why your chart missed the price that liquidated you

liquidation is tied to the venue’s mark price, not necessarily the last traded price displayed by your candle chart. mark and last can separate during a wick, thin book or fast taker flow. if mark reaches the live liquidation threshold while last remains above it, a long can be liquidated even though the last-price candle never prints the liquidation number you recorded.

the inverse is also possible. a last-price wick can trade below a long position’s displayed liq while mark remains above the threshold, leaving the position open. the candle low is therefore not standalone proof that liquidation should or should not have occurred. first identify the chart’s price series, then compare it with venue mark data for the same interval.

the number copied at entry may also be stale by the time the event occurs. funding owed and fees debited from margin reduce the equity supporting the position. under cross margin, unrealised PnL from other positions changes the shared margin balance as well. each mechanism can move the live liq without changing the entry shown in your trade history.

mark price controls the liquidation check on Binance and Bybit perpetual futures covered here. last price is the latest taker fill, while mark is the venue’s reference price for unrealised PnL and liquidation. the two can print different levels at the same moment, so a last-price screenshot cannot reconstruct the trigger by itself.

conditional-order trigger type is selectable and varies by product and interface. check the trigger basis on the account rather than assuming a stop watched the same price as liquidation. a stop keyed to last can remain untriggered while mark crosses liq, or trigger during a last-price wick that never takes mark to the liquidation threshold.

a clean reconstruction therefore needs separate mark and last series. trade candles establish what printed on the book; mark history establishes what the liquidation engine observed. venue documentation explains the mechanism, but it does not make an old last-price candle interchangeable with the corresponding mark series.

take a Binance BTC isolated long with a $97,400 worked-example entry, $10,000 notional and 20x leverage. the initial margin is $10,000 ÷ 20 = $500. this notional sits inside the first BTC bracket, which runs through $300,000 with 0.4% MMR and maximum leverage of 150x.

the isolated approximation is liq = entry × (1 − 1/leverage + MMR). inserting the position gives $97,400 × (1 − 1/20 + 0.004) = $92,919.60. distance from entry is $97,400 − $92,919.60 = $4,480.40, or 4.6%. the 0.4% MMR pulls the threshold inside the naive 5% leverage distance.

the same $10,000 Binance BTC example shows how quickly room contracts as leverage rises. estimated long distance falls from $19,090.40 at 5x to $584.40 at 100x. every output remains an estimate — exchanges compute exact liq from live margin; always double-check on your venue.

the isolated formula above omits exchange fee and funding accruals since entry. Binance taker fee in the facts pack is 0.05%; applied to $10,000 notional, that is $10,000 × 0.05% = $5 for one taker fill. an entry-and-exit taker round trip is 0.1%, or $10 on the same unchanged notional.

against the example’s $500 initial margin, that $10 round trip equals 2% of margin. actual close notional can differ from entry notional, and liquidation-specific charges are venue- and tier-dependent. those charges consume remaining equity rather than creating extra chart distance. exact liquidation-fee treatment must be taken from the account record, not inferred from the candle low.

Binance and Bybit use an 8-hour funding interval in this facts pack. when funding is owed and debited from available margin, the position has less equity and its live liq can move toward mark. funding received can move equity the other way. the applicable rate and payment direction must come from the position’s own funding ledger rather than a later rate snapshot.

yes. MMR is selected from the position’s notional bracket, so the same entry and leverage can produce a different estimated liq when notional crosses a tier. for Binance BTC, MMR rises from 0.4% through $300,000 to 0.5% through $800,000, then 0.65% through $3,000,000. higher listed brackets reach 1% and 2%.

Bybit BTC begins at 0.33% through $300,000, then moves to 0.5% through $2,000,000. the next listed thresholds are $2,600,000 at 0.56%, $3,200,000 at 0.63% and $3,800,000 at 0.67%. this is why copying an MMR from a smaller position can misstate a larger position’s liquidation threshold.

only the first 5 brackets are represented here. for notionals above the last listed bracket, the calculator fallback uses that last bracket and can slightly underestimate MMR. bracket checks matter whenever size changes, partial liquidation changes notional, or the venue updates its risk limits.

a candle only answers the question for its own price series. if the chart uses last price, its low records the lowest taker fill during that interval. a long can show a last-price low below the displayed liq while mark stays above the live threshold. in that case, the wick crossed on the book but not in the liquidation engine’s reference series.

the complaint reverses when mark moves through the live threshold while last does not print there. your last-price chart can show every candle above the copied liq even though the venue’s mark series crossed it. a delayed screenshot adds another mismatch because the liq may already have drifted through funding, fees or cross-margin equity changes.

timeframe compression can hide sequence as well. a candle gives an interval’s open, high, low and close, not the order in which mark, fills, funding debits and margin changes occurred. event-level account records and mark history are therefore more useful than enlarging a last-price candle after the fact.

start with the venue’s liquidation record and account ledger, not the chart screenshot. preserve the displayed liq you recorded at entry, but treat it as one state rather than the final trigger. the reconstruction needs entry, side, leverage, notional, margin mode, bracket, MMR and every equity-changing event before liquidation.

then align the account timeline with mark and last data from the same venue. public forced-liquidation feeds are incomplete, so absence from an aggregator does not prove that the venue did not liquidate the position. the decisive comparison is the account’s live margin state against venue mark, with the calculator serving as an isolated-margin approximation rather than an exchange ledger.

1. record the position’s entry, side, leverage, notional and isolated or cross margin mode.
2. identify the notional bracket and MMR active when liquidation occurred.
3. collect mark-price history and separate it from the last-price candle series.
4. check funding entries, trading fees and any venue-specific liquidation charge in the account ledger.
5. for cross margin, add the unrealised PnL and margin effects of every other open position.
6. compare the reconstructed live liq with mark at the event timestamp, then reconcile the forced close fill separately.