04Advanced
Advanced Position Sizing
Move beyond fixed-lot sizing with the Kelly criterion, volatility-based methods, and systematic scaling strategies.
40 min4 sections
The Kelly Criterion for Optimal Bet Sizing

The Kelly criterion is a mathematical formula that determines the optimal fraction of capital to risk on each trade, maximizing the long-term geometric growth rate of the account. The formula is: Kelly % = W - (1 - W) / R, where W is the win rate (as a decimal) and R is the average win-to-loss ratio. For example, a system with a 55% win rate and a 1.5 reward-to-risk ratio yields a Kelly percentage of approximately 25%.
However, full Kelly sizing is extremely aggressive and produces large drawdowns that most traders cannot psychologically tolerate. A single losing streak can wipe out 40-50% of the account before the edge plays out. For this reason, professional traders typically use "fractional Kelly" -- risking one-half or one-quarter of the Kelly-optimal amount. Half-Kelly achieves about 75% of the growth rate with significantly less volatility, making it a popular choice.
To apply Kelly in practice, you need accurate statistics from at least 100 trades with your specific strategy. The formula assumes that your edge remains constant, which is rarely true in financial markets. Therefore, recalculate your Kelly fraction periodically as market conditions change, and always cap the maximum risk per trade at a level you are comfortable with, regardless of what the formula suggests.
Fixed Fractional & Percent-Risk Models

Fixed fractional position sizing is the most widely used method among professional traders. The trader risks a fixed percentage of their current account balance on every trade -- typically between 0.5% and 2%. As the account grows, position sizes increase proportionally; as it shrinks, they decrease. This automatic adjustment protects the account during drawdowns and accelerates growth during winning streaks.
To calculate lot size under the fixed fractional model: Risk Amount = Account Balance x Risk Percentage. Then, Position Size = Risk Amount / (Stop Loss in Pips x Pip Value). For example, with a $50,000 account risking 1% and a 30-pip stop on EUR/USD (pip value $10 per standard lot), the risk amount is $500 and the position size is $500 / ($10 x 30) = 1.67 standard lots.
A variation is the fixed ratio method, which ties position size increases to a fixed dollar gain ("delta") rather than a percentage. With a delta of $5,000, you increase by one lot for every $5,000 in profit. This method produces slower initial growth but faster compounding once the account reaches a certain size. Choose between fixed fractional and fixed ratio based on whether you prioritize consistent percentage risk or smoother absolute-dollar growth.
Volatility-Based Position Sizing

Volatility-based sizing adjusts position size according to how much a pair is currently moving, ensuring that each trade carries a similar dollar risk regardless of market conditions. The most common implementation uses the Average True Range (ATR). Instead of a fixed pip stop, the stop loss is set at a multiple of the ATR (for instance, 2x ATR), and the position size is calculated to risk the same percentage of capital.
This approach naturally results in smaller positions during high-volatility periods and larger positions when volatility is low. The benefit is that your stop loss adapts to current conditions, reducing the chance of being stopped out by normal market noise during volatile times while keeping your risk constant. It also normalizes risk across different pairs; a position in GBP/JPY (a volatile pair) will be smaller than one in EUR/CHF (a low-volatility pair).
To implement, calculate the 14-period ATR on your trading timeframe, multiply it by your chosen ATR multiple to get the stop distance, and then apply the fixed fractional formula using that stop distance. For example, with a 2x ATR stop on EUR/USD where the daily ATR is 65 pips, the stop is 130 pips. On a $50,000 account risking 1%, the position size is $500 / (130 x $10) = 0.38 standard lots -- much smaller than the 1.67 lots from a 30-pip stop, reflecting the wider stop needed for the current volatility.
Scaling In, Scaling Out & Pyramiding

Scaling in refers to building a full position gradually rather than entering all at once. A trader might allocate one-third of the intended size at the initial entry, add another third on a pullback that confirms the thesis, and the final third on a second confirmation. This reduces the average entry price risk because only a partial position is exposed if the first entry is poorly timed.
Scaling out is the reverse: closing portions of the position at progressive profit targets. A common approach is to close half at 1:1 risk-to-reward and let the remainder run with a trailing stop. While scaling out reduces the average winner size, it locks in profit and improves the psychological experience of trading. Many traders find it easier to hold positions when partial profits are already secured.
Pyramiding is a specific form of scaling in where additional lots are added only as the trade moves in profit. Each new addition is smaller than the previous one to maintain a favorable average price. For instance, the first entry might be 1 lot, the second add (after price confirms) is 0.5 lots, and the third is 0.25 lots. The total risk is managed so that if the trade reverses to the trailing stop, the profit from the initial position offsets the loss on the later additions. Pyramiding is most effective in trending markets and should be avoided in ranges.
Key Takeaways
- Full Kelly sizing maximizes growth but is too aggressive; use half-Kelly or quarter-Kelly in practice.
- Fixed fractional sizing (1-2% risk per trade) automatically adjusts position size with account balance.
- Volatility-based sizing with ATR normalizes risk across different pairs and market conditions.
- Scaling in reduces timing risk; scaling out locks in partial profits and improves trade management psychology.
- Pyramiding adds to winners in trending markets but requires strict rules to avoid giving back profits.