Richard Dennis is the Chicago commodities trader who settled an argument by proving that trading can be taught. In 1983 he recruited a group of novices, trained them for about two weeks in a mechanical trend-following system, funded them with his own money, and let the results speak. The group became known as the Turtles, and the experiment remains the closest thing the industry has to a controlled test of whether trading skill is innate or learnable. This article covers who Dennis is, what the Turtles were actually taught, what happened, and what the experiment does and does not prove. The Turtles are one of several stories worth knowing; our famous traders collection covers the rest.
Who Is Richard Dennis?
Richard Dennis built a substantial fortune trading commodities in Chicago from a famously small starting stake. He borrowed $1,600 from his family, spent $1,200 of it on a seat at the MidAmerica Commodity Exchange, and started trading with the $400 that was left. By the early 1980s he was one of the best-known futures traders in the world. Reported figures for what he went on to accumulate vary considerably across sources, which is worth noting up front: the honest summary is a very large fortune built from a very small base, in an era of powerful commodity trends.
His trading partner, the mathematician William Eckhardt, took the opposite view, and their long-running disagreement was philosophical. Dennis believed successful trading was a teachable skill that could be reduced to rules. Eckhardt believed it was closer to an innate aptitude. Neither could win the argument in conversation, so they designed an experiment.
The Bet That Created the Turtles
Dennis and Eckhardt advertised for trading apprentices in the financial press, including The Wall Street Journal, received applications in the thousands, and selected a small group to train. The first intake was in 1983 with a second class the following year, and the total across both classes was somewhere in the low twenties. Dennis named them Turtles after visiting a turtle farm in Singapore, on the idea that traders could be grown quickly and systematically.
Training lasted about two weeks. The Turtles were then given accounts funded with Dennis’s own capital, commonly reported at around a million dollars each although allocations varied, and told to trade the rules exactly as taught. That last instruction is the part most retellings underplay: the experiment was not about whether novices could think like Dennis, it was about whether they could execute a defined system without improvising.
What the Turtles Were Actually Taught
The system was a complete mechanical trend-following method covering every decision, which is precisely why it could be taught in two weeks. The Donchian breakout rules at its core were the same ones Ed Seykota had put on a computer a decade earlier.
| Component | What the Turtles used |
|---|---|
| Markets | Liquid futures traded on US exchanges |
| Entries | Donchian channel breakouts, with a shorter 20-day system and a longer 55-day system |
| Stops | Volatility-based, defined in advance and expressed in units of average true range |
| Position sizing | A constant percentage of account risk per position, scaled by each market’s volatility |
| Adding to positions | Pyramiding into trades already moving in their favour |
| Exits | Breakout-based exits on winning positions, defined before entry |
Two features deserve emphasis because they are what actually made it work. First, position sizing was volatility-adjusted, so a percentage of risk meant the same thing in a violent market as in a quiet one. This is the ancestor of most modern position sizing methods. Second, exits were specified in advance for both winners and losers, which is what allowed the Turtles to hold large trends without negotiating with themselves.
What Actually Happened
The Turtles were, collectively, a success, and the experiment is generally regarded as having settled the argument in Dennis’s favour. Commonly cited figures put the group’s aggregate profits somewhere between roughly $100 million and $175 million over about four to five years, though sources differ noticeably on both the total and the period, and the numbers should be read as approximate rather than precise.
The more informative detail is the dispersion. The Turtles were taught identical rules at the same time, and their individual results still varied widely. Some followed the system closely and did well. Some deviated, hesitated on entries, or cut winners early, and did materially worse.
One complication belongs with that observation. Dennis did not allocate capital equally, and accounts were resized over time based on how he judged each trader was performing, which created visible tension in the group. So the spread in absolute dollars reflects allocation as well as execution, and the experiment was less controlled than the retelling usually suggests. What survives the complication is the part that matters: every Turtle received the same rules, and what they did with them diverged sharply. The rules were necessary and not sufficient, and the differentiator was the discipline to follow them.
It is also fair to record what came after. Dennis’s own subsequent trading included severe losses in the late 1980s that led him to stop managing money for a period. The lesson available in a set of rules is not a guarantee attached to the person who wrote them, and trend following has long stretches where it does not work.
What the Experiment Proved, and What It Did Not
It proved that a complete, mechanical trading method can be transferred to people with no background and produce profitable results, which was a genuinely open question in 1983 and is the foundation of the systematic trading education industry that followed. It also proved that a system specified down to position sizing and exits is teachable in a way that judgment is not.
It did not prove that the Turtle rules still work in current markets. Those specific parameters have been public for decades, the futures markets of the 1980s carried powerful commodity trends that are not permanently available, and a published edge attracts participants. Nor did it prove that anyone can trade successfully; the spread of results among the Turtles shows that following rules under pressure is itself a skill, and not all of them managed it. Where the Turtles proved a method could be taught, Jim Simons showed how far a systematic edge could be pushed when enough research sits behind it.
What Modern Traders Take From the Turtles
Three things transfer cleanly. Define the complete system, including exits and sizing, before risking money. Size positions by volatility so that risk is consistent across markets and conditions. And accept that a trend-following method will lose more often than it wins, which means the drawdowns are the price of the large winners rather than a sign of failure.
What does not transfer is the parameters. Copying 20-day and 55-day breakouts into a modern stock portfolio without testing them is exactly the mistake the Turtles were trained not to make. Test the principles against your own markets and build your own trading systems on the results, which is the approach we teach in the Trader Success System. If the underlying style appeals to you, our trend trading guide covers the mechanics in current markets.
Frequently Asked Questions
Who were the Turtle Traders?
A group of novices recruited and trained by Richard Dennis in 1983 and 1984 to trade a mechanical trend-following system with his own capital, as an experiment to test whether trading could be taught. They were named the Turtles by Dennis, and the group traded profitably overall.
What were the Turtle trading rules?
A complete trend-following system: Donchian channel breakout entries on 20-day and 55-day windows, volatility-based stops expressed in average true range, position sizing set as a constant percentage of account risk adjusted for volatility, pyramiding into winning positions, and predefined breakout exits.
Do the Turtle trading rules still work?
The principles hold up; the specific parameters are another matter. The original rules have been publicly available for decades, and markets have changed considerably since the 1980s. Anyone considering them should backtest them on current data in the markets they intend to trade rather than assuming the historical results carry over.
What happened to Richard Dennis?
He continued trading and managing money after the experiment, suffered serious losses in the late 1980s that led him to step back from managing outside capital, and returned to trading afterwards. In later years he has been known at least as much for his political and philanthropic activity as for markets. His career is a useful reminder that a successful method and a permanently successful career are different things.
