How to Set a Daily Loss Limit
To set a daily loss limit, choose a figure from your account size before the session, write it down, and enforce it by closing the platform when it is reached. The limit exists precisely for the days when you will disagree with it.
A daily loss limit is a single figure that ends the session when it is reached. It sounds trivial and it is the most consistently useful rule available, because it operates in exactly the circumstances where judgement is least reliable.
Before you start
A number chosen from the account, not from how a session feels. A percentage of the balance, or a multiple of your per-trade risk. Decided when nothing is at stake.
A way to enforce it that does not rely on you agreeing at the time. Closing the platform, a broker setting, a physical break. The enforcement is the rule.
An honest count of what your worst recent day actually was. Most people’s answer is smaller than the record, which is itself worth knowing.
The steps
1. Set the figure from your per-trade risk
Two or three times what one trade risks. That means a bad day is a normal cluster of losses rather than a hole, and it is small enough to make back in a week.
2. Write it down where you will see it
Next to the screen, in the journal, wherever it will be in front of you at the third losing trade rather than in a document you would have to go and find.
3. Decide the enforcement before you need it
Platform closed, positions flat, day over. Some brokers can enforce it directly, which is better than anything relying on your cooperation at the moment it binds.
4. Count it from realised losses only
Open positions fluctuate and would trigger the limit on noise. Count closed trades, and let your per-trade stops handle what is still running.
5. Stop at the number, not near it
“One more, and it is a good one” is the sentence the whole rule exists to prevent. It always sounds reasonable and it is always said at the same point.
6. Add a weekly limit as well
Four consecutive days at the daily limit never breaks the daily rule and is a serious week. A weekly figure catches the pattern the daily one is blind to.
7. Review the number monthly, not during a session
If it is stopping you on ordinary days it may be too tight, and that is a calm-week conversation. Adjusting it while losing is not a review, it is the limit failing.
How to tell it worked
The limit is a written number, decided before the session.
It was reached and observed at least 1 time without being adjusted.
0 trades were taken after it was hit, in the last 90 days.
And a weekly figure exists as well, catching what the daily one cannot.
Why the number matters less than the enforcement
Any reasonable figure works if it is actually applied. A limit that gets overridden on the days it binds provides no protection on exactly the days protection is needed.
Which is why the enforcement mechanism is the real decision. Choosing between two and three times per-trade risk is a detail; choosing something you cannot talk your way past is the rule.
What a bad day actually costs
On this site’s shared series 95% of bars sat below a prior peak, the largest drawdown was 3.76% and the longest recovery took 73 bars. Losing stretches are the ordinary condition rather than an emergency.
Which means the limit is not protecting you from losses. Losses arrive on schedule and the method accounts for them.
It is protecting you from the response to them. A day that runs past the limit is almost never four planned trades that lost; it is two planned trades and four that were taken because the first two lost.
Enforcement mechanisms, ranked
A broker-level lockout is the strongest. Some platforms let you set a maximum daily loss that disables trading when reached, and it cannot be undone until the next session. If yours offers it, use it and stop reading.
A prop-firm rule is next, because breaching it has a consequence you did not choose. That is externally enforced by definition, which is the whole reason those accounts change people’s behaviour.
Physically leaving is third and it works. Close the platform, leave the room, and have decided in advance where you are going. The decision is easier before the third losing trade than after it.
A note in a journal is last and it is close to nothing. It relies on you agreeing with the rule at the exact moment you are least inclined to, which is the one condition every version of this rule was written to avoid depending on.
The original data
Of the 24,971 unique videos in research/search-study-corpus.jsonl, 0 mention daily loss limits in
the title. Risk management generally appears in 410 at a median of 4,079, and revenge trading in a
handful. The counts come from site/corpus_count.py.
0 videos on the single most enforceable rule in trading, against 410 on risk management as a concept. The framework has substantial coverage and the one mechanical thing anybody can implement this afternoon has none.
The answer to the question on that chart is that the setup will happen again and the limit will not. A rule with one exception is a rule with as many exceptions as you need — and the day it feels most worth overriding is the day it is doing the most work.
When it fails
The failure is the limit that has never been tested, and it collapses the first time it binds. The number is written down, the sessions go well, and it never comes up. Then a genuinely bad day arrives: three losses, a strong-looking setup, and a limit that was chosen months ago by somebody who did not know today’s market. Overriding it once feels like judgement rather than a breach — and after the first exception the limit is a preference, which is what it was always going to become without a mechanism.
The second failure is counting floating losses. Noise triggers it.
A third is no weekly figure. Four bad days breaks nothing.
A fourth is adjusting it mid-session. That is the limit failing.
A fifth is a number too large to bind. It never protects anything.
And a sixth is relying on willpower to enforce it. That is the resource already spent.
Related
Risk management is the framework this sits inside. Revenge trading is the behaviour it interrupts. And trading rules covers writing rules that survive contact with a bad day.
The only version that has ever worked for me is the one I cannot argue with. A number in a journal is a suggestion by the third losing trade. Closing the platform, walking away, and having decided in advance that the day is over is the part that actually holds.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.