In 1994 I learned a trading system for trading in the futures markets known as the Turtle Trading System. I immediately put it to use and lucked into hitting a homerun trade in the Coffee market.
Unfortunately, I learned the wrong lessons from that trade, and by 1995 I had abandoned the system.
However, that experience did lead me down a path of knowledge that most people don’t have the opportunity to explore.
Furthermore, while the system has experienced substantial “decay” over the years, and has generally lost its effectiveness, I believe it provides a solid foundation upon which anyone who has an interest in trading can develop their own trading strategy.
Richard Dennis and The Turtles
If you’ve never read the story of the Turtles, you can learn quite a bit about it from Michael Covel in his book The Complete TurtleTrader: How 23 Novice Investors Became Overnight Millionaires.
The short version is that Richard Dennis, whom I had read about in Jack Schwager’s classic Market Wizards had trained a group of individuals how to trade in the futures markets.
Dennis nicknamed the group the “Turtles” after a visit to the Far East where he observed a barrel full of baby turtles. He exclaimed that they were going to grow traders like they grow turtles.
The training program was the result of a bet he made with trading partner William Eckhardt, who didn’t believe people could be trained to trade successfully.
Dennis and Eckhardt presented the Turtles with a set of trading rules designed to capture sizable intermediate term trends that occur in the futures markets from time to time.
Dennis himself had built up a fortune estimated as $200 million from modest means, and was a legend in the business. He essentially taught the group about everything he knew about trading, and staked them all with a $100,000 account.
After a period of time, the traders were evaluated and a few were given significantly more equity to trade. As a group, these traders allegedly produced a total of $175 million in profits to Dennis, and a number of them went on to manage money as Commodity Trading Advisors.
One of the initial traders left the program after a dispute with Dennis. His name was Russell Sands, and it was Sands who then went on to begin teaching the Turtle Trading System in 1993 after a five year moratorium on the system was lifted.
I learned the system from Sands.
Basic Concepts of the Turtle Trading System
The Turtle Systems employed price channel breakouts as a means of entering and exiting positions in the futures markets. You can read more about the mechanical rules of the trading system in former Turtle Curtis Faith’s book… Way of the Turtle
Much of the discussion that follows is derived from Faith’s book and my own knowledge of the system.
Turtle Trading System I
Turtle System I employed a channel breakout entry of 4 weeks, or 20 trading days. In other words, if the market traded within a trading range, once it made a new 20 day high or low in price, a new position would be entered, if the trader did not currently have a position in the market.
P/L Filter (Last Trade Loss)
In order to eliminate some false breakouts and limit trading losses, the traders employed the P/L filter.
The P/L filter involved eliminating trades if the last 20 day breakout in either direction was a winning trade.
So, the trader would not enter a new position if the previous 20 day breakout resulted in a profit. The premise here is that markets tend to consolidate after a trending move.
Once a new position was in place, the Turtles would exit that new position if it immediately turned against them.
The stop loss was calculated as 2 x Average True Range of the last 15 to 20 trading days. Average True Range was also called “N.” The stop loss price was calculated as the entry price – 2N for a long position and + 2N for a short position.
Example: A new trade is entered at $100. The 20 day average true range, or N, is $2. $2 x 2 = $4. The stop loss for a long position is $96 ($100 – $4) and for a short position it is $104 ($100 + $4).
If the trade moved in the right direction, the Turtles would trail a stop at the 10 day low for a long position or the 10 day high for a short position. These were basically just mental stops as the Turtles were trading with too much money to place actual resting stop orders.
This a sequence of trades in the Japanese Yen in 2012-2013 and is reflective of the type of move the Turtle Trading System was designed to exploit.
In September 2012 is a false breakout to the upside. The trader would be stopped out just a couple days later with a 2N stop loss. A breakout to the downside follows in October, but this trade is stopped out at 10 days against in November.
A new breakout to the downside follows in November, and this time the trade is held until late February. As you can see, even when a market is trending strongly, a shorter term system will have of unprofitable trades.
Turtle Trading System II
Turtle System 2 was a simpler breakout strategy as it did not employ any type of filter. The basic trade entry occurred when a market made a new 55 day high, or a new 55 day low in price. The same 2N stop loss as in System I would be applied. The trailing stop was a 20 day low for a long position and a 20 day high for a short position.
Turtle Trading System III
System 3 is a hybrid between Systems 1 and 2. The longer term breakout system would be employed if no position has been taken with the shorter term system. In other words, you would monitor the markets for trades with System 1. However, if the P/L filter indicates you should skip a trade, you would wait for a 55 day breakout. This is called the “Failsafe” trade to ensure participating in every trend.
Turtle Trading System Money Management
The Turtles employed a risk management algorithm that incorporated the concept of N along with account size, and the concept of risk of ruin discussed earlier. N is the 15-20 day moving average of the true range, known commonly as Average True Range.
True Range is computed daily as follows…
True range = maximum(H-L,H-PC,PC-L) where H= High, L=Low, and PC = previous close
The first step in determining position size was to determine the dollar volatility represented by the market’s price volatility. Dollar Volatility is simply N x dollars per point.
Turtles built positions in chunks called units. 1N represented 1% of account equity. Therefore, the unit size for a market is calculated as … unit size = 1% of account/market dollar volatility
Example Position Size:
Account Size = $100,000
Corn…A tick move in Corn = $12.50, and a full point, or one cent = $50
20 day average true range = 5 cents = $250
1% of account equity = $1,000
Therefore, 1 full unit, or N = 4 contracts in the Corn market in this example ($1,000/$250 = 4 contracts)
Since the Turtle Trading System was meant to be traded across a diversified portfolio, this algorithm enables to the trader to normalize volatility and provide the opportunity to have the same chance for a particular dollar gain or loss in each market. This increased effectiveness of diversification.
The Turtles employed “N” as a risk management tool. No more than 4 units, could be held in a single market.
No more than 6 units could be held in a group of closely correlated markets, such as currencies, or interest rates.
No more than 10 units could be held in loosely correlated markets such as coffee, sugar, cocoa.
And no more than 12 units could be held in a single direction, long or short. Nowadays, trading advisors and hedge funds employ much more complex risk management algorithms, but 30 years ago, this was cutting edge strategy.
As indicated, the Turtles would consider themselves “fully loaded” in a market if they held a total of 4 units in that market. They would add units at 1/2 N intervals after the initial entry.
For example, if trading Crude Oil, and the initial price is $100, and N=$2.00, new positions would be added at $101, $102, and $103, as 1/2N = $1.00.
As units are added, the initial stop losses are also raised.
For example, if the initial entry is $100, 2N = $4, and therefore the initial stop is $96. This stop would be raised by 1/2N to $97 as a new position is added, and so forth to $98 and $99.
The Turtle Trading System Today
The heyday for the Turtle Trading System is long gone. The system performed well in the 1980’s, a time period when inflation was a bit higher, and when traders in the futures markets were less sophisticated.
The success that some of the Turtles had managing money led to significant growth in the industry, along with smarter traders. Furthermore, due to the continued decline in inflation, there have been fewer trends to exploit, particularly in the currency markets.
As a result, the performance of the system itself, and the Turtles themselves have declined significantly in recent years.
One of the Turtles closed his shop completely in 2013 after experiencing his third consecutive losing year.
The most successful of the Turtles, R. Jerry Parker, has seen his assets under management decline from a high of over $2 billion in the last ten years, to $200 million today.
Over the last twenty years I’ve studied and researched a variety of trading strategies for the futures markets, along with the performance of many Commodity Trading Advisors. I’ve generally concluded that the futures markets should generally be avoided unless you have substantial research and programming capabilities.
The fact of the matter is that all of the top trading firms in the business now hire the best and brightest out of the Ivy League and equivalent schools to develop quantitative trading strategies, and they pay them a lot of money to do that.
The Turtle Trading System And Stocks
Applying the Turtle Trading System mechanically to the stock market is a difficult proposition due to the sheer number of stocks in the market. However, some aspects of the strategy can definitely be applied to trading stocks.
First and foremost, the system involves a trend following approach to trading. In my view, this type of approach is well suited to trading high momentum, smaller cap stocks. These stocks, once they get rolling, can act like a commodity market, and therefore, the trends can be explosive.
Second, the basic money management strategy with the Turtle Trading System is also well suited for trading these high momentum stocks.
Finally, what I have not mentioned so far is that along with the mechanical system rules provided by Dennis and Eckhardt, the also provided a list of discretionary criteria to help identify the higher probability trades. This is somewhat similar to other trend following approaches to trading stocks, such as CANSLIM and the methodology employed by Nicolas Darvas.
With all this in mind, the Turtle Trading System provides the foundation upon which I’ve developed my core trading strategy for high momentum stocks.