Position Sizing and Risk Per Trade

The correct sequence is: find the setup, mark the level that invalidates it, decide what that loss may cost the account, then divide. Share count is the output of that division, never an input you choose because a number feels right.

Concretely: a $50,000 account risking 1 percent tolerates a $500 loss. If the invalidation level sits $0.40 below entry, the position is 1,250 shares. Move the stop to $1.00 and the same risk buys 500 shares. The risk is constant; the size flexes.

This is why 'I bought a thousand shares' is not a risk statement and 'I risked half a percent' is. Traders who size by share count are letting the volatility of the instrument decide their loss for them.

Between 0.25 and 1 percent of equity per trade covers most professional practice, with active intraday traders often at the lower end because they take many more positions. The number that matters is not per-trade risk in isolation but per-trade risk multiplied by your realistic worst losing streak.

Ten consecutive losers is not unusual for a strategy that wins 45 percent of the time. At 1 percent that streak is a 10 percent drawdown you can trade out of. At 5 percent it is a 40 percent drawdown that requires a 67 percent gain to recover, which effectively ends the strategy.

That asymmetry is the entire argument. Losses compound against you faster than gains compound for you, so the goal of sizing is not maximizing the good case, it is guaranteeing you survive the ordinary bad case.

Five separate positions in the same sector on the same catalyst are one position wearing five tickers. Total portfolio heat — the sum of risk across open trades that would lose together — is the number to cap, typically at two to three times your single-trade risk.

A daily loss limit does the work discipline cannot. Two or three full stops and the platform closes, because the third and fourth trades after a bad morning are almost never taken for the reasons the first one was.

Overnight positions need their own smaller sizing. A stop does not protect you through a gap, so the practical risk on a held position is larger than your marked level implies.

Increase size on evidence, not on feeling. A defensible rule: raise per-trade risk only after a set number of trades at the current size with results the log supports, and cut it back automatically after a defined drawdown. Size should be a function of recent performance, not recent confidence.

Scaling into a position works when each add is priced off a new level with the total risk still inside the same budget. Averaging down without a plan is not scaling, it is converting a defined loss into an undefined one.

Track R rather than dollars — every trade measured in multiples of the risk you took. It makes results comparable across account sizes and instantly exposes the trade where you quietly risked four times normal because you were sure.