Position Sizing
ES: Dimensionamiento de Posiciones PT: Dimensionamento de Posição
The trader’s most important decision: how much capital to risk on each trade. Sizing determines whether the portfolio survives losing streaks, and errors here destroy accounts even with a profitable strategy.
What Is Position Sizing?
Position sizing is the decision of how much capital to allocate to each individual trade and, equivalently, how much risk to take on each one.
It is mathematically more important than entry timing, asset selection or risk-reward analysis, because it determines the portfolio’s survival through the inevitable losing streaks. A trader with a profitable strategy but poor sizing can be ruined; one with a mediocre strategy but excellent sizing can survive long enough to improve it.
Van Tharp demonstrated in Trade Your Way to Financial Freedom (1998) that sizing explains more than 90% of the dispersion in results between traders using the same strategy.
Hence the most quoted rule in the business: never risk more than 1% or 2% of the portfolio on a single trade. Its origin is the mathematics of survival. At 1% risk per trade, it takes 69 consecutive losses to halve the account; at 2%, 35; at 5%, only 14; and at 10%, seven.
Realistically, any strategy — even an excellent one — produces runs of five to ten consecutive losses from time to time. 1% absorbs them and allows recovery. 2% is aggressive but acceptable for experienced traders. 5% or more is financial suicide when the run arrives.
An essential distinction: sizing refers to the risk taken, not the amount invested. With a 100,000 portfolio and 1% risk, the loss limit per trade is 1,000. If the stop sits 5 units away on a share bought at 200, the position is 1,000 / 5 = 200 shares, that is 40,000 invested: 40% of the portfolio committed, but only 1% at risk.
The Basic Formula
The formula connects the risk budget, the distance to the stop and the position size: size = (portfolio × risk percentage) / (entry price − stop price).
A detailed example: a 100,000 portfolio with a 1% risk target gives a budget of 1,000 per trade. You buy at 180 with the stop at 175, so the distance is 5. The size is 1,000 / 5 = 200 shares, and the capital invested 36,000, thirty-six times the risk taken.
If the trade works and price reaches the 195 target, the profit is 200 × 15 = 3,000, or 3% of the portfolio. If it fails and the stop triggers, the loss is 200 × 5 = 1,000, exactly the 1% planned.
The formula varies by instrument. In shares it is the one above, directly. In long options, since maximum loss is the premium paid, the number of contracts is the risk budget divided by the premium per contract: with 1,000 of risk and a call at 5 — 500 per contract — that gives two contracts. In spreads, maximum loss is the width less the credit, or the net debit, and the number of contracts is the budget divided by that maximum loss. In futures the calculation is more complex and involves tick value and margin.
With several simultaneous positions you must watch aggregate risk: the usual rule is not to exceed 5% to 10% total portfolio risk at any moment. At 1% per trade, that allows a maximum of ten concurrent positions.
And you must consider correlation: ten highly correlated positions — ten technology names during a correction, say — behave as a single large position for risk purposes. When correlation is high, sizes must come down.
Kelly Criterion vs Fixed Fraction
Several methodologies exist, each with its trade-offs.
The fixed fraction — the 1% or 2% rule — is the professional standard: you risk a fixed percentage of the portfolio on every trade. Its advantages are simplicity, guaranteed survival through losing streaks, and automatic compounding during good runs, because size grows with the portfolio. Its disadvantage is that it does not adapt to the quality of each individual trade.
The Kelly criterion is the mathematically optimal formula for maximising growth: fraction = hit rate − (miss rate / risk-reward ratio). At a 60% hit rate and a one to two ratio, that gives 0.60 − (0.40/2) = 0.40, or 40% of capital per trade.
The problem is that Kelly assumes perfect knowledge of the hit rate and the risk-reward ratio. In practice the estimates are wrong, and trading at the full percentage leads to catastrophic losses. The usual solution is half Kelly — 20% in the example above — which is still aggressive, or quarter Kelly, which is what most professionals use.
Volatility adjustment sizes the position according to market volatility, using twice the fourteen-session average true range as the stop distance. More volatility means wider stops and smaller positions, and vice versa, which keeps risk constant across different market regimes.
The fixed amount — always risking the same sum — is simple but does not scale as the portfolio grows.
The recommendation for the retail trader is clear: the 1% fixed fraction is the foundation. Master it before experimenting with Kelly or with volatility adjustments. Most accounts that blow up do so from risking more than 2% per trade or from sizing inconsistently: 3% here, 8% there, 1% after a loss.
Sizing With Options
Sizing with options has important nuances.
In long options, maximum loss is the premium paid, so the calculation is direct: contracts = risk budget divided by premium per contract. With 1,000 of risk and a premium of 3 — 300 per contract — that gives three contracts.
In credit spreads — bull put, bear call, iron condor — maximum loss is the width less the credit received. A 5-wide spread with 1 of credit has a maximum loss of 400 per contract, so a 1,000 budget gives two contracts rounding down.
In debit spreads, maximum loss is the net debit and the calculation is immediate.
Naked short options carry potentially unlimited risk, which makes sizing problematic: the margin required and the potential loss bear no relation to each other. Defined-risk spreads are preferable.
Correlation applies here too: ten contracts on the same underlying are one large exposure, not ten small ones. Diversify across underlyings.
The greeks add complexity: gamma spikes near expiration, so size prudently in the final two or three weeks.
The typical errors are four. Using margin capacity instead of risk: "I can buy ten contracts on margin", without the total risk fitting the budget. Ignoring maximum loss and calculating only the entry cost. Oversizing weekly expirations, which can swing 50% or 100% in a session. And confusing net with gross exposure: an iron condor’s net credit looks small, but gross exposure — width times contracts times a hundred — can be enormous.
The practical rule is to always calculate the position’s total maximum loss across every leg, and size it so that it equals 1% or 2% of the portfolio, commissions and slippage included.
Psychology and Common Errors
The psychology of sizing is where many traders fail. There are seven recurring errors.
Increasing size after gains. Doubling the position after several winning trades is emotionally natural but destroys risk discipline. The result is that a single large loss in the next bad run takes everything accumulated.
Averaging down. Adding to losing positions is seductive — "I improve my average price" — but it actually increases risk on a thesis that is already failing. It is one of the most destructive behaviours in the business.
Trying to win it back. Taking larger positions after a loss to recover it quickly almost always produces larger losses.
Revenge trading. Similar to the above but purely emotional: getting angry at the market and trading to get even guarantees negative expectancy.
Cutting size too far after a bad run. The inverse problem: after five losses, the trader feels a cold streak and halves the size or less, missing the winners of the reversal.
Oversizing "high conviction" trades. "This one is different, I know it will work." Breaking discipline on a feeling usually ends badly.
Ignoring correlation. Ten technology names during a crash mean a 10% loss in a single session, when aggregate sector exposure should not exceed 3%.
The discipline framework has four elements: written rules before trading, with the exact percentage, correlation adjustments and concentration limits; a log of every trade comparing planned size with actual, to detect drift; an automatic halt when a violation is detected; and a monthly review of aggregate risk, concentration and correlation.
Van Tharp’s formulation sums it up: sizing explains 90% of results. Mastering that single concept, together with the discipline to respect it, matters more than any indicator, pattern or strategy.
Sizing Methods Compared
Each method balances complexity and adaptability differently.
| Method | Risk per trade | Complexity | Best suited to |
|---|---|---|---|
| 1% of the portfolio | Low | The foundation for any trader | |
| 2% of the portfolio | Low | Experienced traders with a validated strategy | |
| Variable and high | Medium | The theoretical optimum | |
| Half the calculated Kelly | Medium | A less aggressive variant | |
| Inversely proportional to volatility | High | Traders across several markets | |
| A constant absolute sum | Low | Small accounts, for simplicity |
Frequently Asked Questions
Why 1% to 2% and not 5%?
In probability terms, any strategy produces five to ten consecutive losses from time to time, even excellent ones. At 5% risk, ten consecutive losses mean a 40% account drawdown, hard to recover. At 10%, the drawdown exceeds 65% and recovery is all but impossible. At 1% or 2%, it stays between 5% and 20%, perfectly manageable.
The arithmetic of survival imposes conservative sizing. Experienced traders can go to 2% or 3% with a validated strategy, but 1% remains the foundation.
How do I calculate size with options?
In credit spreads: contracts = budget divided by maximum loss per contract, that being the width times a hundred less the credit times a hundred. A 5-wide spread with 1 of credit has a maximum loss of 400, so 1,000 gives two contracts.
In debit spreads: contracts = budget divided by the net debit multiplied by a hundred.
The common rule is to always calculate maximum loss explicitly, not just the entry cost.
What do I do if many positions are correlated?
Correlation is not always obvious: growth stocks correlate with each other, as do small caps or emerging markets. And during the 2008 crisis, supposedly uncorrelated assets all moved together.
Real discipline means treating correlated positions as a single exposure for risk budgeting purposes.
Should I increase size when I am winning?
The wrong way: raising the percentage during a winning run, from 1% to 2% and then 5%. That abandons the discipline and creates enormous risk for when the reversal comes.
The amateur feels confident and raises size aggressively; the professional holds the percentage regardless of their emotions. Compound growth arrives on its own through portfolio appreciation at a constant percentage. Do not force it.
Is the Kelly criterion better than a fixed 1%?
Kelly also assumes stationary probabilities, and real markets change regime. A Kelly calculated on an assumed 40% is probably 30% in reality, which leads straight to ruin.
The usual solution — half Kelly or quarter Kelly — reduces the risk substantially but remains aggressive.
Most professionals use the 1% or 2% fixed fraction as the base, perhaps adjusting upward on very high conviction trades. For the retail trader, the sensible course is to stay with the fixed fraction.