Position sizing for a prop challenge: the only maths that matters

A challenge is not a profit competition, it is a survival test with a profit target attached. The drawdown limit is a hard floor and the target is a soft ceiling — so sizing is the whole game, and it is arithmetic rather than opinion.

You have room (the drawdown limit in dollars) and you will have a losing streak (consecutive losses). Survive the streak and the edge, if it exists, does the rest:

risk per trade ≤ room ÷ (longest expected losing streak × safety factor)

Both inputs are measurable. Room is in the firm's rules. The streak is in your own backtest — and use the worst one in the history, not the average, because the worst one is the one that ends the account. For reference, a gold strategy we run has a longest losing streak of 21 over 818 trades, at a 47.6% win rate.

Fixed sizing does not shrink after losses, so your risk as a percentage of remaining room grows exactly when you can least afford it. The same 1% risk becomes 2% of what is left when you are halfway to the floor, and 4% when you are three-quarters there. Percentage-of-equity sizing shrinks automatically, which is the entire mechanism that keeps you inside the limit.

One strategy, one set of rules, 818 trades of gold over eleven years, no leverage, up to three positions at once. The only thing changing across these rows is risk per trade:

Two things to take from that. First, the signal never changed — win rate is 47.6% in every row. Sizing scales the outcome and the pain; it does not create or destroy an edge. Second, and this is the part that matters for a challenge: only the 0.30% row fits inside a 10% drawdown limit, and that row turns $10,000 into $18,537 over eleven years.

That is the real trade-off, with no flattering version available. Size it to pass, and the profit target in a 30-day window is a long way off. Size it to hit the target, and the drawdown rule ends you first.

1. Maximum drawdown, in %. If it exceeds the firm's limit anywhere in the history, the strategy cannot pass at that size. Reduce until it fits with room to spare.
2. Longest losing streak. Divide your room by it. That is your ceiling for risk per trade, and it is usually near 1%, not the 2-3% commonly repeated.
3. Stop distance against 1 ÷ leverage. If they are close, your stop cannot fire and every loss is a full margin loss.
4. Maximum concurrent positions, and the margin they needed together. The drawdown limit sees them all at once.

Size the strategy so its historical drawdown fits the challenge rule, and most strategies stop being interesting — the profit target moves out of reach in the time allowed. That is not a flaw in the arithmetic; it is the arithmetic telling you the strategy was never a challenge strategy. Knowing that costs nothing here and costs a fee there.