CoinCost
Bybit app
Home/Derivatives/Liquidation risk heatmap

Liquidation risk heatmap

Experimental · assumption-based simulation

Explore price movements and changes in hypothetical risk zones over 24 hours or 7 days.

A hypothetical simulation, not actual position liquidation prices or a prediction. Brighter bands show relative intensity, not amounts or probabilities.

Last completed close—
Available display range—
Visible candles · prior history—

Loading price and OI data.

Price candles · hypothetical risk distribution

Relative intensity: low → high · not USD amounts or probabilities

If the chart is unavailable, check the data status and details below.

Point, tap or use arrow keys to inspect · Esc hides the tooltip · +/− to zoom · each column is calculated after candle close

Select a candle to see OHLC, the price range, relative intensity and the actual OI observation time.

Current hypothetical distribution

0–100 relative intensity · last completed candle

A text summary of the relative distribution appears below.

Shown when usable OI changes are available.

Bybit USDT perpetuals · completed candles only · checking data timestamps

Calculation assumptions, limitations and official sources

This is an experimental CoinCost model. It does not contain exchange-disclosed entry prices, leverage or liquidation orders for individual positions. Stronger colors do not imply a probability of price reaching a level or a return.

Maintenance margin rate (MMR) is assumed to be 0.5%. Candles use completed Bybit Last Trade prices, which differ from the Mark Price used for actual liquidation.

The 24-hour view uses 15-minute candles; the 7-day view uses hourly candles. Available earlier history is calculated before trimming the display, preserving surviving hypothetical cohorts at the visible boundary. The number of earlier candles and the actual displayed range are shown. Missing periods are not fabricated.

The last cached candle is excluded until a later price candle confirms its completion. Elapsed time after receipt alone does not make it a completed candle.

  1. The initial OI total is used only as a baseline, not added to the distribution. Positive OI changes (ΔOI) between consecutive valid samples are split equally between hypothetical longs and shorts, using that completed candle’s close as entry proxy E.
  2. Blend mode allocates 1/4 each to 10×, 25×, 50× and 100×, then half to each side. This is an equal-allocation assumption, not observed leverage, and does not multiply OI by four. A single leverage remains available. With fixed MMR m=0.005, risk levels are long E×(1−1/L)÷(1−m) and short E×(1+1/L)÷(1+m).
  3. New cohorts are added at candle close and cannot be removed by that candle’s earlier high or low. From subsequent completed candles, cohorts whose risk levels are crossed by Last Trade prices are removed, and OI decreases proportionally shrink the remaining distribution. These simple rules can overlap in reducing weights and do not reconstruct actual liquidations or position balances.
  4. Missing price or OI data breaks the distribution and restarts its baseline at the next valid sample. Missing values are not invented. Each column represents a state calculated after its candle closes; it must not be read as a prediction available before that time.

Actual margin, added collateral, fees, funding, tiered maintenance requirements, cross margin and portfolio margin are excluded. MMR 0.5% is this tool’s assumption, not a rate applying to every real position. Intensity is relative within the loaded data window, not liquidation value in USD, actual exposure or probability.

Sources checked: