Position Sizing Principles
Risk a tiny fixed slice per trade so no single loss can hurt you: decide the loss, let the math pick the size.
Notice: you deploy $4,000 but only $100 is truly at risk. Widen the stop and your size shrinks, same risk.
Play with the builder above first. Set your account size, pick how much you're willing to lose, and set where your stop goes. Watch the position size change. That's the whole skill of this module in one widget: you decide the loss, and the math tells you the size. Not the other way around.
The one habit that outlives everything else
Most accounts don't die from a bad call. They die from a good-sized loss on a too-big position. You can be right more than half the time and still get wiped, if the losers are large enough.
Position sizing is the fix. It's boring. It's also the single habit that separates traders who are still here next year from the ones who aren't.
What "position size" actually means
Position size is not "how much margin I put in." It's how much of your account is genuinely at risk on this trade: the money that disappears if your stop hits.
That number, the real loss, is the thing you control. Everything else (leverage, notional, coin) is just knobs you turn to hit it.
The survival rule: ~1% per trade
Here's the rule the whole module is built on:
Risk about 1% of your account on any single trade.
On a $1,000 account, that's $10. On $10,000, it's $100. Not 1% of your margin. Not 1% "sort of." One percent of your total account, as the amount you lose if the trade goes fully against your stop.
Why 1%? Because it makes a losing streak survivable. Lose 10 trades in a row at 1% each and you're down about 10%, annoying, fully recoverable. Do the same at 10% per trade and you're down 65% and need a 186% gain just to get back to even. Small size keeps you in the game long enough for your edge to show up.
"Risk 1%" means the LOSS, not the margin
This is the part everyone gets wrong, so read it twice.
When you open a 100x perp position with $10 of margin, your margin is $10, but that is not your risk. Your risk is whatever you lose before you get out. If your stop is close, you might risk only $10 even while controlling $1,000 of coin. If you have no stop, your risk is your entire margin (a 1% move against 100x = liquidation).
So "risk 1%" is a statement about your exit plan, not your deposit. No stop = no defined risk = you're gambling, not sizing.
Do it in this order: stop first, size second
Beginners size first ("I'll put in $50") and hope. Flip it:
- Find your stop. Where does this trade prove itself wrong? Maybe below a support level, maybe a fixed % away. Say your entry is $60,000 and your stop is $58,800. That's a 2% stop distance.
- Fix your risk. 1% of a $2,000 account = $20. That's the most you'll lose.
- Let the size fall out of the math. You want a 2% move to cost exactly $20. So your position is $20 ÷ 2% = $1,000 of BTC. If price hits your stop, you lose $20. Done.
Notice: you never picked the size directly. The stop and the risk picked it for you.
It works on any coin
A tight-stop BTC scalp and a wide-stop altcoin swing can carry the same $20 risk: the volatile alt just gets a smaller position because its stop is further away. That's the point. Same risk, different size, every symbol. Your account stops caring which coin you traded and starts caring only about the 1%.
The formula (memorize this one line)
Position size = (Account × Risk%) / (Stop distance%)
Worked:
| Input | Value |
|---|---|
| Account | $2,000 |
| Risk % | 1% → $20 |
| Entry | $60,000 |
| Stop | $58,800 (2% away) |
| Position size (notional) | $20 ÷ 0.02 = $1,000 |
| Coin quantity | $1,000 ÷ $60,000 = 0.0167 BTC |
Leverage just sets the margin, not the risk. At 100x this $1,000 position needs only $10 of margin, but your loss on a stop-out is still $20, because that's what the stop enforces. Wider stop → smaller size. Tighter stop → bigger size. Same $20 either way.
Three ways to size: fixed-fractional, fixed-dollar, ATR
Fixed-fractional (recommended default). Risk a constant percent of the current account, e.g. 1%. Your dollar risk grows as you win and shrinks as you lose: automatic compounding, automatic braking. This is what the survival rule uses.
Fixed-dollar. Risk the same dollar amount every trade, e.g. $20, regardless of account size. Simple, but it doesn't scale up as you grow and doesn't protect you as you shrink. Fine for tiny or brand-new accounts.
ATR-based (volatility sizing). Set your stop distance from the Average True Range (a measure of how much the coin typically moves), e.g. stop = 1.5 × ATR. Then feed that distance into the same formula. This makes calm coins get bigger positions and wild coins get smaller ones, automatically. Same 1% risk, sizing that respects volatility.
All three plug into the exact same formula. Only the stop-distance input changes.
Kelly, and why 1% is deliberately below it
The Kelly criterion calculates the mathematically "optimal" fraction to bet given your edge:
f* = W - (1 - W) / (R)
where W = win rate and R = reward-to-risk ratio. Example: win 50% of the time (W = 0.5) at 2:1 payoff (R = 2):
f* = 0.5 - (0.5) / (2) = 0.25 → 25%
Full Kelly says risk 25% per trade. Do not do this. Full Kelly maximizes long-run growth but has brutal drawdowns (a run of losses at 25% each is catastrophic), and it assumes you know your true edge, you don't. Real traders use fractional Kelly: a half, a quarter, or less. Your 1% rule is roughly a twenty-fifth of full Kelly for the example above, intentionally conservative, so a wrong estimate of your edge can't ruin you.
Rule of thumb: when in doubt, size below Kelly, never above.
A real perp example: with fees and funding
Same trade as before: $2,000 account, 1% = $20 target risk, long BTC perp, entry $60,000, stop $58,800 (2%), notional $1,000.
Now the costs a spot example ignores:
| Cost | Calc | Amount |
|---|---|---|
| Price loss at stop | $1,000 × 2% | $20.00 |
| Taker fees (in + out) | $1,000 × 0.05% × 2 | $1.00 |
| Funding (held ~1 day, 3× 0.01%) | $1,000 × 0.03% | $0.30 |
| Real loss if stopped | $21.30 (1.07%) |
Fees and funding push your true risk past 1%. So on perps, size down a touch (use notional ≈ $940 instead of $1,000) to land back on a real $20 loss after costs. The longer you hold, the more funding eats, so wide-stop swing trades need a slightly bigger haircut than quick scalps.
Quick knowledge check
When you "risk 1%," is that your margin or your loss? Your loss: the amount gone if your stop hits. Margin can be far smaller.
What do you decide first: position size or stop? The stop. Size is calculated from the stop distance and your fixed risk, never guessed.
Two coins, same 1% risk, why different position sizes? The wider stop (more volatile coin) gets a smaller position, so the dollar loss stays equal.
Sources
- Van K. Tharp, Trade Your Way to Financial Freedom: position sizing and the fixed-fractional model.
- J. L. Kelly Jr., "A New Interpretation of Information Rate," Bell System Technical Journal (1956): the original Kelly criterion.
- Ralph Vince, The Mathematics of Money Management: fractional Kelly and drawdown math.
- J. Welles Wilder, New Concepts in Technical Trading Systems (1978): the Average True Range (ATR).