Moving average: In finance, and especially in technical analysis, one of a family of similar statistical techniques used to analyze time series data. A moving average series can be calculated for any time series, but is most often applied to stock prices, returns, or trading volumes. Moving averages are used to smooth out short-term fluctuations, thus highlighting longer-term trends or cycles.
5
The Rules
“We have a pretty strict definition of a systematic trader. They basically follow a set series of rules, established in a computer program, that tell you when to buy or sell, how many, as well as when to get out.”
Michael Garfinkle,
Commodities Corporation
While the rules taught by Dennis and Eckhardt were not meant as a statistics class, the Turtles did learn some basic statistics including two “errors”:
A Type I error, also known as an error of the first kind or a false negative, is the error of rejecting something that should have been accepted.
A Type II error, also known as an error of the second kind or a false positive, is the error of accepting something that should have been rejected.1
If the Turtles made those errors on a regular basis, they would be finished with mathematical certainty. Said another way, they learned that it was better to risk taking many small losses than to risk missing one large profit. The concept of statistical errors was an admission that acknowledged ignorance could be quite beneficial in trading.2
At the root of Dennis and Eckhardt’s statistical thinking was Occam’s razor (a principle attributed to the fourteenth-century English logician William of Ockham).3 In more contemporary jargon people express it as, “Keep it simple, stupid!” For Dennis and Eckhardt’s rules to work, to have some statistical reliability, they had to be simple.
Expectation: How Much Does Your Trading Method Earn in the Long Run?
“What can you expect to earn on each trade on average over the long run from your investing decisions or your trading rules?” Or, as a blackjack player would say, “What is your edge?” A first step for the Turtles was to know their edge.
A good analogy is being a batter at the plate in a baseball game, as trades and success rates aren’t much different from batters and their averages. Dennis expanded on this: “The average batter hits maybe .280 and the average system might be successful 35 percent of the time.”4
More importantly what kind of hits did you get in hitting .280. Did you hit singles or home runs? In trading, the higher the expectation, the more you can earn. A trading system with an expectation of $250 per trade will make you more money than a system with a $100 per trade expectation (all other things being equal in the long run). The Turtle rules themselves had a positive expectation per trade because their winning trades were many multiples larger than their losing trades. Expectation (or edge, or expected value) is calculated with a straightforward formula:
E = (PW X AW) - (PL X AL)
Where:
E = Expectation or Edge
PW = Winning Percent
AW = Average Winner
PL = Losing Percent
AL = Average Loser
For example, assume a trading system has 50 percent winning trades. Now, assume the average winning trade is $500 and the average losing trade is $350. What is the “edge” for that trading system?
Edge = (PW X AW) - (PL X AL)
Edge = (.50 X 500) - (.50 X 350)
Edge = 250 - 175
Edge = $75 on average per gain per trade
Over time you would expect to earn $75 for each trade placed. For comparison, another trading system might be only 40 percent accurate with an average winner of $1,000 and an average loser of $350. How would that system compare to the first one?
E = (PW X AW) - (PL X AL)
E = (.40 X 1,000) - (.60 X 350)
E = 400 - 210
E = $190 on average per gain per trade
The second trading system’s “edge” is 2.5 times that of the first even though it has a much lower winning percent. In fact, the second system breaks even with a winning percent of 25.9. The first system breaks even at 41.1 percent. Clearly, when you hear the media and talking heads talking about “90 percent winning trades,” that talk is misleading. Percent accuracy means nothing.
Look at it this way. Think about Las Vegas. A small edge keeps casinos in business. That’s how those monster hotels in Las Vegas and Macau are paid for—by exploiting the edges. Dennis always wanted his trading to resemble being the house.
It didn’t necessarily matter how little the Turtles lost on any individual trade, but they needed to know how much they could lose in their whole portfolio. Eckhardt was clear: “The important thing is to limit portfolio risk. The trades will take care of themselves.”5
Trading Your Own Account Tip #1:
You need to calculate your edge for every trading decision you make, because you can’t make “bets” if you don’t know your edge. It’s not about the frequency of how correct you are; it’s about the magnitude of how correct you are.
Comparing the expected values of various Turtle and Turtle-style trading money management firms to various stock indexes gives more perspective about the importance of expectation:
Table 5.1: Turtle Trader Expectations from Inception to August 2006.
Table 5.2: Stock Index Expectations from Inception to August 2006.
The expectation generated by the trend traders generally beats the monthly expectation of buying and holding market indexes. Why? The average winning months of Turtle traders is much larger than their average losing months.
Entries and Exits: “It’s Always Better to Buy Rallies”
Everyone wants to know, “How do you know when to buy?” The Turtles were taught to enter trades via “breakouts.” A breakout occurs when a market—any market (Cisco, gold, yen, etc.)—“breaks through” a recent high or low. If a stock or futures contract made a fifty-five-day breakout to the upside (long), meaning that its current price was the highest price of the last fifty-five days, Turtles would buy.
If a stock made a fifty-five-day breakout to the downside, meaning that its current price was the lowest price of the last fifty-five days, Turtles would sell short, aiming to profit as the market dropped. In isolation there was nothing special about these simple rules for entry. Philosophically, Turtles wanted to buy a market going up (becoming more expensive) and wanted to sell short a market dropping in price (becoming cheaper).
What about the standard Wall Street refrain of “buy low and sell high”? The Turtles did just the opposite! And unlike most people’s understanding of the markets, pro or beginner, the Turtles actively aimed to make money by “shorting” declining markets. They had no bias to being long or short.
While breakouts were the reason to enter, those breakouts did not mean a trend would continue by any measure. The idea was to let price movement lead the way, knowing at any time the price could change and go in a different direction. If a market went sideways, back and forth, you could see how and why Turtle price breakouts produced many small losses while they waited for a price breakout that might produce the big trend.
No matter what price is the variable that the great traders have lived and died by for decades. Making trading decisions more complicated than the simple heuristic of “price” has always been problematic. Eckhardt knew it was hard to do much better: “Pure price systems are close enough to the North Pole that any departure tends to bring you farther south.”6
Trading Your Own Account Tip #2:
Now that the concept of using price for your decision-making is clear, stop watching TV! Stop looking at financial news. Start keeping track of the open, high, low, and close of each market you are tracking. That is the key data you need to make all of your trading decisions.
Trading Your Own Account Tip #3:
You need to be able to wrap your arms around the concept of “shorting” a market. Or said another way, you have to relish the opportunity to
make money in a decreasing market. Shorting was never unique to the Turtles; they just did it effectively.
System One and System Two: The Two Turtle Systems
The Turtles learned two breakout variants or “systems.” System One (S1) used a four-week price breakout for entry and a two-week price breakout in the opposite direction of the entry breakout for an exit. If a market made a new four-week high, the Turtles would buy. They would exit if/when it made a two-week low. A two-week low was a ten-day breakout—counting trading days only.
While the System One entry rule is straightforward, the Turtles were taught extra rules to confirm whether or not they should take the four-week breakout. The extra rules were called “filters,” and they were designed to increase the odds that when the Turtles took a four-week breakout signal, it would continue as a potentially big trend.
The filter rule: The Turtles ignored the System One four-week breakout signal if the last four-week breakout signal was a winner. Even if they did not take the last four-week breakout signal, or even if it was just “theoretically” a winning trade, the Turtles still didn’t take the System One breakout. However, if the trade before a current four-week breakout was a 2N loss, they could take the breakout (“N” was simply their measure of volatility, discussed in the next section).
Additionally, the direction of the System One four-week breakout was irrelevant in terms of the filter rule. If their last trade was a short losing trade and a new long or short breakout hit, they could take that breakout and get in.
But this filter rule had a built-in problem. What if the Turtles skipped the entry breakout (since the last trade was a “winner”) and that skipped breakout was the beginning of a hugely profitable trend that roared up or down? Not good to be on the sidelines with a market taking off!
If the Turtles skipped a System One four-week breakout and the market kept trending, they could and would get back in at the System Two eleven-week breakout (see below). This fail-safe System Two breakout was how the Turtles kept from missing big trends that were filtered out.
System Two was the Turtles’ longer-term trading system. It used an eleven-week breakout (fifty-five days) for an entry signal and a four-week breakout (twenty days) in the opposite direction for an exit.
Trading Your Own Account Tip #4:
The price “breakout” was Turtle jargon to describe a market that had just made a new high or new low over “x” period. Do traders use other values beyond twenty and fifty-five days for entry? Yes. The selection of these values for your trading will always be subjective. Test or practice these rules on paper and/or trading software (such as wealth-lab.com and mechanicasoftware.com) so you can see the ups and downs and gain confidence. The Turtles typically put half of their money toward each system.
Each Turtle had discretion over which of the two systems, System One (S1) and System Two (S2), Dennis and Eckhardt gave them to use. Mike Carr combined S1 and S2, allowing for more entry and exit points. He was trying to smooth out his trading results.
Jeff Gordon preferred S1, but mixed in S2 for smoother returns. Gordon, like some other Turtles, traded System Three (S3). He said the systems were Dennis’s attempt to teach the Turtles that they should follow his methodology and not venture off the reservation, so to speak. Gordon added, “It was do anything you want to do, but don’t lose more than $50,000 doing it. Once you crossed $50,000 and one dollar, you were out of the Turtle program.”
Dennis called System Three (S3) the dreaded flair account. Erle Keefer saw why S3 was not taking hold, adding, “You could do whatever it was and everybody did it to a small degree, but within about six weeks everybody just canned that account.”
It was canned because the Turtles had already lived emotionally losing seven out of ten trades. They knew that was the right thing to do. Keefer minced no words: “That was the only way you were ever going to hook the real trend. We saw it work. I don’t know anybody that’s writing really good books called Counter Trend Trading to Win.”
Trading Your Own Account Tip #5:
Feel free to experiment on breakout lengths. Do not fixate on specific values. The key will be to accept a breakout value and stick with it consistently. Testing and practice are wise for confidence. Trust, but verify.
It is not surprising that over the years some traders—those who knew the Turtle rules—became obsessed with the specific System One and System Two entry and exit values as if they were the long-lost Holy Grail. Traders who fixate like that are looking at the tree instead of seeing the whole forest. For Turtle trading to work, the simplest of entry rules must continue to work. To get into debates about whether entering on a fifty-day breakout or a fifty-one-day breakout is better is misguided. In reality, a minor change of a variable in any robust trading system should not cause significant performance changes. If it does, you are in trouble.
Jerry Parker uses “robustness” as his guiding precept: “I think it’s important to stay fairly simple—not a lot of variables. I think the reason we make money? It’s the simple moving average systems. They need to continue to do well.”
Parker used the Mount Lucas Management Index to make the point. It is a trend-following index based on a fifty-two-week moving average that goes back to the 1960s. Parker knows that core concept is his edge:
Two-thirds of that is what drives our profits. Our little filters to get in early, to get out quicker, volatility filters, if that is how we’re going to essentially generate returns, we’re going to be in bad shape. The core simple moving average or breakout systems [are key]. I think making our parameters longer term is important, but the minute it takes too much thinking and too much analysis and too much fancy work, it is going to be … a very bad situation.7
The market ‘gurus’ who pretend that a complex approach must be used to make money miss Parker’s point about keeping it simple. They want the equivalent of quantum physics for trading rules. That kind of thinking is mental masturbation, or as trader Ed Seykota calls, it “math-turbation.”
Consider this September 1995 Japanese yen chart to illustrate the System One price breakout in action:
Chart 5.3: Turtle Entry Example Using Japanese Yen.
The market “broke out” to a new four-week high early in 1995. Then the market moved upward until a two-week breakout in the opposite direction in late April. The Turtles exited.
A great example of this process was seen early in Liz Cheval’s career. She bought 350 oil contracts for under $20 a barrel in July 1990 and hung on as prices rose above $40. On October 15, 1990, she started liquidating at $38 a barrel and completed her liquidation at just over $30 with an average exit price of $34.80.8
There was no discussion about OPEC, government reports, or other fundamental factors used in Cheval’s decision-making. It was all about the price action telling her to enter and exit. It is important to note that Turtles always exited after the market went against them, thus having to endure the painful experience of giving back profits. You can’t pick the bottom and you can’t pick the top, so trying to end up with the “middle” of a trend was the goal.
Chart 5.4: Turtle Long Entry Example Using Natural Gas.
Experiencing that trade made Cheval a believer. She said, “I remember giving back $4 million out of $8 million profit of Rich’s money in a few minutes in the silver market in 1987.” That lesson helped her to hang on when crude oil dropped from $30 to $25.9
Consider another example (chart 5.4), from November 2005 in Natural Gas Futures. Each dot represents a new fifty-five-day high. The first breakout happened in mid-July 2005. There was no way to know entering that breakout that a trend would continue higher, but it did, and the Turtles just followed along making money.
However, that breakout could just as easily have been a loser. In fact, Turtles could have had a string of losers in a row with multiple false breakouts. Phil Lu used to say, “When you have a losing trade, you say to yourself, ‘Hey, it seemed like the right thing to do at that time.’” Exactl
y, following the rules means there will be losses. Trend-following trader Larry Hite has long said, “There are four kinds of bets. There are good bets, bad bets, bets that you win and bets that you lose.”
Dealing with and handling losses is not easy. Jerry Parker has lived the struggles of taking losses to win in the long run. He advises:
I used to say we take a small loss, but I think it’s better to take an optimal loss. You don’t want to take one that’s too large and yet you don’t want to have your stops so close that you’re going to get bounced out. [Just] hang on to the trade. Don’t get too excited. If you’re not making very much money, that’s fine. If you got a little loss, that’s fine. If you make a decent profit that turns into a loss, that’s fine. Just hold onto it and then really get aggressive when you’ve been rewarded by a big, huge profit.10
December 2006 Eurodollars is yet another good example to illustrate the Turtle rules in action. However, this time the chart shows the opportunity to make money in a falling market. It was a “go short” opportunity. Each dot represents a new fifty-five-day low. The first “short” breakout occurred in February 2006 and the market continued to fall before reaching a low in June.
Chart 5.5: Turtle Short Entry and Exit Example Using Eurodollars.
The Complete TurtleTrader Page 9
