What a liquidation heatmap actually shows.

What liquidation means in a leveraged market

When a trader opens a leveraged position on a perpetuals exchange, they post a fraction of the full position size as margin. If the market moves against them far enough that the position's losses consume the posted margin, the exchange's risk engine closes the position automatically. That forced close is a liquidation.

Each liquidation adds one-sided pressure to the market. Long liquidations are forced sells; short liquidations are forced buys. When many positions are liquidated at similar prices, the resulting order flow can be sharp. For that reason, price levels at which clusters of liquidations might occur are of research interest — they are places where a move, if it extends to them, can accelerate.

How the heatmap estimates those levels

The problem is that the exchange does not publish, in real time, every open position's entry price and leverage. That data stays private to each trader. What exchanges do publish is aggregate open interest — the total notional size of all open positions in a market — a long/short ratio, and a stream of executed trades.

Enigma's heatmap takes those public inputs, read directly from Enigma's own exchange-data pipeline rather than any third-party API, and applies an isolated-margin model across the range of common leverage tiers offered on perpetuals exchanges. The model combines those estimates into a density distribution along the price axis — a density map of how much forced-liquidation notional is estimated to sit at or near each level. The exact weighting and calibration are part of the product's internal model and are not published here.

On the chart, that distribution is rendered as horizontal bands. Bands below the current price — usually shaded amber or red — represent estimated long liquidation clusters. Bands above — usually shaded green — represent estimated short liquidation clusters. Brighter bands indicate higher estimated concentration; dim bands, lower.

What the heatmap cannot see

Three assumptions bound what the heatmap can tell you.

Isolated-margin only. The model assumes each position is isolated — its own margin, its own liquidation price. In reality, many traders use cross margin, where their whole account balance backs every open position and losses in one position can raise or lower the liquidation price of another. Cross-margin liquidation levels cannot be estimated from public data alone, so they do not appear on the heatmap. A cross-margin trader with a deep equity cushion may not liquidate until far further out than the heatmap suggests.

Distribution assumption. The model spreads estimated open interest across leverage tiers based on observable distributions. If the real positioning is skewed differently — everyone actually using 50× in a market the model thought was mostly 10× — the cluster heights will be off. The heatmap shows a plausible distribution, not a certain one.

Point-in-time snapshot. Live cells are recomputed frequently as new exchange data arrives; the historical record is deep-frozen every five minutes so that past cells are immutable and cannot be repainted. Positions open, close, and scale continuously — a cluster that looked crowded even moments ago may have unwound since. The heatmap you see on the chart is a running estimate, not a guarantee of present positioning.

Reading the bands

A few things the bands usefully surface. Long wicks down into a bright amber band that then fail to break lower are often where estimated long liquidations were, in fact, triggered — the wick is the forced selling, and the failure to continue is the absence of further supply once that cluster has been consumed. The inverse holds for upward wicks through green bands.

Bright bands that sit well away from current price are less actionable — a lot can change before price gets there, and the snapshot will have refreshed several times over. The most useful reading is often near-term: where the clusters are in the next leg up or the next leg down from spot, given the prevailing volatility and range.

The heatmap is a research view, not a signal

Nothing about the presence of a liquidation cluster predicts that price will reach it, or that anything in particular happens when it does. Markets that look primed to cascade into a cluster often reverse before touching it. Markets with no visible cluster can still move sharply for unrelated reasons.

The heatmap is one input among many for thinking about where forced-flow pressure could appear. It is not an entry signal, not an exit signal, and not a recommendation. It sits alongside price, volume, order book depth, and your own strategy logic — all of which have their own limitations, and all of which you are responsible for interpreting.

Where to adjust what you see

Inside Enigma you can toggle the overlay per chart and filter it to the exchanges you care about. The display controls sit in the chart's overlay panel — none of them change what the heatmap estimates; they change how much of that estimate is surfaced on screen.