Understanding Volatility-Based Position Sizing
Volatility-based position sizing is the cornerstone of professional trend following risk management. Unlike fixed-dollar position sizing, which risks a constant amount per trade regardless of market conditions, volatility-based sizing adjusts the number of contracts or shares based on the Average True Range (ATR) of the instrument. The ATR measures the average range between high and low prices over a specified period, typically 14 days. By dividing a fixed risk amount by the ATR, traders ensure that each position carries a similar risk profile. For example, if a trader risks $1,000 per trade and the ATR is $5, they buy 200 shares. If the ATR doubles to $10, they buy only 100 shares. This dynamic adjustment prevents large positions in volatile markets and small positions in quiet markets, creating a consistent risk exposure across all trades. The formula is: Position Size = (Account Equity × Risk Percentage) / (ATR × Point Value). This method also accounts for gaps and slippage by using a multiple of ATR, such as 2× ATR for stop distance. Without volatility-based sizing, a single volatile trade can wipe out weeks of profits.
The ATR Channel Stop and Chandelier Exit
The Chandelier Exit, developed by Chuck LeBeau, is a trailing stop that hangs from the highest high since entry, minus a multiple of ATR. For long positions, the stop is placed at Highest High – (ATR × Multiplier), typically 3× ATR. For short positions, it is Lowest Low + (ATR × Multiplier). This stop moves only in the direction of the trend, never against it, locking in profits as the trend extends. The ATR channel stop is similar but uses a moving average of highs and lows instead of the absolute highest high. Both methods adapt to changing volatility: when volatility rises, the stop widens to avoid premature exits; when volatility falls, the stop tightens to protect gains. A common mistake is using a fixed dollar stop, which becomes too tight in volatile markets and too loose in calm ones. The Chandelier Exit works best on daily and weekly charts, with backtests showing it outperforms simple percentage stops in trending markets like commodities and currencies. Traders should test multipliers of 2.5, 3, and 3.5 ATR to find the optimal balance between whipsaw avoidance and profit capture.
The Turtle Traders’ 2% and 6% Rules
The original Turtle Traders, taught by Richard Dennis and William Eckhardt, used two strict position sizing rules: the 2% Rule and the 6% Rule. The 2% Rule states that no single trade can risk more than 2% of total account equity. The 6% Rule states that total open risk across all positions cannot exceed 6% of equity. For example, with a $100,000 account, each trade risks $2,000, and you can hold a maximum of three such positions before hitting the 6% cap. This prevents correlated positions from causing a catastrophic drawdown. The Turtles also used a unit size based on 1% of equity divided by the dollar value of ATR. If one unit risked 1%, they would add units as the trend progressed, but never exceed 4 units per market and 12 units total across all markets. The 6% rule acts as a circuit breaker: after three consecutive losses, trading stops for the week. These rules are still used by CTAs and hedge funds because they enforce discipline and prevent revenge trading. Without the 6% rule, a trader might hold five correlated long positions in crude oil, heating oil, and gasoline, only to see all five stopped out in a single sector crash.
Correlation-Adjusted Position Sizing
Correlation-adjusted position sizing reduces exposure to markets that move together. If you hold long positions in EUR/USD, GBP/USD, and AUD/USD, these pairs are highly correlated to the US dollar. A single dollar rally can stop out all three. To adjust, calculate the rolling correlation over 60 or 90 days. If two markets have a correlation above 0.7, treat them as one position for risk purposes. For instance, if you normally risk 1% per market, risk only 0.5% each for two correlated markets, or 0.33% for three. Alternatively, use a portfolio heat map: sum the risk of all open positions weighted by their pairwise correlations. If total correlated risk exceeds 8% of equity, reduce or skip new trades. Professional trend followers often use a correlation matrix updated weekly. They also avoid taking new positions in the same sector (e.g., energy, metals, grains) if existing positions already use 50% of the sector risk limit. This prevents a single macroeconomic event, like an OPEC decision or a Fed rate hike, from triggering multiple stop losses simultaneously.
The Role of Stop Loss Orders in Trend Following
Stop loss orders are not optional in trend following; they are the only defense against catastrophic loss. A stop loss order is an instruction to exit a position when price reaches a specified level. In trend following, stops are always placed at the time of entry, never after. The two main types are hard stops (resting orders in the market) and mental stops (trader monitors and exits manually). Hard stops are superior because they remove emotion and ensure execution even during fast markets. However, in illiquid markets or during gaps, hard stops can suffer slippage. For example, a stop at $50 might fill at $48 after a gap down. To mitigate, use stop-limit orders, but these risk not being filled at all. The best practice is to use hard stops with a buffer of 0.5× ATR beyond the technical level. For instance, if the Chandelier Exit is at $48.50, place the stop at $48.00. Never move a stop loss away from the entry price—only move it in the direction of the trend. Moving a stop to breakeven too early is a common error; it guarantees a small loss but also guarantees you miss large trends. Research by Kirk Northington shows that stops placed at 2× ATR from entry have a 65% chance of being hit before a 3× ATR profit target, but widening to 4× ATR reduces that to 35%.
Pyramiding and Stop Loss Adjustment
Pyramiding—adding to a winning position—is a hallmark of trend following, but it requires strict stop loss adjustment. The rule: when you add a new unit, move the stop for the entire position to a level that ensures the total trade risks no more than the original 1% or 2%. For example, you enter long at $100 with a stop at $95 (risk $5). Price rises to $110, and you add a second unit at $110 with a stop at $105. Now you have two units: first unit risks $5 profit locked ($100 entry, $105 stop = +$5), second unit risks $5 loss ($110 entry, $105 stop = -$5). Net risk is zero. As price rises further to $120, add a third unit at $120 with a stop at $115. Now first unit locked +$15, second unit locked +$5, third unit risks $5. Net locked profit +$15. This is called a “free trade” or “risk-free pyramid.” The stop for the entire position is always the most recent unit’s stop. Never widen the stop to accommodate a new unit. Also, limit pyramids to 3–4 units; beyond that, a reversal can erase all profits. The Turtle system used 0.5× ATR increments between units. Pyramiding works best in strong, persistent trends like those in commodities or long-duration bonds.
Equity Curve Trading and Risk Reduction
Equity curve trading is a meta-risk management technique where you reduce position size or stop trading when your account equity curve falls below its moving average. For example, plot a 20-day moving average of your account equity. If equity closes below that average, cut all position sizes by 50%. If it falls below a 50-day moving average, stop trading entirely for two weeks. This prevents you from digging a deeper hole during a drawdown. Trend following systems typically have win rates of 35–45%, meaning long losing streaks are normal. A 10-trade losing streak can reduce equity by 15–20%. Equity curve trading forces you to preserve capital when your system is out of sync with the market. Backtests show that adding a 20-day equity curve filter improves the Sharpe ratio by 0.3 to 0.5 for most trend following systems. However, it also reduces total returns slightly because you miss the first few trades of a new trend. The trade-off is worth it for most traders. Use a simple rule: if equity is below its 20-day MA, risk 0.5% per trade instead of 1%. If below its 50-day MA, risk 0.25%. Resume full risk after equity closes above both averages for three consecutive days.
The Kelly Criterion vs. Fixed Fractional
The Kelly Criterion calculates the optimal position size to maximize long-term growth based on win rate and win/loss ratio. The formula is: Kelly % = W – [(1 – W) / R], where W is win rate and R is average win divided by average loss. For a trend following system with 40% win rate and 3:1 reward-to-risk, Kelly = 0.40 – (0.60 / 3) = 0.20, or 20% of equity per trade. That is dangerously high for real trading. Most professionals use “half-Kelly” or “quarter-Kelly” to reduce drawdowns. Fixed fractional sizing, by contrast, risks a constant percentage (e.g., 1%) per trade. Fixed fractional is simpler and more robust because it does not rely on accurate estimates of win rate and payoff, which vary over time. A 1% fixed fractional risk with a 3× ATR stop is equivalent to a conservative Kelly for most trend following systems. Never use full Kelly: a 20% risk per trade can lead to a 50% drawdown after five consecutive losses, which is psychologically and financially devastating. Research by Ralph Vince shows that optimal f (the fraction that maximizes geometric growth) is often lower than Kelly due to parameter uncertainty. Stick to 0.5% to 2% per trade, and never exceed 2% even with a high-confidence signal.
Stop Loss Placement for Gaps and Overnight Risk
Gaps—price jumps between the close and next open—are the Achilles heel of stop losses. A stop at $50 may be skipped entirely if the market opens at $45. To manage gap risk, use a “stop with a limit” or a “market if touched” (MIT) order. But the best defense is position sizing: assume a worst-case gap of 1.5× ATR. If your stop is 2× ATR away, and a gap of 1.5× ATR occurs, your actual loss is 3.5× ATR, not 2×. So size positions as if the stop is 3.5× ATR. For overnight risk in futures, consider the exchange’s price limits. In limit-down markets, you cannot exit at any price. To avoid this, reduce position size in markets with daily price limits (e.g., grains, softs) by 50%. For stocks, avoid holding through earnings unless you accept gap risk. A common rule: if earnings are within 5 days, cut position size by half or exit entirely. For forex, weekend gaps are common; reduce size by 30% on Friday afternoons. Always check the economic calendar: avoid holding large positions through FOMC, NFP, or OPEC meetings. If you must hold, use options as a hedge—buy a put or call to cap downside.
Trailing Stop Techniques: Parabolic SAR and Moving Averages
Beyond ATR-based stops, trend followers use Parabolic SAR and moving average stops. Parabolic SAR (Stop and Reverse) accelerates toward price as the trend continues. It starts at 0.02 acceleration factor and increases by 0.02 each new extreme, maxing at 0.20. The stop flips when price crosses the SAR. Parabolic SAR works well in strong trends but whipsaws in sideways markets. Use it only after a clear breakout. A 50-day or 100-day simple moving average (SMA) stop is simpler: exit when price closes below the SMA. For long-term trends (months to years), a 200-day SMA is robust. For medium-term (weeks to months), a 50-day SMA. For short-term, a 20-day EMA. Combine them: exit half the position when price closes below the 20-day EMA, and the rest when it closes below the 50-day SMA. This scales out of trades, locking profits while staying in for the big move. Backtests on the S&P 500 from 1980 to 2020 show that a 10-month SMA (approximately 200-day) trend following system produced a 9.5% annual return with a maximum drawdown of 20%, versus buy-and-hold’s 10% return with a 50% drawdown. The SMA stop is not perfect—it gives back 10–15% of the trend—but it captures the majority of large moves.
Risk of Ruin and Monte Carlo Simulation
Risk of ruin is the probability of losing enough capital to be unable to continue trading. For a trend following system with 40% win rate and 1% risk per trade, the risk of a 20% drawdown over 100 trades is about 15%. The risk of a 50% drawdown is less than 1%. But if you risk 5% per trade, the risk of a 50% drawdown jumps to 35%. Monte Carlo simulation runs thousands of randomized trade sequences to estimate drawdowns. Use it to stress-test your position sizing. For example, shuffle your historical trade returns 10,000 times. Plot the distribution of maximum drawdowns. If the 95th percentile drawdown is 25%, you need to decide if you can psychologically survive that. If not, reduce risk per trade to 0.5%. Also simulate worst-case scenarios: 15 consecutive losses. With 1% risk, that is a 14% drawdown. With 2% risk, 26%. With 3% risk, 37%. Most professional CTAs cap risk at 1% per trade and 10% total portfolio heat. Retail traders often use 2% because of smaller accounts, but that is the absolute maximum. Never risk more than 2% on a single trade, no matter how confident you are. The market can remain irrational longer than you can remain solvent.
Psychological Discipline and Pre-Committed Rules
The best position sizing and stop loss plan fails without discipline. Pre-commit to your rules in writing before the market opens. Write down: (1) maximum risk per trade (e.g., 1%), (2) maximum total risk (e.g., 6%), (3) stop loss method (e.g., 3× ATR Chandelier), (4) pyramiding rules (e.g., add 0.5× ATR, move stop to breakeven after 2nd unit), (5) equity curve filter (e.g., cut size by half below 20-day MA). Then follow them mechanically. Do not move stops. Do not add to losers. Do not skip a trade because you “feel” it will fail. Do not take a trade because you “feel” it will win. Keep a trading journal: record entry, exit, stop, size, and emotion. Review weekly. After 100 trades, calculate your actual win rate, average win, average loss, and maximum drawdown. Compare to your plan. Adjust only after 200 trades, not after 10. The market pays for discipline over decades, not for brilliance over days. Trend following is simple but not easy. The simplicity is in the rules; the difficulty is in the execution. Position sizing and stop losses are your only edge against randomness. Treat them as sacred.
