The share of trades with a stop loss is the proportion of your positions opened with a defined loss limit. It is the only discipline statistic in this glossary that is not declarative: it reads directly from your broker's data, and it cannot be flattered.
Out of 100 trades
Contents
A stop loss is an order closing the position at a price defined in advance. The share of trades carrying one therefore answers a very simple and very blunt question: on how many of your trades had you decided the maximum acceptable loss?
This statistic has a property neither trade grading nor plan adherence has: it is verifiable. The stop sits in the data your broker returns, just like the entry price. There is nothing to tick, nothing to remember, and nothing to adjust after the fact.
It also conditions a good part of the rest of the journal. R-multiple, percent of capital risked and MAE in R are all computed from the distance to the stop: on a trade that has none, those three measures do not exist. A rate of 70% therefore means three important statistics describe only 70% of your activity.
Trades with a stop = trades carrying a recorded stop ÷ total number of trades × 100
A mental stop is not a stop. It appears nowhere, it does not execute if the connection drops, and it therefore does not count in this measure.
A stop placed after opening still counts, but it leaves an unprotected window this statistic says nothing about.
Some brokers do not return the stop in their statement. A sudden 0% is then not a discipline problem but a data source one, and that is checked before concluding.
| Share of trades with a stop | What it describes |
|---|---|
| 100% | The only acceptable figure. Every trade had a maximum loss decided. |
| 90 to 99% | A few exceptions. They are almost always the largest losses in the series. |
| 70 to 90% | Three statistics in your journal describe only part of your activity. |
| under 70% | Risk is not managed, it is endured. No other job makes sense before this one. |
| 0% | Check the source first: many brokers do not return the stop in their statement. |
The second row is the most expensive one, and it goes unnoticed. A few trades with no stop are enough, because they are not trades like the others: a protected trade stops at a decided amount, an unprotected one stops when the trader cannot take it any more. The two populations have nothing in common, and the second almost always contains the worst loss in the record.
The last row calls for a check before any diagnosis. A zero or abnormally low rate very often comes from the broker's statement, which does not always transmit the stop, rather than from the trader's behaviour. Just open a trade you know had a stop and see whether the information arrived.
78% coverage, and a quarter ending negative while the protected trades won. Twenty-two trades were enough to flip the result, and that is the most widespread pattern: trades with no stop are few, and they are the largest, because a trade with no limit stays open exactly as long as the loss keeps growing.
This is the least obvious and most useful use. Three important statistics, average R, percent of capital risked and MAE in R, only exist on trades with a stop. Their value is therefore always read with this rate alongside.
An excellent average R over 60% of trades does not describe a good trader, it describes the protected half of a trader. And the remaining 40% are precisely the ones whose risk you would most need to know.
The calculation fits in one line: the P&L of trades with no stop. It is almost always negative, and often by an amount exceeding the whole period's result, as in this page's example.
It is the kind of figure that requires no interpretation. It does not say trade better, it says do not open a position without knowing where it stops.
The first flaw is not in the measure but in its source. Several brokers do not transmit the stop in their statement, or transmit it only for certain order types. An abnormally low rate therefore deserves a technical check before becoming a reproach, and that check takes a minute.
The second is the mental stop. Many traders believe they have a limit because they know at what price they would exit. That is not a stop: it does not execute when the connection drops, it does not execute overnight, and it moves on its own as price approaches it.
The third is believing this rate measures the quality of the protection. It says nothing about your stop loss strategy: not the stop's placement, not whether it was held, not its size. An account can be at 100% with absurdly wide stops, and other measures will say so.
The classic mistake. Counting a mental stop. It appears nowhere, it does not execute when the connection drops or overnight, and it has the particularity of moving away on its own as price approaches. It is not a loss limit, it is an intention, and this measure counts only what is recorded.
| Indicator | Formula | What it measures | When to use it |
|---|---|---|---|
| Trades with a stop | trades with a stop ÷ total | The existence of a limit, verifiable | Knowing what the rest of the journal describes |
| MAE in R | MAE ÷ risk taken | Whether the stop was HELD | Counting stops that did not work |
| MAE | worst point − entry | Whether the stop was well PLACED | Deciding where to place a stop |
| Plan followed | compliant trades ÷ total | General compliance, declared | Quantifying what indiscipline costs |
An abnormally low rate very often comes from the broker's statement, which does not always transmit the stop. Open a trade you know had one and see whether the information arrived. One minute, before any conclusion.
Almost always negative, and often by an amount exceeding the whole period's result. It is the simplest and harshest figure in this group, and it requires no interpretation.
A stop added three minutes later leaves an unprotected window no statistic accounts for. If your platform allows it, a default stop at opening removes the question entirely.
It is the proportion of your positions opened with a defined loss limit, read directly from your broker's data. Unlike trade grading or plan adherence, it is not declared: it is observed.
No. A mental stop appears nowhere, does not execute when the connection drops or overnight, and has the particularity of moving away as price approaches it. This measure counts only recorded stops.
Because R-multiple, percent of capital risked and MAE in R are all computed from the distance between entry and stop. On a trade with no stop, those three measures do not exist and it is excluded from the calculation. A rate of 70% therefore means those statistics describe only 70% of your activity.
Check the source first. Several brokers do not transmit the stop in their statement, or only for certain order types. Open a trade you know had one and see whether the information arrived. It is only a discipline problem once that check is done.
No, it only guarantees the existence of a limit. It says nothing about your stop loss strategy: not the placement, which MAE measures, nor about whether it was held, which MAE in R reveals. An account can be at 100% with absurdly wide or systematically moved stops.
Tradoshi reads the stop directly from what your broker returns, like the entry price. That is why Trades with a stop loss is the only discipline metric in the Reports screen you do not fill in: there is nothing to tick, and nothing to correct after the fact.
The calculation described above happens in the same screen: crossing net P&L with this metric directly gives what unprotected trades cost. And the Risk management screen lets you set the matching rule, so the question stops being asked trade by trade.
To go deeper on this: read the full guide. Every term is gathered in the glossary.