Trading · Risk management
Daily loss limits: when to stop trading for the day
What a daily loss limit does and does not do: a simulation of a steady trader and a tilting one with and without a limit, and how to set and keep the number.
On this page
A daily loss limit is a rule a trader sets before the session: if losses for the day reach a fixed amount, trading stops until tomorrow. It is the personal version of a circuit breaker. Exchanges halt the whole market after a large fall so that participants can stop and think (Market-wide circuit breakers: the 7%, 13% and 20% halt rules). A daily limit does the same for one person.
The rule is often described as if it improved a strategy. It does not, and seeing why makes it easier to use well. A limit cannot add edge to a method. What it can do is cap the worst day and, more importantly, stop a trader at the moment when their decisions are most likely to get worse. This lesson separates those effects with a simulation, then covers how to choose the number and how the rule usually fails.
What the rule is
A daily loss limit has three parts.
- A measure. Realized losses only, or realized plus open losses. Counting open losses is stricter and harder to game.
- A threshold. A dollar amount, a percentage of the account, or a number of R (the planned risk on one trade, Risk-reward ratio and R multiples: the math of a trade). Typical statements are "3R a day" or "2% of the account a day".
- A consequence. No new trades until the next session. Stricter versions also close open day trades.
Variations use the same structure with a different trigger.
| Rule | Trigger | What it guards against |
|---|---|---|
| Daily loss limit | Day's loss reaches a set amount | One catastrophic session |
| Consecutive-loss stop | A set number of losing trades in a row | Trading through a state of mind |
| Maximum trades per day | A count of trades | Overtrading |
| Weekly or monthly loss limit | Period loss reaches a set amount | A slow slide across many sessions |
| Give-back rule | A set share of the day's peak profit is lost | Turning a good day into a losing one |
Elder's 6% rule, described in Portfolio heat: adding up the risk of every open trade, is a monthly version: no new trades for the rest of the month once losses and open risk together reach 6% of equity.
Firms apply the same idea from outside. In the United States the SEC's Rule 15c3-5, adopted in 2010, requires a broker-dealer that gives access to the markets to maintain risk controls, including pre-set credit or capital thresholds for each customer and controls that reject orders exceeding them. Proprietary trading firms commonly add a daily loss level at which a trader's access is cut off for the session. A personal limit is the same control without the enforcement.
What a limit cannot do
Suppose every trade has the same positive expectancy and the outcome of one does not affect the next. Then the trades that would have followed a bad start are as good as any others. A rule that skips them removes positive-expectancy trades, and the average must fall.
The simulation makes this concrete. A "steady" trader takes up to eight trades a day. Each wins 1.2R half the time and loses 1R the other half, an expectancy of +0.10R per trade. With no limit, that is +0.8R a day on average.
- Without a limit: average day +0.79R; the worst day in a hundred was −5.8R; 12% of 21-day months lost money.
- With a stop at −3R: average day +0.72R; the worst day in a hundred was −3.8R; 15% of months lost money.
The limit did what it says. It cut the bad tail of the daily results by about a third. It also cost about 9% of the average day, because on roughly one day in seven the trader stopped early and missed trades that were, by assumption, as good as the morning's. Slightly more months ended negative.
So if a trader is truly unaffected by losses, a daily limit is a small tax paid for a smoother ride. That is an honest reason to use one. It is not the main reason.
What a limit can do
The assumption that trades are identical after losses is the weak point, and research on professionals contradicts it. Coval and Shumway studied proprietary traders in the Chicago Board of Trade's Treasury bond futures pit. Traders with losses in the morning took above-average risk in the afternoon. Thaler and Johnson found the same pull in experiments and named it the break-even effect: after a loss, bets that offer a chance to get back to even become unusually attractive. Traders call the resulting state tilt, and Trading psychology: loss aversion, overconfidence and other biases covers the biases behind it.
The second half of the simulation adds a simple version of that behavior. The "tilting" trader is the same as the steady one until two losses in a row. From then until the next win, the trader takes worse trades (40% winners instead of 50%) at double size.
| Trader and rule | Average day | Trades per day | Worst day in 100 | Losing months |
|---|---|---|---|---|
| Steady, no limit | +0.79R | 8.0 | −5.8R | 12% |
| Steady, stop at −3R | +0.72R | 7.3 | −3.8R | 15% |
| Tilting, no limit | +0.23R | 8.0 | −14.0R | 39% |
| Tilting, stop at −3R | +0.33R | 6.9 | −4.8R | 34% |
| Tilting, stop at −2R | +0.39R | 5.9 | −3.8R | 30% |
Three results matter.
Tilt is expensive. The tilting trader has the same method as the steady one and keeps less than a third of its average day (+0.23R against +0.79R). The bad trades are few. Their size does the damage.
For this trader the limit adds money. A stop at −3R raised the average day by about 45% and a stop at −2R by about 70%. The limit works here because it removes trades whose expectancy really is negative: the ones taken in the tilted state.
The tail shrinks dramatically. Without a limit the worst day in a hundred was −14R, the result of two losses followed by six double-size losses. With the −2R limit it was −3.8R.
Even with the tightest limit the tilting trader earned half of what the steady trader did. A limit contains the damage. It does not repair the cause.
The simulation is a sketch. The tilt rule is invented for illustration, and real behavior after losses varies from person to person and from day to day. What it shows reliably is the logic: a loss limit pays exactly to the extent that trades after losses are worse than trades before them.
Setting the number
A workable limit meets two conditions.
It sits beyond an ordinary bad day. A limit that triggers on normal variation ends too many sessions and costs too many good trades. Start from the record. With eight even-odds trades a day, ending at −3R or worse takes two or fewer wins, which happens on 37 of every 256 days, about one in seven. A trader who finds that too frequent can move the limit to −4R. A discretionary trader can read the same thing from a journal: the limit belongs just past the loss seen on a typical red day.
It sits well inside the loss that would threaten the account. In money, the limit is the number of R times the risk per trade. At 1% risk a −3R day costs 3% of the account. Five such days in a row cost about 14%. A trader should check that a full week of limit days leaves the plan intact, using the drawdown tables in Losing streaks and drawdown recovery: what a trader should expect.
For swing traders, who place few trades a day, the natural unit is the week or the month, and the limit usually counts open risk as well as realized losses.
How the rule fails
- Moving it mid-session. A limit changed at the moment it would bind was never a limit. The number is set when calm and applies when not.
- Sizing up to avoid it. Doubling the next trade to get back above the line before it triggers is the break-even effect in action, and it converts a controlled day into the −14R row of the table.
- Treating it as a budget. A trader who thinks "I have $600 to lose today" takes marginal trades until it is gone. The limit is a ceiling for bad days, not an allowance for ordinary ones.
- Counting only closed trades. Holding a large open loss so that it does not count toward the limit defeats the rule and adds the risk of an even larger loss.
- No rule for the next day. A limit ends one session. Without a review of what happened, the same state of mind arrives at the next open.
Enforcement helps. Where a broker or platform offers account-level loss or order-size limits that cannot be lifted during the day, they do the job mechanically. Without that, the practical substitutes are physical: closing the platform, leaving the desk, and writing the day's review before placing another order (Trading plan and trading journal: what to write down).
Limits compared
| Tool | Scope | Stops | Does not stop |
|---|---|---|---|
| Stop-loss on a trade | One position | A single trade's loss growing | Many small losses in a row |
| Portfolio heat cap | All open positions | Too much risk open at once | Losses realized one after another |
| Daily loss limit | One session | A day of compounding mistakes | A slow decline across weeks |
| Monthly limit or drawdown rule | The account | Trading on through a failed method | Anything within a single day |
The four work at different scales and are usually used together.
How Finabase does it
Finabase is a research and paper-trading tool and does not enforce trading limits.
- Paper trading. The paper account records every order and shows the account's value and its history (Practise with Finabase paper trading on the web). It will not stop you at a loss level, which makes it a fair test: write the limit down, trade a session, and see whether you held to it when nothing forced you to.
- Alerts. Price alerts reduce the need to watch every tick, which for many people is where overtrading starts.
- Portfolio. A portfolio shows the value and gain of the positions you enter at live prices, the figures a swing trader would compare with a weekly or monthly limit.
Frequently asked questions
What is a good daily loss limit?
One that is just beyond a normal losing day and small enough that several in a row do not endanger the account. Many day traders state it as two to four times their risk per trade, which at 0.5% to 1% risk per trade is roughly 1% to 4% of the account. The right figure depends on how many trades are taken a day and on the account's own record.
Does a daily loss limit make a strategy more profitable?
Not by itself. If trades after losses are as good as any others, the limit slightly lowers the average result while cutting the worst days. It raises results only when it prevents trades that are worse than usual, which is the common case for traders who press after losses.
Should the limit include open positions?
Counting open losses is stricter and closes a loophole, because otherwise a trader can avoid the limit by not closing a losing trade. Day traders usually count everything. Swing traders often use a weekly limit on realized losses plus a cap on open risk.
What should I do after hitting the limit?
Stop placing trades, and review the session while it is fresh: which losses followed the plan and which did not. The purpose of the pause is to separate bad luck from bad decisions before the next session.
Summary
A daily loss limit ends trading for the day at a preset loss. Arithmetic says it cannot improve a strategy whose trades are equally good at all times: in the simulation it trimmed a steady trader's average day from +0.79R to +0.72R while cutting the worst days by a third. Its real purpose is behavioral. Evidence from professional traders shows more risk-taking after losses, and for a simulated trader who presses after two straight losses a −2R limit raised the average day by about 70% and reduced the 1-in-100 bad day from −14R to −3.8R. The number belongs beyond a normal bad day and well inside a dangerous one, and it protects only if it is fixed before the session and not negotiated during it.
Check your understanding
4 questions on this lesson. Each answer comes with the reason, and nothing is scored.
A trader’s decisions are exactly as good after losses as before, and every trade has a positive expectancy. What does a daily loss limit do to the average daily result?
- Raises it, because it prevents losses
- Lowers it slightly, because it skips some trades that had a positive expectancy
- Leaves it unchanged
Show the answer
B. Lowers it slightly, because it skips some trades that had a positive expectancy If trades after a bad start are as good as any others, stopping early removes good trades. In the simulation the steady trader’s average day fell from +0.79R to +0.72R, in exchange for a smaller worst day.
For which trader does a daily loss limit increase the average result?
- One who takes worse or larger trades after a run of losses
- One who never changes behavior
- One who trades only once a day
Show the answer
A. One who takes worse or larger trades after a run of losses The limit pays to the extent that trades after losses are worse than trades before them. It removed the negative-expectancy trades of the tilting trader and raised that trader’s average day by 45% to 70%.
An account of $25,000 risks 1% per trade and uses a 3R daily limit. What is the limit in dollars, and roughly what would five limit days in a row cost?
- $250 a day; about 5% of the account
- $750 a day; about 14% of the account
- $2,500 a day; about 50% of the account
Show the answer
B. $750 a day; about 14% of the account 1R = $250, so 3R = $750, or 3% of the account. Five such days compound to 1 − 0.97^5 ≈ 14%. A limit should be checked against a run of limit days, not only one.
Near the daily limit, a trader doubles the size of the next trade to get back above the line. What does the lesson call this, and why is it harmful?
- Scaling in; it lowers the average entry price
- The break-even effect; it is the behavior the limit exists to stop and it produces the largest losing days
- Hedging; it reduces risk
Show the answer
B. The break-even effect; it is the behavior the limit exists to stop and it produces the largest losing days After a loss, bets that offer a chance to get back to even become unusually attractive. Sizing up to avoid the limit turns a contained day into the kind of −14R day the simulation shows for an unlimited tilting trader.
Sources and further reading
- Joshua D. Coval and Tyler Shumway, “Do Behavioral Biases Affect Prices?,” Journal of Finance 60(1), 2005
- Richard H. Thaler and Eric J. Johnson, “Gambling with the House Money and Trying to Break Even,” Management Science 36(6), 1990
- SEC — Risk Management Controls for Brokers or Dealers with Market Access, Rule 15c3-5 (Release No. 34-63241, November 3, 2010)
- Alexander Elder, Come Into My Trading Room (Wiley, 2002) — the 6% rule
- Finabase simulation — 40,000 trading days per setting, fixed seed (scripts/learn/figures/trading-risk-depth.ts)
