Why not a 10% stop on everything?
Short · Exits & selling · October 11, 2026
Olympia publishes educational content for learning purposes only. Past performance is not indicative of future results. Nothing here is investment advice.
Because ten percent is a different bet on every stock. On calm names, eighty-five of every hundred winners never dipped past about seventeen percent before the move paid; on the wildest third, about forty-three. So you measure the stop in the stock's own range instead.
The lesson in writing
Why not just use a ten percent stop on everything? Because ten percent is a different bet on every stock.
What it is
On calm names, eighty-five of every hundred winners never dipped past about seventeen percent before the move paid. On the wildest third, about forty-three.
So one flat number is miles away on one name and inside the ordinary wiggle on another — which is the whole problem with it.
How it's worked out
You measure the stop in the stock's own ATR — average true range, how far it normally moves in a day. It's the same ruler behind the entry and target bands.
On this one, about three ATRs down stopped out twenty-five of every hundred winners; seventy-five percent didn't. Four and a half ATRs down, fifteen out, eighty-five didn't. Five and a half, ten out, ninety didn't.
That's the protection going in: the stop sits outside the stock's normal wiggle, so on an ordinary day it can't just take you out.
Where Olympia calculates it
Olympia calculates it where you log a trade — the ATR stop menu, with the counts in the box under it.
The counts come from the measured ladder and the multiples come from the name's own range, so the same rung is a different number of ATRs on a calm stock than on a wild one.
The dip figures: Olympia's stop-width study, 24 August 2026 — served-board winners, split in thirds by ATR, measured at the depth that eighty-five of every hundred never went past before the move armed. The stopped-out counts are the measured ladder's three rungs. The ATR multiples are an example name's, on an example fill. Historical, not predictive.
