What 12,000 backtests reveal about prop firm payouts

If you spend any time in prop trading communities, you have seen the two genres of payout discussion. There are the withdrawal screenshots, a single five-figure number with a caption about consistency. And there is the expectancy math, win rate times average win minus loss rate times average loss, which is correct as far as it goes but ignores the part that actually decides whether you get paid: the path.

A strategy with a perfectly good expectancy can still violate an account rule before it ever pays out, simply because of the order its winning and losing trades happened to arrive in. Expectancy is a property of the average. Survival is a property of the sequence. We wanted to measure the sequence.

What 12,000 backtests actually means

We took eight portfolio configurations spanning futures and forex prop accounts, and ran each one through 1,500 Monte Carlo paths using block bootstrap resampling on a 12-month sample. Eight portfolios times 1,500 paths is 12,000 simulated account-years.

The point of resampling is to stop trusting the single backtest that happened to occur. One historical run gives you one ordering of trades. Resampling shuffles that history thousands of times while preserving its short-term structure, so you see what the same underlying edge can produce across a distribution of orderings rather than the one lucky or unlucky sequence you lived through.

For every path, instead of asking what it returned, we tracked three things: how many payouts it produced per year, how many days passed before the first payout, and whether the account violated a rule before it paid out at all. Every figure in this article is a median outcome — the middle of the distribution, not the best case.

The trade-off nobody puts on a chart

Here is the finding that reorganized how we think about sizing. Payout frequency and blow risk are not independent. They are two ends of the same dial. The configurations that paid out most often were, with one structural exception, the same ones that blew most often.

Read the futures portfolios along that dashed line and the relationship is plain. You do not get to select high frequency and high survival at the same time by turning up the size. Sizing up pulls both levers together: more contracts means you reach the payout threshold faster, and it means a normal losing stretch eats more of your drawdown limit on the way there.

Put two same-size accounts side by side to see it in money rather than abstraction.

The defensive book pays out under twice a year but its three-year survival rate sits near 99.9%. The growth book pays out three times as often, and you buy that frequency with a blow rate that climbs into the mid-single digits and a survival rate that drops fifteen points. Same idea, opposite point on the dial. Neither is wrong. They are answers to different questions.

Time to first payout is the hidden cost

The second number almost nobody factors in is how long you wait. Across the configurations, the median time to first payout ranged from under a month to over four months, and it lines up exactly with how conservatively the account was sized.

The most defensive book made you wait roughly 132 days before it produced anything eligible to withdraw. The most aggressive paid in under a month. That is intuitive once you see it, but it changes the decision. When you choose a sizing profile you are not only choosing how much risk you carry — you are choosing how long you sit with an account that has produced nothing yet. For a trader who needs the account to contribute income, four months of waiting is a real cost, not a footnote.

So which side do you size for

We are not going to tell you the defensive book is correct and the aggressive one is reckless, because that depends entirely on your situation. A trader living off withdrawals and a trader compounding a durable account are optimizing different things and should sit at different points on the dial.

What the 12,000 paths make hard to ignore is that the choice is real and most people make it by accident. They size for the fast, large withdrawal because that is the visible, exciting number — and that is the exact profile with the lowest survival to actually collect it. The payout you can repeat beats the bigger one you blow before you reach it. The only way to know which profile you are actually holding is to look at the whole distribution, not the one backtest that happened to go well.

If you want the full per-portfolio breakdown — payout frequency, survival, and time to first payout across all eight configurations — that's what the portfolio overview lays out, each one sized against its drawdown limit first.

FAQ

How often do prop firms actually pay out?

Across eight systematic portfolios run through simulation, the median was roughly four payouts per year, ranging from about 1.7 for the most defensive configuration to over eight for a decorrelated multi-strategy book. It depends heavily on how aggressively the account is sized, and it trades off against survival.

What is time to first payout and why does it matter?

It is the median number of days from starting an account to the first eligible withdrawal. In simulation it ranged from under a month for aggressive sizing to over four months for the most defensive. It is the hidden cost of conservative sizing — the safer the account, the longer you typically wait before it produces anything.

Does a higher payout frequency mean a better account?

No. Higher frequency is strongly correlated with higher blow risk — the configurations that paid out most often were generally the same ones that violated rules most often. More payouts isn't strictly better, it's a different point on a frequency-versus-survival trade-off.

How were these numbers calculated?

Eight portfolio configurations were each run through 1,500 Monte Carlo paths using block bootstrap resampling on a 12-month sample, for 12,000 simulated account-years. For each path the model tracked payouts per year, days to first payout, and whether the account violated a rule before paying out. All figures are median (P50) outcomes.

Are these payout figures guaranteed?

No. All figures are hypothetical, modeled outputs from simulation on a historical sample. Past results do not guarantee future returns, and a resampled simulation cannot reproduce a market regime that never occurred within the sample. The numbers describe a distribution of outcomes, not a promise.