Daily loss limit (daily drawdown): from balance or starting capital, and when does it reset?
10 min read · Published 28 Oct 2026
A daily loss limit counts from your account at the daily reset, usually from the balance or the higher of balance and equity; equity triggers the breach. Usually the reset is at midnight server time or at 17:00 in New York. That is 23:00 in Central Europe and 22:00 in the weeks when US and European clock changes differ. In our simulation, starting capital versus balance as the base mattered only with a limit of exactly three stops. Pausing for the rest of the day instead of losing the account raised the pass rate from 32% to 54%.
Abbreviations: balance counts only closed trades, while equity adds the running result of open positions. R is a result in units of risk.
Three rules, the same day: a step-by-step example
In our comparison of 43 prop programs (as of 28 Oct 2026), 39 have a daily limit. It is most often 5% (13 programs), 4% (12) or 3% (11). The floor for the day is the account level at the daily reset minus the limit amount. That level is the balance (15 programs), the higher of balance and equity (18) or equity alone (3). In 3 programs the rules do not say. The amount is a percentage of starting capital in 15 programs and a percentage of the level at the reset in 21.
The account is 100,000 and the daily limit is 5%. At the reset the balance is 104,000 and one position is open. You close nothing during the day.
- Case 1: at the reset the position is at +2,000 (equity 106,000), then it falls to −3,500 (equity 100,500).
- Case 2: at the reset the position is at −2,000 (equity 102,000), then it falls to −5,500 (equity 98,500).
| rule | case 1: floor | equity 100,500 | case 2: floor | equity 98,500 |
|---|---|---|---|---|
| balance at reset − 5% of starting capital | 99,000 | 1,500 buffer | 99,000 | breached by 500 |
| balance at reset − 5% of that balance | 98,800 | 1,700 buffer | 98,800 | breached by 300 |
| higher of balance and equity at reset − 5% of it | 100,700 | breached by 200 | 98,800 | breached by 300 |
In case 1 the third rule counts from equity of 106,000. The floating profit from before the reset becomes part of today’s limit. Giving back those 2,000 is a loss of the new day, so the drop of 5,500 breaches the 5,300 limit.
In case 2 all three rules count from the balance of 104,000, because equity is lower. Yesterday’s floating loss eats today’s limit from the first minute. Equity fell only 3,500 since the reset, yet the floor broke, although nothing was closed. With equity alone as the reference, the floor would sit at 96,900.
An amount based on the balance grows after profits and shrinks after losses: at a balance of 96,000 it is 4,800 instead of 5,000.
When the day resets and what time that is for you
The new day for the limit starts at the reset, not at your midnight. In the 39 programs with a daily limit it looks like this:
| reset in the firm’s rules | programs | Central Europe (Poland, Germany) | UK |
|---|---|---|---|
| midnight server time | 14 | usually 23:00, depending on the server clock | usually 22:00 |
| 17:00 in New York | 10 | 23:00, but 22:00 on 8–28 Mar and 25–31 Oct 2026 | 22:00, but 21:00 on the same dates |
| midnight Central European time | 7 | 00:00 all year | 23:00 all year |
| midnight UTC | 4 | 01:00 in winter, 02:00 in summer | 00:00 in winter, 01:00 in summer |
| other or not stated | 4 | — | — |
In 2026 the US moved to summer time on 8 March and Europe on 29 March. In autumn Europe went back on 25 October, and the US follows on 1 November. In those weeks 17:00 in New York is 22:00 in Central Europe and 21:00 in the UK. A position you open in Warsaw or Berlin at 22:30 on Tuesday 27 October therefore already belongs to the new day.
Midnight server time depends on the firm’s broker. At our broker the server clock runs on New York + 7 hours, so its midnight falls at 17:00 in New York. For a trader in Poland the gap to local time is then two hours instead of one for 3–4 weeks a year (chapter 2). Firms usually write “17:00 EST”, although EST is formally winter time. Only 2 of the 10 programs with a 17:00 reset say explicitly that the reset moves with New York in summer.
Our measurement: starting capital, balance or pause
The engine of our challenge calculator ran two sets of trades. The first: a 40% win rate, a 1.5 R win and a 1 R loss, so 0 R per trade. The second: a 44% win rate, so +0.1 R. Trading 4 days a week, 4 trades a day. Phase: 10% target, 4 minimum days, no time limit, a static 10% maximum loss limit. Each cell is 20,000 runs, and in each run the same sequence of trades went through five versions of a 5% daily limit.
Chance of passing the phase, % (in brackets: share of runs ended by the daily limit; column headers: expectancy and risk per trade):
| 5% daily limit | 0 R, 1% | 0 R, 1.5% | +0.1 R, 1% | +0.1 R, 1.5% |
|---|---|---|---|---|
| no daily limit | 50.1 | 48.4 | 80.9 | 72.0 |
| from starting capital, account lost | 50.1 | 31.5 (63.1) | 80.9 | 48.8 (47.7) |
| from balance, account lost | 50.1 | 31.5 (63.1) | 80.9 | 48.8 (47.7) |
| from starting capital, paused until end of day | 50.1 | 53.8 | 80.9 | 76.5 |
| from balance, paused until end of day | 50.1 | 53.2 | 80.9 | 76.2 |
At 1% risk the limit never triggered, because four stops make 4%. At 1.5% four stops already make 6%. With no edge, a day with four losses comes roughly every eighth trading day. The limit then ended 63 runs in 100, and the chance fell from 48% to 32%. A +0.1 R edge did not protect: 72% against 49%.
With account loss, starting capital and balance gave the same result in every pair of runs. With a balance between 90,000 and 110,000, the balance-based amount is between 4,500 and 5,500, and the day’s loss grows in steps of 1,500. A loss of 4,500 therefore fits both versions, and 6,000 breaches both.
A difference appears when the limit holds exactly as many stops as you plan. Take a 3% limit, three trades a day and 1% risk. With the amount based on starting capital, three losses end the day exactly on the floor. With no edge, 51% of runs then passed. From the balance, the limit is smaller below starting capital, for example 2,940 at 98,000, and the chance fell to 32%. But a loss of 1.02% instead of 1% is enough, which is 0.02 R of cost or slippage. Then the chance fell to 10% from starting capital and to 23% from the balance. Above starting capital, the balance-based limit grows.
The pause beat having no limit, because the engine closes positions exactly on the floor and so cuts off the worst days. In “cautious” mode every position goes to its stop before it closes, including winners. Then, at 0 R and 1.5%, the pause gave 41%, no limit 45% and account loss 21%.
This is a simulation with an assumed distribution, not the history of real accounts: every trade ends at exactly +1.5 R or −1 R. The engine does not calculate the “higher of balance and equity” rule, because it closes all positions within the day.
What typical guides don’t tell you
A morning profit grows your buffer. The floor for the day stays put from reset to reset. In the first rule of the example, with a balance of 104,000, it sits at 99,000. After a morning profit of 3,000 the balance is 107,000, so 8,000 is left to the floor, not 5,000.
A single position can breach the limit on its own. In our carry test, a stop on the daily close did not protect against a move within one session. The worst position lost 6.8 R (chapter 38). At 0.5% risk that is 3.4% of the account, more than a 3% limit. The limit sees equity at every moment, so what counts is the stop order in your terminal, not an exit plan.
A limit of exactly N stops has no buffer. Cost and slippage push the N-th loss past the floor, and from the balance the limit is even smaller below starting capital. Count your stop together with its cost.
Check it yourself
The challenge calculator. In the “Parameters” tab enter: “Win rate %” 40, “Average win (R)” 1.5, “Average loss (R)” 1, “Trades per trading day” 4, “Trading days per week” 4, then “Use these parameters”. Rules: “Number of phases” 1, “Profit target %” 10, “Min. number of days” 4, “Daily loss limit %” 5, “Based on” starting balance, “Maximum loss limit %” 10.
In the “Odds at different risk levels” table, in the 1.50% row, the “Pass” column shows 30% and the “Daily limit” column 67%. Switch “When the daily limit is hit” to “paused until end of day” and you will see 55%. The calculator uses 2,000 runs, hence gaps of up to 4 points against our table. With 3 trades a day and a 3% daily limit, the 1% row shows 50% from starting balance and 30% from the balance.
The program comparison shows, next to each daily limit, what its amount is based on.
A tool in the terminal. During the phase, it helps to have a tool that calculates the day’s floor from equity under the firm’s rule and knows the reset time.
Questions
Is the daily loss limit based on balance or equity?
The breach is checked on equity, so the loss on open positions counts. Of the 39 programs in our comparison, 29 say so explicitly. The reference is the balance at the reset (15 programs), the higher of balance and equity (18) or equity alone (3).
What time does the daily loss limit reset?
That depends on the firm: at midnight server time, Central European time or UTC, or at 17:00 in New York. The last one is usually 23:00 in Central Europe and 22:00 in the UK. From 8 to 28 March and from 25 to 31 October 2026 it is an hour earlier.
Does an open position count towards the daily limit?
Yes, its loss enters equity before you close it. Held through the reset, it carries its result into the new day. Yesterday’s floating loss shrinks today’s buffer. Under the “higher of balance and equity” rule, a floating profit raises the floor.
What happens if you exceed the daily loss limit?
In 37 of 39 programs the account is lost; in two, trading stops until the end of the day. In our simulation the pause raised the pass rate from 32% to 54%, and in “cautious” mode from 21% to 41%.
Educational content, not investment advice. Past results do not guarantee future results. The numbers in the example are made up for the arithmetic. The simulation describes an assumed distribution of trades, not real accounts or the terms of any particular firm. Firms change their rules, so check them at the source.