What Is Profit Factor in Trading? (And Why It Matters More Than Win Rate)
Win rate is the metric most new traders fixate on, and it's often the least useful one. Profit factor tells you what actually matters: whether a strategy makes money.
The definition
Profit factor (PF) is gross profit divided by gross loss. A PF of 1.5 means the strategy makes $1.50 for every $1.00 it loses, across all trades combined. A PF below 1.0 means the strategy is losing money overall, regardless of how often it "wins."
Why win rate alone is misleading
Consider two strategies:
- Strategy A: 70% win rate, but wins average +0.5R and losses average -1.5R. Net expectancy: (0.7 × 0.5) - (0.3 × 1.5) = 0.35 - 0.45 = -0.10R per trade — a net loser, despite winning most of the time.
- Strategy B: 40% win rate, wins average +2.5R, losses average -1R. Net expectancy: (0.4 × 2.5) - (0.6 × 1) = 1.0 - 0.6 = +0.40R per trade — solidly profitable, despite losing most of the time.
A high win rate with small wins and large losses can lose money. A low win rate with well-managed risk and larger wins can be very profitable. Profit factor captures this relationship; win rate on its own doesn't.
What counts as a "good" profit factor
Above 1.0 is technically profitable but often not enough margin to survive real-world costs — spread, slippage, commission. A reasonable minimum bar for a strategy intended for live trading is around 1.3, with meaningful sample size behind it (100+ trades). Profit factors above 2.0 are strong; anything dramatically higher (5, 10, 50+) on a retail strategy is usually a sign the backtest was scored with a shortcut that inflated the numbers, not evidence of an extraordinary edge.
The trap: profit factor from a thin sample
A profit factor of 3.0 means very little if it comes from 15 trades. Small samples are dominated by variance — a couple of lucky trades can produce an eye-catching PF that has no real predictive value. Always check profit factor alongside sample size, not instead of it.
Meridian Signals — 13 ICT forex strategies, each validated on a 10-year walk-forward before publishing a single signal. $87/mo, 30-day no-conditions refund.