Definition
Profit factor
Profit factor is gross profit divided by gross loss over closed trades. What 1.0, 1.5 and 3.0 actually mean, how costs move it, and the one test that exposes a fake one.
Profit factor is gross profit divided by gross loss across a set of closed trades — equivalently, win rate times average win, divided by loss rate times average loss. A factor of 1.0 is break-even before costs; systems that survive live trading usually sit between 1.3 and 2.0 after fees. A factor above about 3 on a backtest of fewer than a few hundred trades is more likely overfitting or a single outlier trade than a durable edge.
- Also known as
- PF · profit/loss factor · gross profit to gross loss ratio
How it is calculated
Profit factor = Gross profit ÷ Gross loss
= (p × Avg win) ÷ ((1 − p) × Avg loss), p = win rate
p = 0.40, Avg win = 300, Avg loss = 100:
Profit factor = (0.40 × 300) ÷ (0.60 × 100) = 120 ÷ 60 = 2.0
Expectancy per trade = (0.40 × 300) − (0.60 × 100) = 60
Net of 10 per trade in costs, over 100 trades (40 wins at 290, 60 losses at 110):
Net profit factor = 11,600 ÷ 6,600 = 1.76; net expectancy = 50
1.0 = break-even before costs. Size-free, but blind to trade count and loss order.
Worked example: 40% win rate, average win 300, average loss 100 → (0.40 × 300) ÷ (0.60 × 100) = 2.0; with 10 of costs per trade, 1.76.
Expectancy as a ratio, with the same blind spots
Profit factor carries the same information as expectancy, expressed as a ratio instead of currency per trade. At a 40% win rate with an average win of 300 and an average loss of 100, the factor is (0.40 × 300) ÷ (0.60 × 100) = 2.0 and the expectancy is 60 per trade; the ratio is comparable across account sizes, the expectancy is not. The ratio pays for that by discarding trade frequency and the order of losses: two systems with an identical 2.0 can have very different drawdowns, and the Kelly growth rate needs win rate and payoff separately, which no single factor gives back. QANTERION publishes win rate and drawdown cap per strategy on its strategy pages (crypto momentum: 37% win rate, 3% cap) precisely because win rate alone says nothing about the factor.
Net of costs, and minus the best trade
Subtract fees and slippage from every trade before computing the factor. On the same 100 trades, 10 of costs per trade turns 40 wins of 300 into 40 wins of 290 and 60 losses of 100 into 60 losses of 110: the factor falls from 2.0 to 11,600 ÷ 6,600 = 1.76. On a high-frequency system with thin margins a gross factor of 1.6 can be below 1.0 net, which is the whole difference between a strategy and a fee generator. Then remove the single largest winner and recompute; if the factor collapses, the edge was one trade.
How it gets misread
The common way to game the factor is to chase win rate. A martingale-style system reports a 95% win rate and looks superb until the losing run arrives, at which point the gross loss from a handful of trades outweighs every small win and the factor drops below 1: the ratio is set by the size of losses, not their count. The other misreading is quoting a factor from hypothetical results without the trade count. The NFA requires a specific disclaimer on hypothetical performance figures (Interpretive Notice 9025) because such numbers are routinely presented without the context a reader would need to judge them.
See also
Sources
- A New Interpretation of Information Rate — J. L. Kelly, Jr., Bell System Technical Journal (hosted by Princeton University)
- Optimal Gambling Systems for Favorable Games — Proceedings of the Fourth Berkeley Symposium (UC Berkeley Digital Collections)
- Interpretive Notice 9025 — Compliance Rule 2-29: Use of Promotional Material Containing Hypothetical Performance Results — National Futures Association
Definitions are educational. Nothing here is investment advice, and no metric described on this page predicts future results.
Definitions are the easy part
Knowing what drawdown means is not the same as having a system that halts on it. QANTERION applies these limits while a strategy runs.