Tight stop, expensive stop: what 4,893 trades say about the stop-loss
The usual rule of thumb says the stop belongs closer than the target. Aim for 7%, get out at 4%, and on average you win more than you lose. It sounds compelling. I ran the numbers, and it does not hold.
What was measured
4,893 purchases of US stocks between September 2022 and August 2026, all picked by the same pattern: the last quarterly results beat expectations by at least 10%, the price stood in the lower third of its range of the past twelve months, and it had fallen by at least 5% over the five days before. Buy at the next day's open; sell at the target, at the stop, or after 20 trading days at the latest.
Target and stop are not set in fixed percent but in multiples of the stock's usual daily swing — the average distance between the day's high and low over the last 20 days. For these stocks it averaged 4%. The target always sits at three times that, roughly 12%. Only the stop changes.
The assumptions lean to the cautious side: if target and stop fall on the same day, the stop counts. If a day opens below the stop, the worse opening price applies. Each trade pays 0.2% in costs.
The result
| Stop | roughly | stop triggered | target reached | return per trade |
|---|---|---|---|---|
| 1× daily swing | −4% | 69% | 25% | +0.21% |
| 1.5× | −6% | 56% | 31% | +0.32% |
| 2× | −8% | 45% | 34% | +0.62% |
| 3× | −12% | 26% | 37% | +0.92% |
| 4× | −16% | 14% | 38% | +1.15% |
| 6× | −24% | 4% | 38% | +1.30% |
| no stop | — | — | 38% | +1.34% |
Every row down is better than the one above. The tightest stop triggers in two out of three trades and leaves the least of the return.
Why
A stock that usually swings 4% in a day reaches −4% on a perfectly ordinary day. A stop at that distance does not measure whether the idea was wrong. It measures whether the stock moved. And this selection bets on a recovery after a drop — recoveries rarely run in a straight line. The tight stop sells into the dip the stock then climbs out of.
The effect is largest for the most restless stocks. In the third with the highest daily swing, a stop at 1.5× returned +0.17% per trade; at 4× it returned +1.45%.
So the arithmetic behind the rule of thumb is only half right: a tight stop keeps each loss small. It also turns trades that would have won into losses — and that weighs more.
What a wide stop costs
It limits the loss later. The worst 5% of trades ended at −11% or worse with the stop at 1.5×, and at −19% with the stop at 4×. Overnight gaps jump over any stop, which is why both figures sit below the stop itself.
Whoever stops wider has to buy smaller. The risk of a trade is position size times distance to the stop. A stop at −16% with half the position risks the same amount as one at −8% with the full position. The stop is then no longer a tool for keeping losses small but a guard against a crash.
What the numbers do not say
- They hold for this selection. For stocks bought for other reasons it may look different.
- Stocks that were delisted during those four years are missing from the price data. Those are exactly the cases where a stop saves you. "No stop" therefore looks better than it was.
- In falling markets the selection does not work, with any stop: in 2022 it lost per trade no matter how wide the stop was.
- Four years is not much. The wide stop (4×) beat the tight one (1.5×) in each of the five calendar years; the size of the returns varied a lot.
- Not investment advice. The calculation shows what was — not what will be.
How this shows up in rulio
For buy candidates, rulio suggests target and stop from the stock's daily swing instead of fixed percent: three times up, four times down. Both can be changed in the order ticket.
Sources
- Own analysis: daily prices (open, high, low, close) of all liquid US stocks via Alpaca, reported quarterly results from the Nasdaq calendar; September 2022 to August 2026.