Home / Algorithmic Risk / Daily Loss Limits
Loss controlIntermediate

Daily Loss Limits

A daily loss limit is a pre-set floor on a day's cumulative loss at which all trading stops for the rest of the session, capping how much a bad day, whether from a losing streak or a malfunction, can cost.

Quick answer: A daily loss limit is a pre-set floor on a day's cumulative loss at which all trading stops for the rest of the session, capping how much a bad day, whether from a losing streak or a malfunction, can cost.

In simple words

A daily loss limit is a line in the sand for a single trading day: once your losses hit it, you stop trading until tomorrow. It exists because bad days cluster, one loss triggers a worse decision, which triggers another, and an automated system can run that spiral at full speed. Setting the limit in advance, when you are calm, means the decision to stop is already made before the painful moment arrives. It is the daily version of not letting one bad session turn into a catastrophe.

Purpose

A daily loss limit exists to contain the damage of a single session, because losses cluster and an unmonitored algo can keep executing a losing pattern all day, turning a routine bad day into an account-threatening one.

Visual explanation

Daily Loss Limits

A day's equity falling to the preset loss floor, at which the horizontal halt line stops further trading for the session.

Drawdown CurveEquity0% — high-water markmax drawdowntime →

Professional explanation

Why the day is the right unit to cap

Losses are not evenly distributed; they cluster into bad sessions driven by adverse regimes, volatility spikes or a strategy that has simply stopped working today. Capping the day contains this clustering, because it stops the strategy before a single session's damage compounds into a drawdown that takes weeks to recover. The daily limit also interacts with human and machine tilt: after a run of losses, both a discretionary trader and a naively coded martingale-style algo tend to press harder, and a hard daily floor removes that option. It turns an open-ended bad day into a bounded, known worst case.

The formula and how to set the level

The rule is simply: halt when the day's realised-plus-unrealised P&L reaches minus (Loss percent times capital). Setting the percentage is the real work. Too tight and normal intraday noise stops the strategy on days it would have recovered, wasting edge; too loose and the limit fails to protect. A common heuristic is to anchor the daily floor to a small multiple of the average winning day or to a fraction of the maximum drawdown budget, so a single stopped day is recoverable within a few normal days. The figure is a risk-appetite choice, not a law, and should be justified against the strategy's own daily P&L distribution.

Realised, unrealised, and what counts

A subtle design question is whether the limit measures realised losses only or also marks open positions to market. Counting only realised losses can let a large unrealised loss sit open past the point the limit was meant to protect, while counting unrealised losses can trip the halt on a temporary mark that would have recovered. Most robust designs mark open positions to market so the limit reflects true economic exposure, and pair the halt with a rule on whether existing positions are held or flattened when the floor is hit. The limit must also define its baseline precisely: loss from the day's open equity, not from an intraday peak.

Halt behaviour: stop entries or flatten the book

Hitting the limit raises the same hold-versus-flatten question a kill switch faces. The gentler version blocks new entries but lets existing positions run to their own stops, which avoids forced exits into bad liquidity but leaves residual risk. The stricter version flattens everything, guaranteeing the loss stops growing but accepting slippage and the chance of exiting just before a recovery. Which is right depends on the strategy: a mean-reversion book may be harmed by flattening at the worst moment, while a trend book with wide stops may prefer a clean flat. The choice should be pre-decided and coded, not improvised.

Interaction with position sizing and streaks

A daily loss limit is not a substitute for per-trade sizing; it is a backstop above it. If each trade risks 1 percent, a 3 percent daily limit is roughly three full losing trades, meaning the limit and the sizing must be set together so the daily floor corresponds to a sensible number of normal losers. Set the per-trade risk too high and the daily limit is hit on two trades, making the strategy fragile; set it sensibly and the daily limit only bites on a genuinely abnormal day. The daily limit and the trade-level risk are two rungs of the same ladder and should be designed as a pair.

Formula

Halt trading when Day P&L ≤ −(Loss% × Capital)

Day P&L = realised plus marked-to-market unrealised profit and loss since the day's opening equity (in ₹); Loss% = the fraction of capital allowed to be lost in one session (a risk-appetite choice, e.g. 0.02–0.04); Capital = account equity at the day's open in ₹. When Day P&L falls to the negative threshold, block new entries and apply the pre-decided hold-or-flatten rule.

Hard vs soft daily loss limit

AspectHard limitSoft limit
EnforcementAutomatic halt, no overrideAlert or guideline the trader can bypass
Protects againstTilt, malfunction, revenge tradingOnly honest, disciplined operators
Failure modeMay stop on a recoverable dayIgnored exactly when it matters most
Best forAutomated strategiesExperienced discretionary traders, with caution

Practical example

Illustrative example (Indian market)

A strategy runs on Rs 5,00,000 with a daily loss limit set at 3 percent, so the floor is minus Rs 15,000. Each trade is sized to risk about 1 percent, roughly Rs 5,000, trading Nifty (Rs 75 per point). On a choppy expiry day the strategy takes three losses of about Rs 5,000 each as the market whipsaws; day P&L reaches minus Rs 15,000 and the limit trips, blocking further entries. Without the limit the same conditions might have produced two or three more losing trades before the close, deepening the day to minus Rs 25,000 to Rs 30,000, a 5 to 6 percent hole needing several clean days to recover. The limit converted an ugly but survivable day into a bounded one.

For an F&O algo, marking to market matters because SPAN margin and MTM are settled through the day, so a large unrealised loss on a short-option position can trigger a margin call before any daily limit you coded. Set the daily loss floor comfortably inside the level at which the broker would auto square-off, so your own control acts first rather than the broker's.

Limitations

  • Set too tight, it stops the strategy on normal noise days it would have recovered, wasting edge
  • Counting only realised P&L can leave a large unrealised loss open past the intended floor
  • It caps one day but says nothing about a slow bleed of small losses across many days
  • Flattening at the floor can crystallise a loss just before a recovery, especially for mean-reversion
  • A soft limit protects nobody if the operator overrides it under the emotional pressure of a losing day

Common mistakes

  • Measuring the day's loss from an intraday peak instead of the day's opening equity
  • Setting the daily floor without reference to the strategy's own daily P&L distribution
  • Making the limit a soft alert that can be, and always is, overridden on a bad day
  • Ignoring unrealised marks so a big open loss sits past the limit until the broker acts
  • Setting per-trade risk so high that two losers breach the daily limit, making the strategy fragile
  • Resuming trading the same session after a manual override once the pain eases

Professional usage

Professional desks encode the daily loss limit as a hard, non-overridable halt tied to marked-to-market P&L from the day's opening equity, not an intraday peak. They calibrate the level against the strategy's daily P&L distribution so it corresponds to a genuinely abnormal day rather than routine noise, and set it inside the broker or margin auto-square-off level so their own control fires first. The halt behaviour, block entries versus flatten, is pre-decided per strategy, and the daily limit is designed jointly with per-trade sizing so the two rungs are consistent.

Key takeaways

  • A daily loss limit caps a single session so a bad day cannot become an account-threatening drawdown
  • Halt when day P&L reaches minus (Loss percent times capital), measured from the day's opening equity
  • Mark open positions to market so a large unrealised loss cannot sit past the intended floor
  • Make it a hard, non-overridable halt for an algo, and design it jointly with per-trade sizing

Frequently asked questions

What is a daily loss limit?
A daily loss limit is a pre-set floor on a single day's cumulative loss at which all trading stops for the rest of the session. It caps how much one bad day, from a losing streak or a malfunction, can cost, converting an open-ended bad session into a bounded worst case.
How do I calculate a daily loss limit?
Halt when the day's profit and loss reaches minus (Loss percent times capital). For example, a 3 percent limit on Rs 5,00,000 capital sets the floor at minus Rs 15,000. The percentage is a risk-appetite choice that should be justified against the strategy's own daily P&L distribution.
What is a good daily loss limit percentage?
There is no universal figure; a common heuristic anchors the daily floor to a small multiple of the average winning day or a fraction of the maximum drawdown budget, so a stopped day recovers within a few normal days. Too tight wastes edge on noise, too loose fails to protect.
Should a daily loss limit count unrealised losses?
Most robust designs mark open positions to market so the limit reflects true economic exposure. Counting only realised losses can let a large unrealised loss sit open past the point the limit was meant to protect, defeating its purpose.
Should hitting the limit flatten positions or just stop new trades?
It depends on the strategy. Blocking new entries but letting positions run to their stops avoids forced exits into bad liquidity but leaves residual risk, while flattening guarantees the loss stops growing but risks slippage and exiting before a recovery. The choice should be pre-decided and coded.
Why does an algo need a daily loss limit more than a human?
Because an unmonitored algo can keep executing a losing pattern all day at full speed, and a naively coded strategy may even press harder after losses. A hard daily floor removes that option and stops the machine before a routine bad day compounds into a serious drawdown.
Is a daily loss limit the same as a stop-loss?
No. A stop-loss exits a single position at a price level, while a daily loss limit halts all trading once the whole day's losses reach a floor. A stop-loss manages one trade; the daily limit caps the aggregate of a session, including a string of separately stopped trades.
Should a daily loss limit be measured from the day's open or from a peak?
From the day's opening equity. Measuring from an intraday peak makes the limit trip after a normal give-back from a good morning, which is not what it is meant to protect against. The baseline is the open, so the floor reflects a true losing day.
What is the difference between a hard and a soft daily loss limit?
A hard limit is an automatic, non-overridable halt, while a soft limit is an alert or guideline the trader can bypass. Hard limits protect against tilt and malfunction; soft limits protect only disciplined operators and tend to be ignored exactly when they matter most.
How does the daily loss limit relate to margin in F&O?
In Indian F&O, marked-to-market losses on positions like short options can trigger a broker auto square-off through the day. Set the daily loss floor comfortably inside that level so your own control fires first, rather than being liquidated on the broker's terms and timing.
Can a daily loss limit stop me on a day I would have recovered?
Yes, and this is its main cost. Set too tight, it halts the strategy on normal intraday noise, forgoing a recovery. This is the trade-off for protection, and it is why the level should be calibrated to the strategy's daily P&L distribution rather than guessed.
Does a daily loss limit protect against a slow bleed?
No. It caps a single session but says nothing about many days of small losses that individually stay under the floor yet accumulate into a large drawdown. A daily limit must be paired with drawdown and exposure controls to cover that slower failure.
Should I resume trading after hitting the daily limit?
Not the same session. The limit exists precisely to stop the day, and resuming after the pain eases reintroduces the tilt it was meant to prevent. Resume the next session, ideally after reviewing why the limit was hit.

Voice search & related questions

Natural-language questions people ask about Daily Loss Limits.

What is a daily loss limit?
It is a line in the sand for one trading day. Once your losses hit it, you stop trading until tomorrow, so one bad day cannot snowball.
How much should I set my daily loss limit at?
Enough that a stopped day recovers within a few normal days, often a few percent of capital. Too tight stops you on noise, too loose fails to protect.
Should my algo close positions when the limit is hit?
It depends. Closing stops the loss growing but can give bad fills, while just halting new trades leaves open risk. Decide which for each strategy in advance.
Why does an algo need a daily loss limit?
Because an unwatched algo can keep making losing trades all day at full speed. A hard daily floor stops it before a bad day becomes a disaster.
Can I keep trading after I hit my daily loss limit?
No, that defeats the point. Wait for the next session, and first work out why you hit the limit at all.
Is a daily loss limit the same as a stop-loss?
No. A stop-loss exits one position at a price. A daily loss limit stops all your trading once the whole day's losses reach a floor, however many trades caused it.
Should the daily limit count open positions too?
Ideally yes. Mark open positions to their current price, or a big paper loss can sit open past your floor until the broker acts on margin.

Sources & references

    Last reviewed 12 July 2026. Educational content only — not investment advice. Markets and rules change; verify current conventions with SEBI, NSE/BSE and your broker.

    ← Previous
    Kill Switch Design
    Educational content only — not investment advice. Examples use illustrative numbers and simplified models. Risk-management techniques reduce but never remove risk, and trading derivatives involves substantial risk of loss. See our Risk Disclosure and SEBI Disclaimer.