Position sizing and risk of ruin
Strategy gets the attention. Sizing decides whether you are still in the game after a bad month.
3 min read4-question quiz150 XP
You cannot control whether a given trade wins. You can control how much it costs when it loses. Position sizing is the one lever that is entirely yours.
Contracts = floor( Account x Risk % per trade / Max loss per contract )
Use the real worst case at the stop or at expiry. For long options and spreads, that is usually the full debit or width minus credit.
| Account | 1% budget | Long call ($359) | Bull call spread ($294) | Bull put spread ($407) | Iron condor ($306) |
|---|---|---|---|---|---|
| $10,000 | $100 | 0 | 0 | 0 | 0 |
| $25,000 | $250 | 0 | 0 | 0 | 0 |
| $50,000 | $500 | 1 | 1 | 1 | 1 |
| $100,000 | $1,000 | 2 | 3 | 2 | 3 |
Calculator
How many contracts fit your plan?
- Risk budget
- $500
- Contracts
- 1
- Dollars at risk
- $306
- 10 straight losses leave
- 90.4%
The arithmetic of drawdowns
Losses hurt more than the headline number because recovery is measured against a smaller balance.
| You lose | You then need to gain to get back |
|---|---|
| 10% | 11.1% |
| 20% | 25% |
| 33.3% | 50% |
| 50% | 100% |
| 75% | 300% |
Losing streaks happen to everyone with any strategy. Here is what ten losses in a row leave behind at different risk-per-trade levels, using plain compounding:
| Risk per trade | Account left after 10 straight losses |
|---|---|
| 1% | 90.4% |
| 5% | 59.9% |
| 10% | 34.9% |
Expectancy beats win rate
Expectancy (in R) = p x W - (1 - p) x L
p is the win rate, W the average win and L the average loss, both measured in R, where 1R is the amount you risk per trade.
- Hypothetical: win 40% of the time, wins average 2R, losses average 1R. Expectancy = 0.4 x 2 - 0.6 x 1 = +0.2R per trade.
- Hypothetical: win 85% of the time, wins average 0.2R, losses average 1R. Expectancy = 0.85 x 0.2 - 0.15 x 1 = +0.02R. Still positive, but one extra loss in twenty trades turns it negative.
- A high win rate with small wins and big losses is the credit-spread shape. A low win rate with large wins is the long-option shape. Neither is better by itself.
Portfolio heat
Add up the maximum loss of everything open. If you hold long SPY calls, long QQQ calls, and a long NVDA call, that is mostly one bet: the market goes up. Treat correlated positions as one position when you count risk, and set a cap on total open risk, not just risk per trade.
Recap
Key takeaways
- 01Contracts = floor(account x risk % / max loss per contract). If the result is zero, the trade does not fit.
- 02Recovering from a 50% drawdown requires a 100% gain. Avoid the holes.
- 03Ten straight losses at 1% risk leave 90.4%; at 10% risk they leave 34.9%.
- 04Judge a strategy by expectancy and drawdown, not win rate. Count correlated positions as one when summing risk.
Terms in this lesson
- Position sizing
- Choosing how many contracts to trade so that a loss equals a pre-set fraction of the account.
- R-multiple
- A result measured in units of the amount risked. A trade that makes twice its risk is +2R.
- Expectancy
- Average result per trade: win rate x average win minus loss rate x average loss.
- Drawdown
- The decline from a peak in account value to a subsequent low.
- Portfolio heat
- Total open risk across positions, counting correlated positions together.