True range: capturing a day's real movement
Start with a single day. The obvious measure of movement is the distance from the day's high to its low, but that misses gaps: a market that closes at 100 and opens the next morning at 105 has moved, even if it then trades in a tight range. True range fixes this by taking the largest of three numbers:
- Today's high minus today's low
- The distance from yesterday's close up to today's high
- The distance from yesterday's close down to today's low
Average True Range (ATR) smooths that daily figure over a window, classically 14 or 20 days. The Turtles used a 20-day exponential average and called the result N. If gold's N is $18, gold has typically been moving about $18 a day.
From volatility to position size: the unit
The Turtles' insight was to define position size so that one N of movement equals 1% of the account no matter what the market is. They called that quantity a unit:
A worked example with a $100,000 account, so 1% is $1,000:
- A quiet market: corn has an N of 5 cents, and one futures contract is worth $50 per cent. One N of movement is $250 per contract, so a unit is $1,000 ÷ $250 = 4 contracts.
- A wild market: crude oil has an N of $2.00, and one contract is worth $1,000 per dollar. One N of movement is $2,000 per contract, so a unit is $1,000 ÷ $2,000 = half a contract, meaning you trade the smallest size available or skip it.
Same account, same formula, wildly different position sizes. That is the point. Both positions now move about $1,000 on a typical day, so the portfolio's risk is spread evenly instead of being quietly concentrated in whichever market happens to be most volatile.
The 2% rule
Sizing connected directly to the stop. Every Turtle position carried a stop 2N from entry, and since one N was sized to 1% of the account, a stopped-out unit lost about 2%. That number was chosen to make losing streaks survivable. Ten consecutive losers, which any breakout system will eventually produce, drew the account down roughly 18%, painful but recoverable. The same streak risking 10% per trade is a 65% hole that most traders, and most systems, never climb out of.
Van Tharp's Trade Your Way to Financial Freedom generalizes this idea into expectancy and "R-multiples," and it remains the standard reference for thinking about sizing independently of any particular entry signal.
Why volatility sizing outlived the system
The 20-day and 55-day breakouts from the Donchian channel entries have been public for decades and their edge has faded in many markets. The sizing idea has not. Practically every systematic futures fund today, and most risk-parity and volatility-targeting strategies in the equity world, run on the same core concept the Turtles learned in 1983: measure each market's volatility, then size positions inversely to it. It is arguably the experiment's most durable technical legacy, and it is explained from the inside in Way of the Turtle.
Adapting it beyond futures
The formula ports cleanly to stocks and ETFs: shares = (account × risk%) ÷ (ATR × dollars per share per point, which for a stock is just 1). A $50,000 account risking 0.5% per ATR on a stock with an ATR of $3 takes $250 ÷ $3 ≈ 83 shares. The details, and the caveats that come with losing futures leverage and diversification, are covered in our guide to Turtle trading on stocks and ETFs.
See sizing inside the full system
N also set the stops, the pyramids, and the portfolio limits. It all connects.