The Turtle Traders and Early Trend Following

Richard Dennis had turned a small stake into tens of millions of dollars trading commodity futures, and he believed his edge came from a learnable, rule-based process, not innate market intuition. His trading partner, William Eckhardt, disagreed — he thought successful trading required a talent that couldn't simply be taught. They settled it with an experiment: in December 1983, Dennis recruited around twenty novices, many with no trading background at all, trained them for two weeks in a specific mechanical trading system, gave them real capital, and let them trade. They became known as the Turtles.

The system they were taught was a breakout strategy, a specific, disciplined form of trend following, and several Turtles went on to run successful trading firms for decades using variations of it — reasonably strong evidence, as far as informal experiments go, that Dennis was right.

The core rules

The system centered on channel breakouts: buy a market when its price breaks above the highest price of the last 20 trading days (a sign of a new uptrend starting), and sell short when it breaks below the lowest price of the last 20 days. Positions were sized not as a fixed dollar amount, but scaled inversely to each market's recent volatility (measured via a rolling average of the daily trading range, called "N" in Turtle terminology) — a more volatile market got a smaller position, so that a fixed dollar amount of expected daily swing was roughly equal across every position in the book, regardless of whether it was a wild commodity or a calmer one. Losses were cut with a stop roughly two volatility units against the entry; winners were allowed to run, with the position often held until an opposite breakout signaled the trend had reversed.

Worked example

Suppose a commodity has a 20-day high of $52 and the recent average daily range (N) is $1. A Turtle-style trader buys when price breaks above $52. Position size is set so the total position, if it moves by N in the trader's favor, changes the account by roughly a fixed fraction (commonly 1%) of total capital — meaning a market with a wider daily range (bigger N) automatically gets a smaller position in contracts than a calmer one, keeping risk roughly consistent market to market. If price falls back and hits a stop set at two N below entry ($50), the position is closed at a small, planned loss. If instead the trend continues and price runs to $65 before finally breaking below its own 10-day low, the position is closed there, capturing most of the intervening $13 move — the strategy's asymmetry comes from cutting the frequent small losses quickly while letting the occasional large trend run uninterrupted.

What this means in practice

The Turtle experiment is often cited as the founding case study for the modern managed-futures and CTA (commodity trading advisor) industry: an entire category of systematic, trend-following funds trace their intellectual lineage to variations on these breakout-and-volatility-sizing rules. It also popularized position sizing by volatility rather than by fixed dollar or contract amount, a technique now standard across nearly all systematic strategies, not just trend following.