What forced selling does to the next forced sale, given a book and a set of leveraged positions you state.
Nothing here is observed
Bids
Asks
Price and size per level. The result is stated against exactly the levels you give — three levels answer for three levels, and nothing is assumed beyond them. A published book is twenty levels a side, aggregated so they reach far enough for a cascade to run through rather than describing the spread.
The impulse that starts it. A negative shock closes longs; a positive one closes shorts.
How little equity a position may keep before the venue closes it.
Why it stopped: Every position on that side was closed.
The shock reached $99.00 without crossing a resting level, so it consumed no depth. A shock that goes further would: price cannot move past an order without taking it, and everything it takes is gone before the first forced sale.
| Pass | Price after | Forced onto the market | Depth it walked | Buckets left |
|---|---|---|---|---|
| 1 | $98.00 | $1,485.00 | $1,480.00 | 2 |
| 2 | $96.00 | $2,450.00 | $2,420.00 | 1 |
| 3 | $95.00 | $960.00 | $950.00 | 0 |
What this does not say
How this calculation works — the formula, its inputs, what it assumes, and what it does not establish.
State an order book, a set of leveraged positions and a shock, and watch what the forced selling does to the next forced sale. Every figure is one you typed: no venue is queried, no real position is read, and the result says which of three conditions ended the cascade rather than only where price landed.
A stop-loss is an order its owner shapes and can move. A liquidation is a market order the venue must fill regardless of price. At twenty times leverage a five-percent move closes the whole position, and that forced order pushes price into the next leverage tier — which is the difference between a one-percent move and a six-percent one.
Price does not travel from 100 to 95 while bids at 99, 98, 97 and 96 are still resting. Reaching a price means consuming everything better than it, so that depth is gone before the first liquidation and the first forced sale meets a thinner book than the one that was typed in. Skipping that step flatters every number after it.
A venue carries an insurance fund and auto-deleveraging, and resting orders arrive precisely because price moved. None of the three is modelled here, because each would mean inventing a number nobody gave — and all three absorb part of a cascade. The distribution of open positions across leverage is not observable anywhere either, so it is an assumption you state rather than a fact anyone can look up.
Run a shock through an order book and leveraged positions, and see what forced selling does to the next sale. MoonMath labels the data basis for this tool as Observed, Formula.
No. The Liquidation Cascade Simulator is an educational scenario tool, not financial advice, a recommendation, or a forecast. Provider coverage, liquidity, routing, fees and market conditions can make real outcomes materially different.
No. The book and the positions are the ones you typed. Nothing is observed and nothing is predicted; it answers what the arithmetic does to the numbers you gave it, the way the market-cap simulator answers what an inflow would have to be.
No. A heatmap claims to know where positions are, which no venue publishes. This asks you to state a distribution and then says plainly that the distribution is an assumption, so every number resting on it is one too.
Because running out of order book is a different statement from running out of positions. The first means the arithmetic reached the edge of what you described and the final price is the deepest level you gave, not a price a market reached.