Position sizing answers a question that's separate from choosing a direction or picking leverage: how much collateral should go into this specific trade. Two traders can agree completely on which way an asset is headed and still produce very different outcomes because one sized the position to survive being wrong and the other didn't. Sizing is the risk decision; leverage and direction are the market decisions, and treating all three as one combined choice is where most avoidable losses on a perpetual actually come from.
It helps to treat "how much to risk" and "how much leverage to use" as independent choices. Leverage determines how much notional exposure a given amount of collateral controls; it does not by itself determine how much of the account is actually at risk if the trade goes wrong, that depends on where the stop is placed. A trader can use 10x leverage and risk a small, controlled fraction of the account, or use 2x leverage and risk a dangerously large fraction, depending entirely on position size and stop placement.
A common approach fixes the maximum amount of the account willing to be lost on a single trade, often a small single-digit percentage, and works backward from there. Given a stop-loss distance, the position size is whatever notional makes the loss at that stop equal the fixed risk amount. A wider stop, appropriate for a more volatile asset, requires a smaller position to keep the dollar risk constant; a tighter stop allows a larger position for the same dollar risk. This keeps risk consistent from trade to trade even as volatility and stop distance change.
Leverage changes how much margin is required for a given notional exposure. A $10,000 notional position with a stop 2% away risks $200 whether it's opened with $1,000 margin at 10x or $5,000 margin at 2x. What leverage changes is capital efficiency: how much of the account's total collateral sits idle and how much sits deployed. Confusing margin used with risk taken is one of the more common sizing errors.
Every market on Canborsa carries its own minimum and maximum position size, ranging from as little as $10 minimum on some assets up to a $20,000 maximum notional on others, alongside a leverage cap set per asset, from 3x on the most volatile tokens up to 30x on the S&P 500 index. These aren't arbitrary numbers, they reflect the liquidity and volatility profile of each specific market, and they set a hard ceiling on how large a single position can get regardless of how much collateral is available to post.
The Kelly criterion is a formula from probability theory that calculates the fraction of a bankroll to risk on a bet with a known edge and known odds, to maximize long-run growth. Applied to trading, it requires an estimate of win rate and average win-to-loss size, inputs that are rarely known with real precision in live markets, which is why full Kelly sizing is uncommon in practice. Its more common use is as a ceiling: most traders who reference Kelly use a fraction of the calculated size, often a quarter or a half, precisely because the formula is unforgiving of overestimated edge and produces aggressive position sizes even with modest input errors. Treated this way, Kelly is less a sizing formula to follow exactly and more a reminder that even a genuine statistical edge doesn't justify betting the size a naive read of that edge might suggest.
Sizing each position in isolation misses a risk that only shows up at the portfolio level: two positions in correlated assets, two crypto longs, or a stock long alongside an index long, don't diversify each other the way two uncorrelated trades would. A drawdown that hits one is likely to hit the other at the same time, so the combined risk is closer to the sum of the two than either one alone would suggest. Treating a cluster of correlated positions as one combined risk budget keeps total portfolio exposure from silently exceeding what was intended.
An account with $10,000 in total collateral sets a 2% max risk per trade, or $200. The trade setup calls for a stop 4% away from entry. Position size works out to $5,000 notional ($200 ÷ 4%), which at 5x leverage requires $1,000 in margin, leaving the rest of the account uncommitted. The 0.14% round-trip fee on that $5,000 notional adds $7, a small but real addition to the $200 risk already budgeted for the stop.
Position size should follow from a fixed dollar risk per trade and the distance to the stop-loss. Leverage sets the margin required for a given notional.
A sound approach fixes a small, consistent percentage of the account at risk per trade, sizes the position so the stop-loss distance produces exactly that dollar risk, and treats correlated open positions as a combined risk.