Crypto daily loss limits: calculation and reset rules

A crypto daily loss limit is a rule that stops specified trading activity when the day's measured loss reaches a preset amount. For an equity-based rule, calculate daily P&L as current equity minus opening equity minus net external cash inflows, including changes in open-position value and costs. Define the reset time, breach threshold and post-breach actions before relying on the limit.

The number alone is incomplete. A “$200 daily limit” can describe a warning, a block on new entries, or an instruction to close positions. It can measure closed trades, total equity, or loss from an intraday peak. Those choices produce different outcomes on the same account. A useful rule makes the accounting and the resulting action explicit.

Define the account, valuation and clock

Start with the measured boundary: one strategy allocation, one exchange subaccount, or a consolidated portfolio. If several bots and manual trades share the account, an account-level equity rule includes their combined results. A strategy-only figure needs a separate, reconciled allocation ledger.

Choose a reporting currency and a consistent equity source. Include the covered assets, open-position value and liabilities under that source's definitions. If collateral itself changes value in the reporting currency, that movement affects account equity even without a new trade. “Available balance” and margin used are different quantities; neither automatically measures total account value.

Write an exact session boundary, such as 00:00 UTC, plus a policy for an unavailable opening snapshot. Preserve the timestamp of the valuation and when the control received it. A daily rule evaluated only at day-end can miss an intraday breach that later recovers.

Provider conventions differ. Kraken Prop documents a daily allowance based on 3% of the balance at 00:30 UTC, with open losses and costs included in the equity test. That is a rule for its named Prop account, not a universal setting for crypto exchange accounts. Kraken Prop's maximum daily loss explanation

Calculate the day's result without treating deposits as profit

For the illustrative equity-based policy in this article, freeze an opening equity value and a dollar budget at the session start. Use:

• E0: opening equity.
• Et: current equity using the same valuation method.
• Ct: cumulative external deposits minus withdrawals since the opening snapshot, valued in the reporting currency when they occur.
• B: the fixed dollar loss budget for that session.

Then calculate daily P&L = Et − E0 − Ct. Remaining headroom is B + daily P&L. This policy breaches when headroom reaches zero or becomes negative. Its current equity floor is E0 + Ct − B.

Bybit's Unified Trading Account reporting uses the same equity-change and net-flow structure for Today's P&L, measured from midnight UTC. Its documentation supports the accounting relationship; a reporting display does not establish that a loss-limit control enforces it. Bybit's Account P&L Analysis

A deposit raises both equity and net external inflows. A withdrawal lowers both. Neither changes the measured trading result at that instant. Value a coin transfer at its event-time reporting-currency amount, preserve its identity, and avoid counting it twice. A transfer between two accounts inside a consolidated boundary is internal.

Trading fees and funding remain costs under this net-result policy. Calling them withdrawals would hide a loss. Conversely, do not subtract a fee again if the equity value already reflects it.

This is a dollar-loss control, not a cash-flow-adjusted percentage return. Deposited capital can change subsequent exposure and results even though the transfer itself is neutral. For performance across changing capital, use the separate cash-flow-adjusted drawdown ledger.

If a policy sets the budget as a percentage of opening equity, calculate that dollar amount once and specify when it can change. Recomputing the allowance after every deposit can replenish a breached budget. The example below instead uses a fixed $200 for the entire session; its 2% relationship to opening equity is illustrative, not a suggested risk level.

A worked ledger with floating loss and two transfers

This hypothetical cash-settled derivatives account begins at $10,000 equity at 00:00 UTC. Its cash balance is $10,100 and its existing open position has $100 of unrealized loss. All values use one reporting currency; there are no other assets, liabilities or valuation changes beyond those shown. The daily loss budget is $200. Trading-result rows include the stated costs, and transfers have no separate charge.

At 09:00, assume gross closed trading P&L is −$70, fees and funding total $10, and unrealized P&L has moved from −$100 to −$140. The day's result is −70 − 10 + (−140 − (−100)) = −$120. Only the change in the opening position's unrealized value belongs to the day's incremental result.

At 12:00, unrealized P&L has deteriorated to −$220. With the same closed P&L and $10 of costs, the result is −70 − 10 + (−220 + 100) = −$200. The flow-adjusted equity floor is $10,000 + $1,500 − $200 = $11,300, exactly the observed equity.

The final row adds $10 of execution cost after detection. Actual daily loss becomes $210, exceeding the trigger. The account still has more equity than at midnight because it received net new capital. A raw comparison of $11,290 against $10,000 would conceal the loss.

The limit is a trigger, not a guaranteed final loss cap. An existing order can fill, a price can move, or a closing action can incur a cost after the threshold is observed. Assess that possibility with portfolio stress testing, including communication delays and joint moves across positions. Headroom should not be treated as cash available to risk in full on the next entry.

Decide what the reset does to an open position

A reset changes the measurement window. It does not make the position's economic loss disappear.

Continue the hypothetical example with no further activity. At the next reset, equity is $11,290, the cash balance is $11,510, and unrealized P&L remains −$220. Compare two deliberately different policies, each with a fixed $200 allowance:

This table recalculates accounting headroom only. It does not authorize a restart after the previous session's breach.

Under the equity baseline, the carried position starts with unrealized P&L of −$220 already reflected in the opening value. Another $30 deterioration is a new daily loss of $30. Subtracting the entire $250 floating loss again would count the opening loss twice.

Under the cash balance convention shown, the existing $220 floating loss already exceeds the allowance. These rows illustrate why a trader cannot infer the new day's room from the word “reset.” They are not replicas of a particular provider's complete rules.

For a separate real implementation, FTMO's 2-Step rules calculate the floor from the balance at midnight CE(S)T less 5% of initial simulated capital, then compare equity against it. This differs from both a percentage of current equity and a UTC-midnight assumption. Read the rule for the specific program and account. FTMO's Trading Objectives

Keep a longer-horizon loss or drawdown rule alongside the daily calculation if the policy requires one. Several losses on successive days do not become harmless because each belongs to a different session. A volatility-targeting rule addresses position exposure before a move; a daily stop responds to the measured result. Their settings and purposes should remain explicit.

Make a breach persistent until the restart decision

For an operator-defined control, a useful design is to latch the breach: store the first crossing and block new risk even if equity subsequently rebounds. Otherwise, a fluctuating mark can repeatedly stop and restart entries around the threshold. Document whether the next clock reset merely starts a new accounting window or also authorizes trading again.

The response needs its own evidence. Block the covered new entries, identify outstanding entry orders, and reconcile any fill that occurred while the action was being processed. Decide explicitly how protective orders are treated. If the policy calls for reducing exposure, verify the resulting quantities and costs after the reduction.

These are separate operations. A halt on new submissions does not prove accepted orders were canceled or positions closed. The trading-bot kill-switch guide explains that execution boundary. Kraken Prop, by contrast, documents automatic position closure, order cancellation and disabled trading after its breach; those documented consequences must not be assumed for another control. Kraken Prop's breach actions

Restart should follow reconciliation of equity, transfers, outstanding orders and the cause of the stop. A new session identifier should not erase an unresolved request or release queued entries whose signals have expired. Retain the old session's breach record even when a new baseline is valid.

A worksheet to validate the rule before use

Fill this out for the actual account and control. A blank field is an unresolved operating decision.

Replay the ledger with a deposit, a withdrawal, an overnight open loss and a threshold touch. Then introduce a duplicate transfer event and a stale equity mark. The reconciliation should expose those faults; a green status based on incomplete inputs is not reliable evidence of available headroom.

For the fixed-start policy above, intraday gains increase the distance to the same opening floor. A separate rule that limits profit giveback from the day's peak will behave differently. Test both paths if both controls exist, and retain the values that caused each decision.

The useful final record is concrete: a timestamped opening value, a reconciled flow ledger, the current equity and headroom, and a verified account state after any breach. Those records turn “stop after a bad day” into a rule that can actually be checked.