Bittensor liquidation heatmap

What the map shows

A perpetual futures position is borrowed money. Every leveraged trade carries a price at which the exchange closes it automatically, because the margin behind it has run out. That price is not a secret: it follows mechanically from the entry price and the leverage used.

This map is the liquidation heatmap for the selected venue's TAO perpetual over the chosen window. It takes the open interest added in each period, works out roughly where those positions were opened, and projects the price at which each standard leverage tier would be liquidated. Stack all of those projections together and you get a surface: bright where a lot of forced closing sits waiting, dark where there is almost none.

Bittensor trades in the hundreds of dollars and still covers a small-cap's range in a day. Its bands are therefore far apart in dollars and close together in percent, which is the only unit worth reading them in.

Time runs left to right, price runs bottom to top: The dotted line is the actual TAO price over the same window. Bands that survive to the right edge are leverage that price has not yet reached.

How to read it

• Bright bands above price are shorts. If TAO rallies into one, those shorts get bought back by the exchange, which adds buying to a move that is already up.
• Bright bands below price are longs. If TAO falls into one, those longs get sold, which adds selling to a move that is already down.
• A band that ends abruptly was consumed. Price traded through it, the positions were closed, and the liquidity is gone. That is why the surface is cleared behind the price line rather than left painted.
• Thin, tight bands very close to spot are the 100x crowd. They are large in count and small in notional, and they get taken out constantly.
• Wide bands far from spot are the 10x crowd. Rarer, much bigger, and the ones that produce the headline cascades.

What is different about Bittensor leverage

TAO's unit price changes how the map looks. A ladder quoted in dollars puts ten dollars between rows that are around three percent apart, so clusters that appear to be neighbours are, in the terms that matter to a position, a long way from each other.

Its range clears leverage quickly. A session that covers ten percent takes out the clusters on one side and leaves the other untouched, and the crowd rebuilds into the move. Expect this map to look different at the end of a day than it did at the start, more so than on any major.

What the map cannot see is when the move starts. It shows where leverage is stacked and how much of it sits on each side; the attention that reaches those levels arrives on its own schedule.

TAO carries the 24 hour to 90 day windows on every book on the picker; the 180 day and one year maps exist only on All exchanges and the three deepest venues, and only for Bitcoin and Ethereum.

Why clusters move price

A liquidation is not a normal sale. It is a market order the exchange sends on the trader's behalf, at whatever price the book offers, with no regard for slippage. When a dense cluster is hit, thousands of those orders fire inside a few seconds against a book that has usually already thinned out.

That is the cascade mechanic. Price reaches a cluster, the cluster fires, the firing pushes price further in the same direction, and the push reaches the next cluster. It stops when it runs out of stacked leverage or when resting bids and offers are deep enough to absorb the flow.

Using it without getting run over

1. Read the distance column in percent: TAO trades in the hundreds, so a cluster ten dollars away is roughly three percent from spot, not the near-miss it looks like.
2. Treat clusters as risk, not as entries: Knowing where forced selling sits tells you where a move is likely to accelerate. It does not tell you the move is coming.
3. Do not park a stop inside a cluster: That is precisely the price where slippage is worst.
4. Watch which side is heavier: When one side of price holds most of the leverage, the path of least resistance usually runs toward it.
5. Check the window: A cluster built over thirty days is a different animal from one built in the last six hours.
6. Size for the cascade, not the level: If you are long into dense long liquidity below, assume the fall through it will be faster than normal.

What it cannot tell you

• A blended level is not one order book: All exchanges adds the venues together, so a bright band there can be mostly one book. Only that book's own price trading through it triggers those positions, so pick the venue out before sizing around a level.
• It cannot see cross-margin: A trader with collateral spread across several positions is liquidated on portfolio health, not on a single price.
• It assumes standard leverage tiers: Real traders use 7x and 33x and every number in between.
• It is not a forecast: A dense cluster at 8% below spot says what happens if price gets there. It says nothing about whether it will.