The Complete TurtleTrader
Page 10
By overlaying the twenty-day breakout exit on the Eurodollar chart, the full trade can be seen in context. An initial “short” breakout occurred in February, with a twenty-day high breakout in mid-March. That forced an exit. That first breakout resulted in a small loss.
However, the market resumed downward in late March and another breakout signal was hit. Turtles got right back in short again. The final exit occurred when a clear twenty-day breakout high was made in July. This is seen on the chart by the slightly larger dot. Profits generated on the second trade covered the loss from the first trade and then some.
That’s the process. The Turtles could not afford to ignore the second breakout just because the first breakout resulted in a loss. They had to get back on the horse. The second breakout was the trade they were hoping for, and there was no way to predict it.
It was all a waiting game. Erle Keefer described to me their day-today process in rapid-fire summation terms: “First, you use channel breakout theory with a couple of filters. Second, you are going to size your bet by volatility. Third, you are going to have two hard stops on every trade. You are going to have the natural liquidation and you are going to have the firm hard stop. That’s what saved everybody. Rich’s systems inherently said, ‘You got to stay in the game all the time as you never know when trends are going to hit.’”
Random Entries
When a breakout occurred, whether long or short, there was no way to know what would happen next. Maybe the market would go up for a short time and then go down, giving a loss. Perhaps the market suddenly goes higher, giving a nice profit.
Eckhardt witnessed many systematic traders spending great deal of time searching for the “good” places to enter. He cautioned against it: “It just seems to be part of human nature to focus on the most hopeful point of the trading cycle. Our research indicated that liquidations are vastly more important than initiations. If you initiate purely randomly, you do surprisingly well with a good liquidation criterion.”11
Dennis actually challenged the Turtles to randomly enter the market and then manage their trades after getting in. That was a real Zen moment for many Turtles. If they applied appropriate risk management, they could handle the worst that came down the pike once they were in any trade.
Trading Your Own Account Tip #6:
Stop worrying only about how you enter a trade. The key is to know at all times when you will exit.
Risk Management: How Much Do You Bet on Each Trade?
Risk management has many names. You will find it called money management, bet sizing, or even position sizing. It was the very first concept Eckhardt addressed in class and ultimately the most important.
Turtle risk management starts with the measurement of daily market volatility. The Turtles were taught to measure volatility in terms of “daily ranges.” It was nicknamed “N” (also known as the Average True Range, or ATR). They were taught to take the maximum of the following for any market to derive “N”:
The distance from today’s high to today’s low
The distance from yesterday’s close to today’s high
The distance from yesterday’s close to today’s low
If the result is a negative number, it is turned it into an “absolute value.” In mathematics, the absolute value of a real number is its distance from zero on a number line. So, for example, 3 is the absolute value of both 3 and -3.
The maximum value of the three choices is the “true range,” or technically the absolute distance (either up or down) the market traveled in a given twenty-four-hour period. The Turtles then took a twenty-day moving average of true ranges. This gave a sample volatility for the last few weeks for each market traded.
Trading Your Own Account Tip #7:
You can determine the average true range for any stock or futures contract. Simply take the last fifteen true ranges, add them up, and divide by 15. Repeat each day, dropping off the oldest true range. Many software packages will do this automatically.
Eckhardt explained the logic behind “N”: “We found that volatility is something that can be described as a moving average process. Our incorporation of a volatility element in our trading—something that tells us how large our positions should be—has both kept us out of trouble during the tough times and allowed us to capture large gains when things are going our way.”12
The Turtles were taught multiple uses of “N,” but first they had to calculate it. Consider this example of “N” calculation:
Table 5.6: September 2006 Kansas City Wheat Futures ATR Calculations Example.
If the “N” for corn was 7 cents and the market was up 5.25 cents, then the market was up three-quarters of an “N.” That’s Turtle jargon. So “N” is a volatility measurement and a useful rule of thumb to classify how far a market has trended. Erle Keefer rattled off Turtle jargon: “When we put a bet on, we never said, ‘I am putting on a $1,000 bet.’ We were taught to think in terms of ‘N.’ ‘I got a one-half N on.’ We were taught that way because for most people, if they start to think, ‘I’ve got $34 million in bonds on,’ then the concept of money gets into their lizard brain and they start saying, ‘Oh, my God!’ We learned the correct way to think was, ‘How much did the market move today?’ It didn’t move thirty-one ticks in the bonds, it moved one and one-quarter ‘N.’”
The below chart shows “N” plotted below a bar chart of Dell. Notice how “N” can and does change. These values had to be up dated. Eckhardt updated his volatility estimates every day. He said, “That’s my routine. Two or three times a year I make an adjustment intra-day.”13
Chart 5.7: Chart Showing Dell Daily Bars with Daily ATR.
Once they had a feel for “N,” the Turtles were instructed about how much to “bet.” They bet a fixed 2 percent of their capital on hand on each trade. If they had $100,000, they would bet (or risk) 2 percent ($2,000) on each trade. Each 2 percent bet of their equity was called a “unit.” The “unit” was jargon that they used every day to measure risk.
They had unit limits on any market sector and unit limits on the total portfolio. The unit fluctuated so that every day the Turtles knew how many contracts to have on based on how much money they had in their trading account at that instant.14
Trading Your Own Account Tip #8:
Take your account (whatever size it is) and multiply by 2 percent. For example, a $100,000 account would risk 2 percent, or $2,000 per trade. It is always better to bet a small amount initially on any trade in case you are wrong—which can easily be greater than 50 percent of the time. While the Turtles typically used a 2 percent bet, you can reduce your risk and reduce your return by decreasing that number to, for example, 1.5 percent, etc.
The Turtle risk management dictated their stops, their additions to positions, and their equalization of risk across their portfolios. For example, a corn futures contract (a standard corn contract is worth $50 per cent) with an “N” of 7 cents has a risk of $350 (7 cents X $50). If the Turtles received a corn breakout signal (using a 2N stop), they would have had a “contract risk” of $350 X 2, or $700.
Assuming a $100,000 account, they would have had an “account risk” of $2,000 (2% X $100,000). The number of contracts to buy or sell is determined by taking the 2 percent account risk and dividing it by the contract risk. That gives 2.67 ($2,000/$700) futures contracts. Turtles rounded down to the nearest whole number. So when their breakout signal was hit, they traded two corn contracts for their $100,000 account.
The rules made a corn unit equal to a gold unit equal to a CocaCola unit. This was how Dennis was able to trade markets as “numbers” with no fundamental expertise in any of those markets. It was how the Turtles were able to trade such a wild cross-section of unrelated markets with only two weeks of training.
Table 5.8: Contract Calculation Method Using ATR in $ Terms.
However, the Turtles learned another use of “N” beyond a measure of volatility. It was also used as their primary stop (or exit rule, as first mention
ed with S1 and S2). The Turtles used a 2N stop. This simply means that their primary stop, or hard stop, was two times the daily “N.”
For example, if there was a breakout in corn, and assuming a closing price of $250, Turtles quickly determined their “N” stop. If the “N” was 7 cents, a 2N stop would have been 14 cents. The stop would have been 14 cents behind the entry price. An entry at $250 would have a hard stop at $236 (250 - 14). You would exit if the stop at price level $236 was “touched.” No second-guessing. No overthinking. Follow the rules.
Trading Your Own Account Tip #9:
Assume you are trading Google stock and its ATR is 20. A 2ATR (2N) stop would be 40. If you lose 40 points on Google, you must exit, no questions asked.
Chart 5.9: Chart Showing Soybean Daily Bars with Daily ATR.
On the other hand, a small “N” allowed Turtles to trade a larger position or take on more units. Soybean units purchased in August (chart 5.9) at the beginning of the breakout were 2.50 times larger than units that could have been bought at the end of the trend. This example is a great reminder of the relationship between market volatility and unit size: A low “N” value always means more contracts (or shares).
Jerry Parker found that his best trends often start with very low volatility at the initial breakout entries. He said, “If the recent volatility is very low, not $5 in gold, but $2.50 in gold, then we’re going to throw in a very large position.”15
Parker’s analysis kept showing that a low “N” measurement at the time of entry was a good thing. He said, “I can have on a really large position. And when volatility is low, it usually means that the market has been dead for a while. Everyone hates the market, has had lots of losers in a row, tight consolidation. And then as it motors through those highs, we get on board.”16
Unit Limits
It didn’t matter whether the markets were futures, commodities, currencies, FOREX, or stocks. One unit of corn, through the Turtle rules, had now roughly the same risk as one unit of dollars, bonds, sugar, or any other market in the Turtles’ portfolio.
However, the Turtles could not trade unlimited units. Each unit, after all, represented 2 percent of their limited and finite capital. The Turtles had unit guidelines to keep them from overtrading. For example, they were limited to four to five units for any one market traded.
Thus, trading like a Turtle could leave you with a $100,000 portfolio that might have purchased one bond contract, but a $1 million portfolio might have purchased five. As the bond contracts gained in value, others would be added.17
Examples of Initial Risk Determination
The following examples show the basic Turtle trading process in action.
Assume a trading account of $150,000, risking 1.5 percent on each trade and seeking to trade Swiss franc futures using a 2N stop. The Swiss franc has a single “N” dollar value of $800.
$150,000 X 1.50% = $2,250
2N stop = $1,600
The number of contracts to trade on this unit is 1.40, rounded down to 1.0.
Assume a trading account of $25,000, risking 2.0 percent on each trade and seeking to trade mini corn futures using a 3N stop. Mini corn has a single “N” dollar value of $70.00.
$25,000 X 2.0% = $500
3N stop = $210
The number of contracts to trade on this unit is 2.38, rounded down to 2.0.
The unit rules make good intuitive sense once the light bulb goes off. However, that light bulb did not turn on immediately. One Turtle described the learning curve: “When somebody says, ‘N is volatility and N is your unit size,’ I say, ‘How do I know the difference between them?’ It’s really like wrapping yourself around a conundrum, but after a while it was easy. Pretend I am talking to Liz Cheval, ‘I have got a half unit on and I am three N up’ or ‘I have got a half N unit on and I am half N positive.’ I totally understand the difference between them. Don’t even have to think about it. This is burned into your brain.”
Pyramiding: “Adding to Winners”
Once they understood S1 and S2 entry and exit rules, once they understood “N” and units, Eckhardt then instructed the Turtles to pile profits back into winning trades. This maxing out of their big winners was part of what helped to create the Turtles’ fantastic positive expectancy, or “edge.”
For example, a market bought at a price breakout level of 100 could have additional units added as the market moved through price levels of 102, 104, and 108. Assume a long breakout entry at 100 with an “N” value of 5. Assume that you will add another unit each 1N move. A new unit will be added at 105, 110, etc. Turtles could pyramid a maximum of 5 units. They set their stops at V2N on the first day of trading and from that point forward, 2N stops were used. Then, once the second unit was bought, both stops were brought up to the new unit’s 2N stop. As new units were added, all stops were brought up to the stop of the newest unit added.
This process protected open profits, but not to the extent that it would jeopardize catching a very big trend. This thinking also aimed to guarantee that profits would be plowed back into those big unpredictable trends. This was how Dennis and Eckhardt taught the Turtles to “bet their left nut.”
Trading Your Own Account Tip #10:
If you want to make Turtle-like money, you will need to use leverage. The key is to always manage your leverage use and not let it get past your limits.
Sample Trade to Demonstrate Pyramiding
This sample trade illustrates how the Turtles pyramided their winning trades.
First Unit
Starting account size: $50,000.
Account risk of 2%, or $1,000 per signal.
Long signal generated in live cattle at 74.00.
1N value is 0.80, 1 point in live cattle is $400, so the dollar value of 1N is $320.
2N value is 1.60, dollar value of $640.
Contracts to trade: $1,000/640 = 1.56 rounded down to 1.0.
Add the next position at 1N, or 74.00 + 0.80 = 74.80.
Stop setting is 74.00 - 1.60 = 72.40.
Table 5.10: Purchase First Unit of Live Cattle at $74.00.
Addition of Second Unit
Account value is now $50,320 ($50,000 + unit one gain of $320).
Account risk of 2%, or $1,006.40.
Second position added at 74.80.
1N value remains 0.80, or $320.
2N value remains 1.60, or $640.
Contracts to trade $1,006.40/$640 = 1.57, rounded down to 1.0.
Add the next unit at 1N, or 74.80 + 0.80 = 75.60.
Stop setting on both positions is 74.80 - 1.60 = 73.20.
Table 5.11: Purchase Second Unit of Live Cattle at $74.80.
Addition of Third Unit
Account value is now $50,960 ($50,000 + unit one gain of $640 + unit two gain of $320).
Account risk of 2% or $1,019.20.
Third position added at 75.60.
1N value decreased to 0.70, or $280.
2N value decreased to 1.40, or $560.
Contracts to trade $1,019.20/$560 = 1.82 rounded down to 1.0.
Add the next unit at 1N, or 75.60 + 0.70 = 76.30.
Stop setting on all units is 75.60 - 1.40 = 74.20.
Table 5.12: Purchase Third Unit of Live Cattle at $75.60.
Addition of Fourth Unit
Account value is now $51,800 ($50,000 + unit one gain of $920 + unit two gain of $600 + unit three gain of $280).
Account risk of 2%, or $1,036.00.
Fourth unit added at 76.30.
1N value remained 0.70, or $280.
2N value remained at 1.40, or $560.
Contracts to trade $1,036.00/$560 = 1.85 rounded down to 1.00.
Add the next unit at 1N, or 76.30 + 0.70 = 77.00.
Stop setting on all units is 76.30 – 1.40 = 74.90.
Table 5.13: Purchase Fourth Unit of Live Cattle at $76.30.
Addition of Fifth and Final Unit
Account value is now $52,920 ($50,000 + unit one gain of $1,200 + unit two gain of $880 + unit thr
ee gain of $560 + unit four gain of $280).
Account risk of 2%, or $1,058.40.
Fourth unit added at 77.00.
1N value increased to 0.85, or $340.00.
2N value increased to 1.70, or $680.00.
Contracts to trade $1,058.40/$680 = 1.55, rounded down to 1.00.
Stop setting on all units is 77.00 – 1.70 = 75.30.
Table 5.14: Purchase Fifth Unit of Live Cattle at $77.00.
Turtle stops were adjusted to break even with each 1N market move up.
Position Exit
Live cattle rallies to 84.50 and their exit criteria are met.
Table 5.15: Exit Live Cattle at $84.50.
With this kind of pyramiding, you could have a $300,000 account that was long five units containing Canadian dollars, U.S. dollar indexes, the S&P 500 index, unleaded gas, orange juice, yen, Swiss francs, gold, soybean oil, and cotton. By Turtle trading logic, they would be net long one unit, or 2 percent of the total portfolio.18
Risk of Ruin: “Will You Live or Die?”
Aggressive pyramiding of more and more units had a downside. If no big trend materialized, then those little losses from false breakouts would eat away even faster at the Turtles’ limited capital. How did Eckhardt teach the Turtles to handle losing streaks and protect capital? They cut back their unit sizes dramatically. When markets turned around, this preventive behavior of reducing units increased the likelihood of a quick recovery, getting back to making big money again.