How to Pass a Futures Prop Firm Evaluation: 9 Steps

To pass a futures prop firm evaluation reliably, you need a positive trading edge, risk per trade of about a tenth of the loss buffer, and the discipline to stay inside every rule until you hit the profit target, then stop. A futures prop firm evaluation (also called a challenge or combine) is a simulated account in which you must make a set profit before losing a set amount. A trader with no edge at all passes a $3,000-target, $2,000-loss evaluation 40% of the time, so a pass on its own proves little.

Key takeaway

A trader with zero edge passes a futures prop firm evaluation with probability buffer / (target + buffer), where the buffer is the distance down to the loss limit: 40% for a $3,000 target and $2,000 maximum loss, and about 24 to 26% once the loss limit trails. With a real edge (an illustrative 55% win rate at 1:1), cutting risk from $1,000 to $200 per trade lifts the pass probability from 52% to 87%, while oversizing drags every trader back toward the zero-edge odds.

Is passing a prop firm evaluation just luck?

For a trader with zero edge, passing is pure luck, with odds fixed by two rulebook numbers: the profit target and the buffer, the distance from the starting balance down to the loss limit. The setup is the classic gambler's ruin problem, a random walk (a path of independent up and down steps) between two walls that end the game. Eric Weisstein's gambler's ruin entry on Wolfram MathWorld gives the fair-game result: "your chances of going bankrupt are equal to the ratio of pennies your opponent starts out to the total number of pennies." In an evaluation, going bankrupt means hitting the loss limit, and the opponent's pennies are the distance to the target.

A zero-edge trader's expected profit is zero however the run ends, so with pass probability P, P × target - (1 - P) × buffer = 0, which solves to P = buffer / (target + buffer): 2,000 / 5,000 = 40% for a $3,000 target and $2,000 buffer.

The formula is exact under six assumptions: independent trades, a 50% win rate at 1:1 (zero expectancy, meaning the average trade earns nothing), a fixed risk R per trade, no commissions or slippage, a static floor, and a target and buffer that are whole multiples of R. Position size changes only how long the coin flip lasts: the average number of trades is (buffer / R) × (target / R), 2 × 3 = 6 at $1,000 risk and 10 × 15 = 150 at $200 risk, with the same 40% at the end.

With no edge, where the path ends depends only on the distances to the two lines. A $2,000 buffer against a $3,000 target passes two times in five, before costs and before a trailing floor makes it worse.

What are the zero-edge pass odds for common evaluation sizes?

With zero edge, pass odds run from 50% when the target equals the maximum loss down to 33.33% when the target is twice the loss, and a trailing loss limit removes another 11 to 18 percentage points. The $3,000 / $2,000, $6,000 / $3,000 and $9,000 / $4,500 rows match Topstep's $50K, $100K and $150K Trading Combine parameters as published in September 2026; the other rows are illustrative.

Target / maximum loss	Target ÷ loss	Static floor	Trailing, locks at start	Trailing, never locks	Average trades (static, R = loss / 10)
$1,500 / $1,500	1.0	50.00%	38.55%	38.55%	100
$3,000 / $2,500	1.2	45.45%	32.13%	31.86%	120
$1,500 / $1,000	1.5	40.00%	25.70%	23.94%	150
$3,000 / $2,000	1.5	40.00%	25.70%	23.94%	150
$6,000 / $3,000	2.0	33.33%	19.28%	14.86%	200
$9,000 / $4,500	2.0	33.33%	19.28%	14.86%	200

With target T, maximum loss D and risk R = D / 10 per trade, a static floor gives P = D / (T + D). A floor that trails the closed-balance high trade by trade needs each new high, R above the last, before a ten-step fall, a 10/11 chance, so P = (10/11)^(T/R) if it never locks and P = (10/11)^10 × D / T if it locks at the starting balance once the high reaches +D. Average trades = (D / R) × (T / R).

Only the target-to-loss ratio matters, so the $6,000 / $3,000 and $9,000 / $4,500 rows are identical. With ever smaller bets the trailing values fall only slightly, to e^(-T/D) = 22.31% and, with the lock, e^(-1) × D / T = 24.53% at a 1.5 ratio (Lehoczky 1977, The Annals of Probability).

Firm-published results sit below even the trailing model. Topstep's 2025 trader performance disclosure states that "16.8% of all Trading Combines initiated were successfully completed and afforded the opportunity to advance to the Funded Level," under the 25.70% zero-edge trailing value and consistent with costs, oversizing and rule breaches. The comparison is suggestive only: the count includes every combine initiated, possibly abandoned ones, and Topstep trails on end-of-day balance (milder than per-trade trailing) but liquidates on real-time unrealized P&L (harsher).

A trailing floor takes roughly a third off the coin-flip odds. The reported completion rate sits below every trailing bar, which is what costs, rule breaches and negative edges add on top of chance.

Does trading smaller raise your chance of passing?

Trading smaller raises the chance of passing only for a trader with a positive edge; with zero edge size changes nothing, and with a negative edge smaller size makes failure more certain. The unfair-game form of gambler's ruin gives P = (1 - r^b) / (1 - r^(a+b)), where p is the win probability, r = (1 - p) / p, and b and a are the buffer and target counted in units of risk (b = 2,000 / R, a = 3,000 / R). Small bets give expectancy (average profit per trade) more trades to outweigh luck before either wall is reached, whatever its sign.

Pass probability for a $3,000 target and $2,000 static loss limit, 1:1 payoff, illustrative win rates:

Risk per trade	55% win (+0.10R)	53% win (+0.06R)	50% win (zero edge)	50% win, 5% cost (-0.05R)	48% win (-0.04R)
$100	98.20%	91.18%	40.00%	4.17%	7.37%
$200	87.13%	73.57%	40.00%	14.98%	19.17%
$250	81.39%	67.90%	40.00%	18.81%	22.67%
$500	63.76%	54.57%	40.00%	28.54%	30.77%
$1,000	52.20%	47.30%	40.00%	32.34%	35.28%

The 5%-cost column treats trading costs as negative edge: an illustrative cost of 5% of R per trade makes a win pay +0.95R and a loss cost -1.05R (computed exactly, including overshoot of the walls), which lifts the break-even win rate at 1:1 to 1.05 / 2 = 52.5%.

Sizing down helps only a trader with an edge: more trades let the edge show. For a trader without one it does the opposite, which is why "trade small" is correct advice for exactly the people who have something to protect.

Once costs count, a trader with no edge gets the best pass odds by oversizing: 32.34% at $1,000 risk per trade against 4.17% at $100.

Every positive-edge column assumes the edge exists, and the evidence says it rarely does. Chague, De-Losso and Giovannetti (2020) studied every individual who began day trading Brazilian equity futures between 2013 and 2015 and persisted for at least 300 days, and reported that "97% of them lost money, only 0.4% earned more than a bank teller (US$54 per day)". Barber, Lee, Liu and Odean (2014, Journal of Financial Markets) found that less than 1% of Taiwanese equity day traders from 1992 to 2006 predictably and reliably earned positive abnormal returns net of fees. No step below creates an edge; the steps only stop the rules and bad sizing from wasting one.

Step 1: Which rulebook numbers decide whether you pass?

Five numbers decide an evaluation: the profit target, the maximum loss and its type, the daily loss limit, the consistency rule and the minimum trading days; write them down before the first trade. Topstep's $50K Trading Combine, as its help center described it in September 2026, is the worked example here (the Apex evaluation guide covers a different rulebook):

Profit target: $3,000 ($6,000 on $100K, $9,000 on $150K).
Maximum loss and type: $2,000, trailing. Topstep's Maximum Loss Limit article says: "The MLL is a trailing limit. It rises as your end-of-day balance grows, but never moves down. Once it reaches your starting balance, it locks permanently." The limit is monitored in real time, unrealized P&L included.
Daily loss limit: $1,000 and optional; hitting it flattens positions and pauses trading until the next session without ending the Combine.
Consistency: the best day may not exceed 55% of the target, $1,650, or the target rises to best day / 0.55.
Minimum days: Topstep says the Combine can be passed in as few as two days.

Read the maximum-loss type as closely as its amount: the same dollars can stay static, trail intraday, trail at the close or lock, and each version gives different odds (see how the profit target and drawdown interact).

Rules change

Topstep's help center showed a 55% consistency line in September 2026, while older third-party summaries still quote 50%. Check the current rulebook before each attempt.

Steps 2 to 4: How much should you risk per trade, and how many contracts?

Risk about a tenth of the maximum loss per trade, work out from your expectancy how many trades the target needs, and pick the contract whose point value expresses that risk exactly, which on a $50K evaluation means micros.

Step 2, buffer to risk. $2,000 / 10 = $200 per trade, so failing takes ten straight full losses or a long net losing run. Halving again to $100 lifts the illustrative 55% trader from 87.13% to 98.20% in the edge table, at the price of about 2.5 times as many trades (290.99 against 117.84 expected trades to finish). Sizing from the nominal $50,000 misleads: 1% of it is $500, a quarter of the buffer (more in position sizing and the 1% rule).

Step 3, trades from expectancy. Expectancy per trade = win rate × average win - loss rate × average loss. At 55% and 1:1, 0.55 × 1 - 0.45 × 1 = +0.10R, or $20 at $200 risk, so the $3,000 target takes 3,000 / 20 = 150 trades on average if the floor never intervenes. Your own trade record supplies these inputs; if it cannot produce a positive number, stop here. Trades per day turn the count into a calendar (how long passing takes).

Step 4, contract choice. A 10-point stop on the Micro E-mini S&P 500 ($5 per index point, one tenth of the $50-per-point E-mini, according to CME Group's contract specifications) risks $50 per contract, so four micros express exactly 10 × $5 × 4 = $200. The same stop on one E-mini risks $500, a quarter of the buffer on a single trade.

Size comes out of the chain, it never goes into it. The same ten-point stop on one E-mini would spend a quarter of the buffer on a single trade and drag the pass odds back toward chance.

Steps 5 to 7: How do daily limits, consistency rules and news releases fit in?

Set a personal daily loss stop inside the firm's limit, cap each day's profit below the consistency threshold, and stand aside around scheduled economic releases.

Step 5, personal daily stop. An illustrative stop of three full losses, 3 × $200 = $600, sits well inside Topstep's $1,000 daily limit. In the independent-trades model a daily stop changes nothing; its value is behavioral, because chasing a bad day invites bigger size. Where a firm's daily limit fails the account rather than pausing it, that limit is a second floor. Write the stop into your trading plan before the session starts.

Step 6, daily profit cap. Topstep's consistency article is blunt: "55% is a hard line. It is not rounded, and there is no buffer." Topstep's own example shows a $1,800 best day raising the $50K target to $1,800 / 0.55 = $3,272.73, stated as $3,273. A personal cap a little below the line, for example $1,500, leaves room for a last open trade. Other firms write the rule differently (consistency rules explained).

Step 7, scheduled releases. Ederington and Lee (1993, Journal of Finance), studying interest-rate and currency futures, found that after a major announcement "volatility remains substantially higher than normal for roughly fifteen minutes and slightly elevated for several hours." A stop sized for normal conditions is undersized in those minutes, and slippage past a stop is the overshoot the formula assumes away. Some rulebooks also restrict trading around releases (news blackout windows).

Steps 8 and 9: How do you manage a trailing drawdown, and when do you stop?

Manage a trailing drawdown by tracking the corridor between floor and target, which narrows each time a new high lifts the floor, and stop the moment every pass condition is met.

Step 8, the trailing floor. On a Topstep-style $50K account (end-of-day trailing, locking at the start), the account opens at $50,000 with the floor at $48,000 and the target at $53,000, a $5,000 corridor. After an end-of-day high of $51,200, the floor rises to $49,200 and the corridor shrinks to $3,800. From $51,200, a zero-edge trader's odds are 2,000 / 3,800 = 52.63% if the floor stays put, against 3,200 / 5,000 = 64.00% at the same balance with a static floor, and (10/11)^4 × 2/3 = 45.53% if the floor keeps trailing in $200 steps. After an end-of-day high of $52,000, the floor locks at $50,000, the corridor is $3,000, and the odds from $52,000 are 2,000 / 3,000 = 66.67%.

Every new end-of-day high drags the floor up behind you, so progress buys less room than it looks like. Once the floor locks at the starting balance, the corridor stops shrinking and the arithmetic finally works in your favour.

The first $2,000 of profit, before the lock, is where trailing costs the most, so keep size flat there rather than pressing after a good day. The squeeze hits edge traders too: at an illustrative 55% win rate and $200 risk, a per-trade trailing floor with a lock cuts the pass probability from 87.13% to 68.92%. Intraday, end-of-day and static floors are compared in trailing vs static vs EOD drawdown.

Step 9, stop at the target. Stopping locks in a 100% pass probability, and any further trade can only lower it. If a minimum-days rule is still open, meet it with the smallest position the rulebook counts as a trading day.

Is it worth buying multiple evaluations to get lucky?

Buying repeated evaluations without an edge is a paid coin flip: it raises the chance of an eventual pass but delivers a trader with no edge into a funded account, where the same arithmetic runs again. At the 40% zero-edge odds, at least one pass comes 64.00% of the time within two attempts and 92.22% within five, and at Topstep's 16.8% per combine the expected number of attempts to a first pass is 1 / 0.168 = 5.95. Four independent attempts at 16.8% give 1 - 0.832^4 = 52.08%, near the 51.8% of 2025 Topstep participants who advanced at least once, a loose match because attempts per trader vary and one trader's attempts are not independent. The same disclosure reports that 33.3% of funded-level participants received a payout. Why passes turn into losses is covered in why funded traders fail, and the funded-side math in risk of ruin for a funded account.

A trade copier does not change the pass arithmetic, and for a trader without an edge it is not the answer. Thor, this blog's own product, copies one master account to many evaluation or funded accounts with per-account sizing, which scales an edge already demonstrated. Copying a zero-edge strategy into five evaluations does not buy five tickets: identical trades under identical rules and sizing pass or fail together, so the chance of at least one pass stays near 40%, not 92.22%. Check each firm's rules on copy trading before connecting any copier.

Go deeper

Frequently asked questions

What percentage of traders pass a futures prop firm evaluation?

No verified industry-wide pass rate exists. Figures such as 5-10% passing, repeated on many blogs, usually cite no dataset or method. Firm-published numbers are the reliable ones, and they must be read per attempt or per person, since the two differ by a factor of about three at Topstep (16.8% of combines against 51.8% of participants in 2025).

Can you pass a prop firm evaluation in one day?

Not under a consistency cap, because a single day that makes the whole target is 100% of the profit; at Topstep's 55% line a one-day $3,000 result would raise the target to about $5,455. Where no such rule applies, finishing in one day needs size large enough to reach the target in a few trades, which drags the odds toward the zero-edge figure.

How many contracts should I trade on a $50K evaluation?

Far fewer than the maximum the firm allows. Topstep's $50K Combine permits up to 5 E-minis or 50 micros, yet 5 E-minis with a 10-point stop risk 5 × 10 × $50 = $2,500 on one trade, more than the whole $2,000 buffer.

Is the gambler's ruin model realistic for real trading?

The model is exact only for its stated assumptions, but its zero-edge result is sturdier than it looks. For any zero-expectancy strategy whose trades end close to the walls, the average-profit argument still gives roughly buffer / (target + buffer) whatever the win rate or payoff shape; what the model misses is overshoot from gaps and slippage, costs, and an edge that changes over time.

Which matters more, the profit target or the maximum loss?

For a trader with an edge, the depth of the buffer measured in trades matters more than the target-to-loss ratio. At an illustrative 55% win rate and a buffer ten losses deep, a 1.5 ratio ($3,000 / $2,000 at $200 risk) passes 87.13% of the time and a 2.0 ratio ($9,000 / $4,500 at $450 risk) passes 86.77%, while the zero-edge odds for the same two pairs fall from 40.00% to 33.33%.

Why do traders fail evaluations close to the target?

The odds are not the cause, since they are highest there. A zero-edge trader at +$2,800 on a $3,000 target with a static $2,000 floor passes 4,800 / 5,000 = 96% of the time, and 2,800 / 3,000 = 93.33% once a trailing floor has locked at the start, so a late failure points to execution, such as size raised to finish quickly.

How long does passing take at a realistic trading pace?

Divide the trades your expectancy requires by the trades you take per day: the illustrative 150 trades at +$20 each take 30 trading days at five trades a day and 50 at three. Larger size shortens the calendar but, for a trader with an edge, lowers the pass probability, so speed is paid for in odds.

Is end-of-day trailing drawdown easier than intraday trailing?

Generally yes. An end-of-day floor rises only with the closing balance, so an open profit that reverses before the close does not lift it, while an intraday trailing floor ratchets up on every unrealized high.

Is a 55% win rate at 1:1 a realistic goal?

No typical figure can be quoted: the 55% used here is illustrative, and no study checked for this article measures prop-firm traders' win rates. What matters is the sign of your own expectancy over enough trades, and over 20 trades the standard error of a win rate is about 11 percentage points (the square root of 0.25 / 20), far wider than the 7-point gap between a 55% and a 48% trader.