Are There “Free” Trades? The Truth About High Win Rates

Short answer. There are no free trades. When someone says a setup is so obvious that you can close it with your eyes shut at 1:1 or 1.5:1, one of three things is going on: the target sits close and the stop far, so the win rate is bought with geometry; the sample is small or filtered by memory; or the large loss exists and has not happened yet. The win rate you need to break even is fixed by arithmetic, and no setup escapes it.

If you heard a trader say that after a drawdown on a funded account they switched to “free money” trades, it is worth separating two things. Cutting size after a loss is sensible. Searching for trades that cannot lose is not a strategy, because the market does not issue them.

What win rate do I need to break even?

Expectancy per trade, in units of risk (R), is the win rate times the reward, minus the loss rate, minus costs. Set it to zero and the break-even win rate is (1 + cost) ÷ (1 + reward:risk). At 1:1 it is 50%; at 1.5:1 it is 40%; at 2:1 it is 33%. Add a cost of 0.1 R per trade (spread, commission and slippage as a fraction of the stop) and those become 55%, 44% and 37%. That is our calculation from the formula, not a finding from a study.

So “I win 70% of my 1:1 trades” is only impressive if it is measured over enough trades and the stop and target were fixed in advance. The next sections explain why those two conditions are so often missing.

How much win rate can I buy with a closer target?

For a random walk, the chance of reaching a target before a stop is the stop distance divided by the sum of both distances, so the win rate falls from 80% to 20% as the target moves from a quarter of the stop to four stops away. We checked it on real candles: GER40 at 5 minutes, gold at 1 hour and Bitcoin at 1 hour, 2022–2026, with random entries and random direction, a stop of 1 ATR and a target at various distances.

The win rate matched the geometry within a few points everywhere, and the expectancy after costs was negative at every distance, between −0.02 and −0.20 R. Before costs it ranged from −0.07 to +0.03 R, which is zero within noise; the slightly low win rates at near targets come from our rule that a candle touching both stop and target counts as a stop. The point is not that markets cannot be beaten. It is that a 75% win rate on its own is cheap to produce and tells you nothing about skill. Costs are assumptions (0.006%, 0.012% and 0.08% per round trip), and they weigh more when the stop is tight, which is why near targets with tight stops fare worst.

Isn’t moving the stop to break-even a free trade?

In trader slang, a “free trade” is one where the stop has been moved to the entry price, so it can no longer lose. We tested it on the same random entries: stop 1 ATR, target 1 ATR, and the stop moved to the entry once price went 0.5 ATR in favour. Results across the three instruments: about 34–37% wins, 35–36% losses and 28–30% break-even exits, with expectancy before costs between −0.01 and +0.01 R, the same as without the move. The rule changes how the results look, turning a third of the trades into zeros and lowering the loss rate, and it changes nothing about the average.

That is not an argument against break-even stops, which may suit your psychology or your risk limits. It is an argument against the idea that they create an edge.

What do researchers say about high-win-rate profiles?

The shape of “win often, lose big rarely” is well studied under other names. Goetzmann, Ingersoll, Spiegel and Welch, in a paper about gaming performance measures, give a simple example: a portfolio that sells rarely-paying derivatives, such as deep out-of-the-money options, is better than a near-identical portfolio whenever the rare event does not happen, so it looks better by almost any measure. They conclude that “any portfolio, even an ex ante optimal one, can be beaten often.”¹ A high hit rate is easy to manufacture and proves nothing about the quality of a strategy.

Brunnermeier, Nagel and Pedersen found the same shape in currencies. Carry trades, which collect an interest-rate difference, show negative skewness: high interest-rate differentials predict crash risk. The paper links this to the trader saying that exchange rates go “up by the stairs and down by the elevator.”² A strategy that collects small, frequent gains and gives them back in a few sharp drops looks the same on a trade log: many winners and a tail.

Why does the “free trade” profile feel so good on a prop challenge?

Because in some setups it passes more often. We simulated a challenge with a 10% target and a 10% loss limit from the starting balance, 1 R equal to 1% of the account, up to 300 trades and a cost of 0.05 R per trade. A strategy that wins 50% at 1:1 (expectancy −0.05 R) passed 25% of the time and breached the limit 72% of the time. A strategy that wins 90% of trades at +1 R and loses 10% at −9 R (also −0.05 R after costs) passed 53% of the time, with a median of 10 trades. Ten quick wins hit the target before the first big loss, and one 9 R loss does not on its own breach a 10% limit.

The catch is everything after. Many firms also apply a daily loss limit; 5% is a figure commonly seen, and rules differ by firm, so check yours. Under a 5% rule, a single −9 R loss at 1% risk ends the attempt on its own, so passing requires ten wins in a row: 0.9 to the power of 10, about 35%. On a live account, the chance of at least one 9 R loss in 20 trades is 88% (1 − 0.9²⁰). The profile is at its best in the one phase where a fast pass pays and at its worst afterwards. How the rules shape risk-taking is covered in prop firm drawdown rules and how evaluations change risk-taking.

What should I do after a drawdown instead?

Coval and Shumway studied proprietary traders at the Chicago Board of Trade and report that traders appear highly loss-averse and regularly take above-average risk in the afternoon to recover morning losses.³ We only read the abstract, and it concerns proprietary traders at an exchange, not funded-account traders, but the pattern is the one to avoid: responding to a loss by changing the risk profile of the book. A sturdier response is the opposite one:

• Cut size, not rules. Lower variance with smaller position size, which helps whatever the strategy. The reasoning is in Kelly criterion and position sizing.
• Do not change the payoff to raise the win rate. Moving the target closer buys a number, not an edge, as the figures above show.
• Fix stop and target before you enter. Then the sample cannot be filtered by memory; the pull to cut winners early is covered in why traders cut winners early and hold losers.
• Log every trade, including the “obvious” ones you skipped. A setup that is obvious only after it worked is hindsight.
• Compare to the break-even rate, with an interval. A measured 80% over 30 trades has a 95% interval of roughly 63% to 90%; over 100 trades about 71% to 87% (Wilson interval, our calculation). How many trades you need is in how many trades to validate a strategy.
• Watch the largest loss against the average win. If one loss equals ten wins, the profile is the one in the figure above.

What this article does not show

We did not find evidence that “free” trades exist, and we did not prove that none can exist: a real edge is possible, and it shows up as a win rate above the break-even rate for its payoff over a large, honestly logged sample. Our test used random entries to show the geometry; it says nothing about any specific setup. The passing-rate simulation depends on its assumptions (1% risk, 300 trades, costs of 0.05 R) and on firm rules that differ from firm to firm. We did not hear the podcast the question came from and make no claim about what its guest did.

This is educational material, not investment advice.

Related. Prop firm drawdown rules · How evaluations change risk-taking · Kelly criterion · Why traders cut winners early · How many trades to validate a strategy · Do day traders make money?