What Is a Good Profit Factor? It Depends on Your Win Rate

A good profit factor (gross winning dollars divided by gross losing dollars) is one that stays above 1.0 after costs over enough trades to trust, and there is no universal pass mark: the same 1.5 earns 0.30R per trade at a 40% win rate and 0.20R at 60% (R is your average loss). The bigger qualification is sample size: in our computation, a strategy whose true profit factor is 1.3 still prints below 1.0 in 16% to 26% of 50-trade samples, so a net 1.3 over 300 trades says more than a 2.0 over 40.

Key takeaway

Profit factor has no universal pass mark because it blends win rate and payoff: PF = win rate x payoff / loss rate. On Discount Trading's ES fees as of October 2026, our computation shows a 55% one-to-one bracket needs a target above about 3 ticks to beat commissions and about 13 ticks once one illustrative tick of slippage is added. Judge any figure net of costs and against its trade count, because 50 trades cannot separate a 1.3 edge from 1.0.

What is profit factor and how is it calculated?

Profit factor is the total profit of your winning trades divided by the total loss of your losing trades, so 1.0 is break-even and anything above it made money over that sample. The NinjaTrader 8 Help Guide defines it as Gross Profit / Gross Loss and says "It gives you an idea of how much more money your strategy earns than it loses."

With W as your win rate, PF = (W x average win) / ((1 - W) x average loss). Call the payoff b (average win divided by average loss) and it shortens to PF = W x b / (1 - W). A threshold that ignores win rate and payoff is missing two of the three inputs; costs are the third.

Platforms disagree about what goes into the sums. TradeStation's Profit Factor report field "Takes into account commissions (and slippage, if specified for strategies)." NinjaTrader 8 lists Commission as a separate statistic and does not say whether profit factor is net of it, so check your own report before comparing numbers.

What is a good profit factor for your win rate?

A good profit factor depends on your win rate: at 70%, a strategy whose winners are half the size of its losers still shows 1.17, while at 30% you need winners 3.5 times your losers just to reach 1.5. The map below shows the profit factor of each win rate and payoff before costs.

Win rate	Payoff 0.5	Payoff 1.0	Payoff 2.0	Payoff 3.0	Break-even payoff
30%	0.21	0.43	0.86	1.29	2.33
35%	0.27	0.54	1.08	1.62	1.86
40%	0.33	0.67	1.33	2.00	1.50
45%	0.41	0.82	1.64	2.45	1.22
50%	0.50	1.00	2.00	3.00	1.00
55%	0.61	1.22	2.44	3.67	0.82
60%	0.75	1.50	3.00	4.50	0.67
65%	0.93	1.86	3.71	5.57	0.54
70%	1.17	2.33	4.67	7.00	0.43

Computed by Phoenix Technologies; method: PF = W x b / (1 - W) and break-even payoff = (1 - W) / W, with illustrative win rates and payoffs, before costs. A 1.5 sits on a different payoff in every row, so a threshold with no win rate attached tells you little.

Picture the table as a heat grid with every cell at or below 1.0 shaded. The edge of the shaded area runs diagonally, from a break-even payoff of 2.33 at a 30% win rate down to 0.43 at 70%.

The break-even line runs diagonally, so the same profit factor demands a payoff of 2.33 from a 30% trader and only 0.43 from a 70% trader.

The grid settles two common claims. A high win rate does not guarantee a profit: 65% with half-size winners gives 0.93. And a flat ladder such as Defcofx's (January 2026), which calls 1.5 to 2.0 or higher good, has no win rate, cost or sample attached. Trader's Second Brain's guide (updated September 2026) is closer: "A 'good' profit factor is not a universal target."

One claim fails on evidence rather than arithmetic: TradeZella's guide (March 2026) labels 1.5 to 2.0 "Excellent performance. Many professional traders and hedge funds operate here." It cites no source, and we found no public dataset of trade-level profit factors for hedge funds or CTAs, so there is no professional benchmark to check it against.

Profit factor vs expectancy: why do two 1.5s earn different amounts?

Profit factor is a ratio of totals, while expectancy (the average result per trade) is what each trade earns, and the same profit factor can carry different expectancies. Two illustrative strategies with a profit factor of 1.5 show the gap: one wins 40% of its trades and earns 0.30R per trade, the other wins 60% and earns 0.20R.

Expectancy = (1 - W) x (PF - 1) x average loss. At 40%, that is 0.6 x 0.5 = 0.30R; at 60%, 0.4 x 0.5 = 0.20R. The 40% strategy needs a payoff of 2.25 to reach 1.5, the 60% strategy only 1.0. Expectancy in R has its own page in our guide to R-multiples.

Picture two bars on one baseline, both labelled PF 1.5. Over 100 trades they earn 30R against 20R.

Identical profit factors hide a 50% difference in what each trade earns, and the bigger earner needs more trades to prove itself.

The higher earner is also the noisier one. In a simple two-outcome model, the 40% strategy swings with a standard deviation of 1.59R per trade against 0.98R, so its return per unit of swing is 0.188 against 0.204. It needs about 113 trades to reach a t-statistic of 2 (a common bar for telling an edge from noise) against 96. Measured in average wins rather than R, the ranking flips to 0.133 against 0.20.

How do commissions and slippage change profit factor?

Costs come off every winner and add to every loser, so they cut profit factor hardest when targets are small. At Discount Trading's non-member rates as of October 2026, an ES round turn costs $3.76 in commission plus exchange and NFA fees ($1.88 per side), or 0.30 ticks, and an MES round turn costs $1.12, or 0.90 ticks. That is one broker's schedule, not a market average; tick values are in our futures tick value chart.

Take an illustrative 1:1 bracket (target and stop both T ticks) that wins 55% of the time, a pre-cost profit factor of 1.222. With cost c in ticks, net PF = W x (T - c) / ((1 - W) x (T + c)), and the target at which it falls to exactly 1.0 is T = c / (2W - 1). The table adds one illustrative tick of slippage per round turn, because no public source gives a typical slippage for retail stop exits.

Contract and costs	Cost (ticks)	4-tick target	8-tick target	16-tick target	Break-even target
ES, commissions	0.30	1.051	1.134	1.177	3.01 ticks
ES, plus 1 tick slippage	1.30	0.622	0.880	1.038	13.01 ticks
MES, commissions	0.90	0.775	0.976	1.093	8.96 ticks
MES, plus 1 tick slippage	1.90	0.436	0.754	0.963	18.96 ticks

Computed by Phoenix Technologies from Discount Trading's October 2026 fees; method: the formula above at a 55% win rate, pre-cost PF 1.222 in every row. An 8-tick ES scalp that reads 1.22 on a gross report is 0.880 once one tick of slippage is in it.

Picture two lines over target size from 2 to 32 ticks: a flat dashed line at the pre-cost 1.222, and the ES after-cost line climbing beneath it. It starts at 0.259 at 2 ticks, crosses 1.0 at 13 ticks and reaches 1.127 at 32. The gap between the two lines is cost, and it only narrows as the target grows.

Costs are a fixed toll per trade, so a scalp with a target under about 13 ES ticks loses money at a win rate that looks profitable on a gross report.

The insider point is contract size. Ten MES contracts carry the same dollar exposure as one ES, but each MES round turn costs about three times as many ticks, so the same 8-tick scalp drops from 1.134 on ES to 0.976 on MES on commissions alone. The fee stack beyond commissions is in the real cost of funded trading.

Defcofx's advice that intraday traders, who take many small trades, should prefer 2.0 or higher mixes two effects. Small trades raise each trade's cost share, which you fix by netting costs out; frequency itself helps, because more trades narrow the uncertainty band.

How many trades does it take to trust a profit factor?

Usually a few hundred. In our bootstrap, a strategy whose true profit factor is 1.3 prints anywhere from 0.58 to 2.94 over 50 trades and lands below 1.0 in 25.9% of 50-trade samples; only at 300 trades does the 95% range tighten to 0.94 to 1.79.

The realistic case below resamples 10,000 simulated trades with variable sizes; the best case is an exact calculation for a fixed bracket.

Trades	95% range, variable sizes	Below 1.0	95% range, fixed bracket	Below 1.0
30	0.45 to 3.84	30.9%	0.65 to 2.60	29.2%
50	0.58 to 2.94	25.9%	0.73 to 2.31	16.1%
100	0.74 to 2.28	18.4%	0.87 to 1.95	9.7%
300	0.94 to 1.79	5.6%	1.04 to 1.63	1.2%

Computed by Phoenix Technologies; method: seeded bootstrap (seed 20261118, mulberry32 generator), 50% win probability, exponentially distributed sizes rescaled to a population PF of exactly 1.3, 20,000 resamples per row; fixed bracket by exact binomial at a 50% win rate and a payoff of 1.3. A rerun with seed 7 moved each figure by about 1 point or less. Below 100 trades, a genuine 1.3 edge and a break-even strategy produce overlapping numbers.

Picture four horizontal bars on one scale from 0 to 4, one per trade count, with a vertical line at 1.0 and a dot at the true 1.3 on each. The 30-trade bar stretches from 0.45 to 3.84, and the 300-trade bar shrinks to 0.94 to 1.79, still dipping just below 1.0. A thinner bar inside each shows the fixed-bracket best case.

A real 1.3 edge prints below break-even in a quarter of 50-trade samples, and even at 300 trades the realistic range still dips just below 1.0.

This undercuts the common sample ladder: TradeZella's guide says that at 50 trades "Patterns are becoming more reliable", yet a true 1.3 strategy prints below 1.0 in 16.1% to 25.9% of 50-trade samples. Our trading journal guide calls 20 trades per setup tag "a practical minimum before trusting any number, and it is still thin", and a profit factor is less stable than a win rate.

The study covered in our review of ICT research reported 1.50 on 60 holdout trades, which this table says to read with care. For the calendar time an edge takes to confirm, see how long it takes to become profitable.

TradeStation already discounts small samples

TradeStation's Adjusted Gross Profit field "is calculated by looking at the number of winning trades, subtracting the square root, multiplied by the average profit", and Adjusted Gross Loss adds the square root to the losing count instead.

For a true 1.3 strategy at a 50% win rate, the adjusted profit factor reads 0.766 at 30 trades, 0.867 at 50, 0.978 at 100 and 1.104 at 300, crossing 1.0 at about 118 trades. It is a worst-case heuristic, not a confidence interval, but it points the same way as the bootstrap.

When is a profit factor too high?

A profit factor is too high when it is bigger than its trade count and its search process can support, and above 3.0 in a backtest the first question is how it was found. TradeZella's flag on anything above 3.0 is partly right: the 95% range of a genuine 1.3 strategy reaches 3.84 over 30 trades by chance alone, so a short sample explains a high figure as easily as skill does.

Bailey, Borwein, Lopez de Prado and Zhu write in the Notices of the AMS (2014): "We prove that high performance is easily achievable after backtesting a relatively small number of alternative strategy configurations". Because analysts rarely report how many configurations they tried, the authors add, "investors cannot evaluate the degree of overfitting in most investment claims and analysis." The paper works with Sharpe ratios, not profit factors, but the logic applies to any statistic picked from many backtests.

US regulation says the same about hypothetical results. The CFTC's advertising rule, 17 CFR 4.41, requires anyone presenting the simulated or hypothetical performance of a commodity pool operator or commodity trading advisor to attach a cautionary statement, and the text the rule prescribes includes: "Simulated or hypothetical trading programs in general are also subject to the fact that they are designed with the benefit of hindsight."

Harvey and Liu, in The Journal of Portfolio Management (2015), note that halving backtested Sharpe ratios is common practice, then write: "We argue that it is a serious mistake to use the usual 50% haircut." The right haircut is nonlinear, so no flat discount fits every backtest.

Our guide to backtest overfitting covers walk-forward testing, and why sim results differ from live the fills a backtest assumes.

Profit factor vs Sharpe ratio: what does each answer?

Profit factor answers how many dollars you won for each dollar you lost; the Sharpe ratio answers how much return you earned per unit of volatility. William Sharpe (1994) defines the ex ante version of his ratio as "the expected differential return per unit of risk associated with the differential return", with risk measured as the standard deviation.

Profit factor uses only two sums, so it cannot see how wins and losses are spread or in what order they came. Ten losses in a row and the same ten losses scattered through a year give one profit factor and two different drawdowns, so read drawdown beside it.

Do prop firms require a profit factor to pass an evaluation?

No: as of November 2026 we found no futures prop firm that publishes a minimum profit factor as a pass rule. Evaluations are typically judged on a profit target and a drawdown limit, with a daily loss limit, a minimum number of trading days or a consistency rule at some firms, and the profit-factor thresholds in search results are blog recommendations, not firm rules.

Profit factor still matters indirectly, because the target and the drawdown test your expectancy and swings; see profit target vs drawdown. Check the firm's current rules before buying an evaluation.

Is a profit factor of 1.5 good? How to check yours

A profit factor of 1.5 is good when it is net of commissions and slippage, comes from a few hundred trades and was not chosen from many backtest variants; from 40 trades or a gross report it tells you little. Here is the check on an illustrative ES journal of 40 trades: 18 wins averaging 10 ticks and 22 losses averaging 7 ticks, a 45% win rate.

Gross sums. Wins: 18 x 10 ticks x $12.50 = $2,250. Losses: 22 x 7 ticks x $12.50 = $1,925. Gross profit factor: 1.169.
Costs. 40 round turns x $3.76 = $150.40. Net wins: 18 x ($125 - $3.76) = $2,182.32. Net losses: 22 x ($87.50 + $3.76) = $2,007.72. Net profit factor: 1.087.
Expectancy. ($2,182.32 - $2,007.72) / 40 = $4.365 per trade, against $8.125 gross, so costs took 46.3% of the $325 gross edge. The identity (1 - W) x (PF - 1) x average net loss of $91.26 returns the same $4.365.
Sample. 40 trades sits between the 30- and 50-trade rows, where a true 1.3 strategy ranges from 0.45 to 3.84 and from 0.58 to 2.94. A net 1.087 here cannot be told apart from 1.0, and it still excludes slippage.

Picture it as a flow: the trade list splits into wins and losses, each passing a cost gate, and the streams meet at the ratio. From there it forks: "how much per trade" reads $4.365, and "can I trust it" reads 0.58 to 2.94 at 50 trades, a band that includes 1.0.

One trade list answers two different questions: expectancy says what each trade earned, and the sample band says whether the number means anything yet.

Run your own figure through these rules:

If your report is gross or leaves out slippage, recompute it net before comparing it with any threshold.
If you trade a 1:1 bracket, work out the break-even target, cost in ticks / (2W - 1). If your target is under it, your net profit factor is below 1.0; at 55% on ES with one tick of slippage, that line is about 13 ticks.
Compare strategies on expectancy, (1 - W) x (PF - 1) x average loss, not on profit factor.
If you have fewer than 100 trades, treat any profit factor as provisional, whether it reads 0.9 or 2.2: a true 1.3 still spans 0.74 to 2.28 at 100 trades.
If a backtest or a vendor shows above 3.0, ask how many variants were tried and for trades taken after the rules were frozen.
Tag trades by setup and compute the figure per tag as well as overall, so one losing setup cannot hide inside a winning total.

Go deeper

Frequently asked questions

What does a profit factor of 1 mean?

It means gross wins exactly equalled gross losses, so the sample broke even. TradeStation's help puts the rule as "By definition, a value greater than 1 means the trades have a positive net profit." Whether 1.0 is break-even after costs depends on whether your platform netted commissions and slippage into the sums.

Is a profit factor of 2 good?

Over a few hundred trades net of costs, yes: it is well above what a modest edge produces. Over 30 trades it proves much less, because a strategy whose true profit factor is only 1.3 has a 95% range of 0.45 to 3.84 at that count. At a 50% win rate, 2.0 simply means winners twice the size of losers.

Is profit factor the same as the risk-reward ratio?

No. Payoff, often called risk-reward, is the average win divided by the average loss; profit factor multiplies it by win rate over loss rate, PF = W x b / (1 - W). The two numbers match only at a 50% win rate.

Can a strategy with a 30% win rate have a good profit factor?

Yes, if its winners are large enough. At 30%, a payoff of 3.0 gives 1.29 and a payoff of 3.5 gives 1.5, while any payoff below 2.33 loses money before costs. Low win rates also mean bigger swings, so these strategies need more trades before the figure settles.

How do you calculate profit factor from a trade journal or spreadsheet?

Add up the net dollar result of every winning trade, add up the net dollar result of every losing trade as a positive number, and divide the first by the second. Use results after commissions and slippage. The same two sums work on any CSV export of your trades.

Does a profit factor below 1.0 mean my strategy has no edge?

It means the sample lost money, not that the edge is absent. In our bootstrap, a strategy whose true profit factor is 1.3 printed below 1.0 in 30.9% of 30-trade samples and 18.4% of 100-trade samples. Below 100 trades, a figure under 1.0 is a reason to check costs and keep size small, not proof either way.

What profit factor do most profitable traders have?

Nobody can say: no public dataset gives the distribution of profit factors among retail futures day traders. Ranges quoted for successful or professional traders in blog posts carry no data source, so treat them as opinion rather than a benchmark.

Does trade size affect profit factor?

Yes, because the sums are in dollars. NinjaTrader 8 computes gross profit as the sum of profit times quantity over winning trades, so a 5-contract trade weighs five times as much as a 1-contract trade. If your size changed during the sample, the figure leans toward your largest trades; measuring in R removes that tilt.