Profit factor

Profit factor is gross profit divided by gross loss. Add up every winning trade, add up every losing trade, divide the first by the second. Cross 1.0 and the strategy made more than it lost, at least on paper. It's one number, easy to compute, easy to slap on a sales page.

That simplicity is the problem. A strategy can drive profit factor higher without improving at all: just refuse to close losing trades. Let them sit open, underwater, off the books, until they eventually recover or get force-closed at month's end. The realized losses shrink, gross loss shrinks, profit factor climbs. Meanwhile the drawdown the trader actually lived through never shows up in the ratio at all.

The number also says nothing about how the profit arrived. A strategy sitting comfortably above 1.0 could have gotten there from dozens of small, consistent wins, or from a single outsized trade carrying every other loser on the book. Same ratio, completely different risk. Profit factor has no concept of variance or sequencing, and no idea how much capital sat exposed while it worked. For that you need something closer to a Sharpe ratio, which at least divides return by the volatility it took to earn it.

We treat profit factor the way we treat win rate: a seller's favorite, because both numbers look great in a backtest built on convenient assumptions and both degrade fast under a real cost model. Add realistic spread, slippage, and commission per trade, and the razor-thin edges that inflated the ratio start bleeding out one basis point at a time. A profit factor built on a handful of outsized trades is especially fragile. Pull the best trade out of the sample and watch the number collapse.

None of this makes profit factor worthless. It just means one ratio alone can't carry a decision. We log it for every strategy we test, but we never let it stand alone. It sits next to drawdown and a cost-adjusted return before anything gets called a pass. A ratio above 1 tells you a strategy made money in the sample tested. It doesn't tell you how, or whether it will again.