only incorporate risk but should do so down to day-to-day variations of equity.]

How do you explain your success as a trader?

(A) We do good research, so we have an edge. (B) We have a rational, practical approach to money management. (C) We pay very low commissions. (D) Our executions are among the best in the business. (E) Most of the people who work here keep a large portion of their net worth in the fund we manage. Personally, I have over 95 percent of my net worth in the fund.

I take it that you’re not pulling out very much money.

I don’t pull out any money. I rent my condo and I drive a cheap car.

Is the money you’re making then more a matter of keeping score, or do you have some sort of ultimate goal?

At this point, it’s more a matter of keeping score, because I can retire today and live very comfortably off the interest for the rest of my life. The fact is that I like to trade. When I was a kid, I loved to play games. Now I get to play a very fun game, and I’m paid handsomely for it. I can honestly say that there isn’t anything else I would rather be doing. The minute I don’t have fun trading, or I don’t think I can make a profit, I’m going to quit.

Have the markets changed since you started in the business?

Volume has gone up dramatically, which is great. The markets also appear to have become more efficient. Some of the patterns I used to trade off are starting to get eliminated as other people start picking up on them.

Can you give me an example of a pattern that has become obsolete?

I used to like to put on positions in the stock market in the same direction as the price movement two days earlier. For example, if the market was up on Monday, I would be prone to be a buyer on Wednesday. This is an example of a pattern that I don’t believe in much anymore.

Your type of system development seems to be heavily based on past patterns. If you test enough patterns, aren’t some of them going to be profitable a large percentage of the time just based on normal probability? Just like if you had ten thousand people toss a coin ten times, some of them are going to get ten heads in a row. It doesn’t mean that those coins have any better chance of landing on heads on the next toss. How do you distinguish between patterns that reflect real inefficiencies in the market and those that are merely coincidental, and inevitable consequence of looking at so many patterns?

A pattern has to make sense. For example, if I find that the price change of the British pound forty days ago is statistically significant in predicting today’s price in the S&P, I wouldn’t put any faith in it. Why would the British pound price forty days ago affect the S&P? So we toss out a lot of these types of patterns even if they have a high percentage of success.

Why would you even bother testing patterns like that?

It’s actually easier to set up giant computer runs to test every combination of market and interval relationships, and then consider the relationships that appear statistically significant, than to decide which individual combinations to test.

Is a statistical emphasis one of the keys to your trading approach?

Yes, because it keeps us rational. We like to see that something has worked in the past before we use real money on it.

How many different models or patterns are you using at any given time to trade the markets?

Dozens.

For diversification reasons?

Yes. We try to diversify everything we possibly can. We like to diversify over both trading strategies and time.

By time diversification, do you mean you use the same pattern for hourly data, daily data, and weekly data?

Right. It also means simply following the markets throughout the entire trading session and being ready to trade if something happens at any time in the day.

What percentage of your trading is automatically determined by specific patterns or systems you use?

Roughly 50 percent. It’s hard to gauge because our systems may tell us to buy one thousand contracts on a given day but it’s my discretion as to when to buy them.

So your skills in entering and exiting positions are an important element in your trading performance?

Absolutely.

If you just followed the systems blindly by placing orders using some automatic entry rule—for example, buying on the opening, or on the close, or at some fixed intervals through the day—instead of timing the entry of the trade, what would you guess would be the degradation of results per month?

It’s very hard to say, but if we blindly followed the systems we might make half of what we do now. Maybe even less. I could give ten CTAs the exact systems we use, and some of them still wouldn’t make any money.

You said that roughly 50 percent of your trades are not system determined. Give me an example of that type of trade.

My favorite is the magnet effect, which we spoke about earlier. When the market approaches a round number or a critical point, I like to play for the market to get to that price.

Have you ever tried systematizing that concept?

No, because I don’t think it can be systematized. I may suddenly realize that a certain level is a key point, or the information may come from one of our floor contacts. We’re always asking our floor people, “What numbers are people looking for?”

Any other examples of a discretionary [i.e., nonsystem] type of trade?

We get constant information from the floor. We probably get a call a minute from our clerks.

Telling you who is doing what?

Right.

Is that helpful?

If, for instance, a lot of players whom we respect seem to be doing the same thing, it might prompt us to take a similar position or increase our position if we’re already on the same side.

Any other examples of nonsystem trades?

Another trade I like to do is to find out when a price move has been caused primarily by locals—we have very good floor contacts, so we get that type of information—and then go in the opposite direction. The reasoning is that the locals are going to want to cover their position before the end of the day, which is going to bring the market back from whence it came. [Most locals go home flat.]

I noticed that you have a whole crew of traders working for you. Yet, it’s my impression that you are the only one here who makes discretionary trading calls. How do you utilize those people? Couldn’t you just watch the markets on your own and use some clerical staff to help you enter the orders?