logo
TradFi
Sign Up to 15,000 USDT in Rewards
Limited-time offer is waiting for you!

Bitcoin Liquidation Heatmaps vs. Recorded Liquidations: What Each Dataset Can Prove

Quick answer: A liquidation heatmap estimates where leveraged positions might be forced to close under a model's assumptions. A recorded liquidation feed reports forced closures observed by its data provider. Neither identifies the order that started a Bitcoin sell-off. To assess a cascade, align mark price, trades, open interest, liquidation events, and executable market depth on one clock.

A search for “Bitcoin liquidations” often produces a large number and a colored map. Both can be useful. Neither is a substitute for an event study. The map estimates exposure near prices; the number counts events after margin thresholds have been crossed. Price can trigger liquidations, while liquidation orders can add to a price move. The direction of causation cannot be read from a 24-hour total.

This distinction matters when Bitcoin trades near levels that attract headlines. A search phrase mentioning $62,000 does not establish that Bitcoin traded there on November 16. A Phemex BTC/USDT screen captured for this analysis showed 65,263.70 at approximately 10:25 UTC+8, with a positive 24-hour change. A separate liquidation screenshot has no visible UTC capture time. It cannot be synchronized with that price screen, so the two images alone do not prove a liquidation-driven rise.

Trade BTC on Phemex!

What is a liquidation heatmap estimating?

Leverage creates a range of prices at which position equity may no longer satisfy maintenance margin. A heatmap provider combines available information—often open interest, price history, volume, leverage assumptions, and its own positioning model—to infer where many positions could be vulnerable. The output is conditional. It depends on a model of entry prices, collateral, leverage, and the extent to which positions remain open.

Actual liquidation thresholds are account-specific. A trader can add collateral, reduce a position, change other positions in cross margin, or close before the threshold. An exchange can change risk parameters. The market may move through a colored area after the positions estimated there have already disappeared. The darker color therefore does not represent a standing order that is certain to trade.

Even the price on the horizontal axis requires care. A heatmap plotted against a last-trade price may be compared with a venue that liquidates on mark price. Phemex's liquidation rules state that mark price drives its liquidation assessment; last price is the most recent trade. The two can diverge during a fast market. A chart wick touching a heatmap band is not evidence that every mapped account was liquidated.

What does a recorded liquidation feed measure?

A liquidation feed is closer to an observation, but it still has a scope. It may cover only participating venues, certain contract types, or the events that the provider can ingest. A dollar figure generally describes the notional value of positions closed, not the capital traders lost and not the amount of Bitcoin withdrawn from the market. A 24-hour figure is also a rolling window: two screenshots taken minutes apart can disagree without either being wrong.

Before quoting a total, preserve five fields: asset, contract universe, long/short side, time window, and provider. A “crypto liquidation” total can include ETH and other assets. A BTC figure can mix coin-margined and stablecoin-margined contracts. A large long-liquidation number does not tell us whether spot BTC sellers initiated the decline.

Dataset What it can support What it cannot establish alone
Heatmap A model estimates concentrations of vulnerable exposure near specified prices. That those positions still exist or will be closed.
Liquidation feed The provider observed forced closures in its stated coverage and window. That liquidations started the price move or equal realized trader losses.
Open interest Outstanding derivative exposure changed. Whether longs or shorts opened without further information.
Trade and order-book data Transactions and quoted depth at a venue and time. Every participant's motive or unseen liquidity.

Read the supplied liquidation screenshot on its own terms

The CoinGlass screen captured for this analysis labels its left panel “Liquidation Heatmap,” but the visible graphic is a coin-level treemap of recorded liquidation amounts over 24 hours. It has asset tiles and dollar totals, not a time axis and price levels. CoinGlass's documentation describes a price-level heatmap as a model-estimated intensity surface across time and price. Those are different products despite the similar label.

At the screenshot's undisclosed capture time, the dashboard showed $312.72 million in total liquidations over the trailing 24 hours: $119.26 million long and $193.45 million short. Short liquidations accounted for about 61.9% of the tracked total. The BTC tile displayed $158.77 million, with 74.7% attributed to shorts. Those are historical rolling-window observations from the supplied image. They are not live figures after publication, and they do not establish that a particular BTC candle was caused by short covering. The dashboard also displays some stock-linked symbols, so its all-asset total should not be casually relabeled as Bitcoin-only or crypto-only.

The BTC tile represents roughly 50.8% of that dashboard's tracked 24-hour total ($158.77 million divided by $312.72 million). That ratio is a property of the provider's displayed coverage and window, not BTC's share of global trading. For a causal argument, retrieve timestamps of the individual BTC liquidations and compare them with matched BTC price and order-book observations.

Long Or Short?

How to reconstruct a Bitcoin sell-off

Start with an event window, not a daily headline. Define the first price break using a specific BTC pair, venue, and timestamp. Collect one-minute or finer observations on that clock. Where data providers use different time zones or publish late, retain their original timestamps and record the conversion.

Next, compare four sequences. Price shows when the move occurred. Volume shows how much traded, though not why. Open interest shows whether outstanding derivative exposure rose or fell. Liquidation events show when a provider observed forced closures. Order-book snapshots and realized slippage help estimate whether available bids were thin enough for sell orders to move the price materially.

Consider two possible sequences. In the first, BTC drops on heavy spot selling, open interest starts falling, and long liquidations arrive minutes later. The evidence is consistent with liquidation as an amplifier. In the second, a cluster of liquidations arrives during a thin period and prices accelerate downward while depth vanishes. That is consistent with a stronger feedback effect. Neither sequence proves what the first seller intended. The difference is in timing and the size of the change relative to available liquidity.

An especially common error is to use a 24-hour liquidation total to explain a five-minute candle. The total may include closures hours before and after that candle. The analyst must compare matched windows. Likewise, a chart of open interest denominated in USD can fall mechanically when BTC falls even if the number of contracts changes less. Check whether the series is in contracts, BTC, or dollars.

A simple test for a cascade

A defensible cascade claim needs a testable sequence: a price decline reaches margin thresholds; forced closures are recorded; marketable liquidation orders meet limited bids; the next price move exposes additional accounts. The analyst should then test competing explanations, such as a macro announcement, a large spot order, an index-price move, or a venue-specific liquidity gap.

The evidence table for one episode should include the first observed price break; the same-minute mark and index prices; liquidation timestamps by contract; open-interest changes in native units; and bid depth at fixed distances from the midprice. If these fields are unavailable, say that the sequence cannot be resolved. A single heatmap image does not fill the gaps.

There is a second limit: published liquidation events need not equal all forced selling. A position can be reduced before liquidation, closed by a risk desk, or hedged elsewhere. Conversely, a closing trade observed during a drop need not be a liquidation. The visible feed is a sample of a process that also involves discretionary decisions.

What traders can use this for

Heatmaps are scenario tools. They can help identify where leverage may be concentrated and where to inspect liquidity. Recorded liquidations describe stress that already occurred. The useful trading question is not “Where will price be pulled?” but “How much evidence supports a forced-flow mechanism at this time, on this contract, and with this depth?”

A decision rule based only on a bright heatmap band ignores changes in collateral, positioning, and order-book liquidity. A decision rule based only on a liquidation total is late by construction. Traders who use these datasets should specify the venue, contract, trigger price, and maximum loss they can tolerate before placing an order.

Conclusion: A Bitcoin liquidation cascade is a process to investigate, not a label to attach to a red candle. Model estimates and recorded closures answer different questions. Time alignment and market depth are what turn a plausible account into an analysis.

Disclaimer: This article is for informational purposes only and does not constitute financial advice. Digital-asset and derivatives trading can result in substantial losses.

Sign Up and Claim 15000 USDT
Disclaimer
This content provided on this page is for informational purposes only and does not constitute investment advice, without representation or warranty of any kind. It should not be construed as financial, legal or other professional advice, nor is it intended to recommend the purchase of any specific product or service. You should seek your own advice from appropriate professional advisors. Products mentioned in this article may not be available in your region. Digital asset prices can be volatile. The value of your investment may go down or up and you may not get back the amount invested. For further information, please refer to our Terms of Use and Risk Disclosure