Loading
Move one price, revalue every book it touches, and test each account against its maintenance requirement.Note 16
The distance to a break divided by one daily standard deviation of the instrument that would cause it. Eleven per cent of something that moves one per cent a day and eleven per cent of something that moves six are the same number and not the same danger, and every dashboard in this asset class (including the tab beside this one) shows them identically.
Ranked by percentage, the books nearest a break look like the platform’s biggest risk. Ranked by their own instrument’s daily moves, the tight ones are small: 1 book sits within three daily sigmas holding $22.2k, while 99% of the measured equity is beyond six.
Two rankings of the same books, and they disagree because they measure different things. A percentage is a distance; a sigma is a distance divided by how far this instrument usually travels in a day. Both are on this page and neither replaces the other; a small book inside one ordinary day is a certainty nobody needs to size, and a large book six sigmas out is a tail nobody should ignore.
Equity here is the measured book’s own account value, not the capital at risk to a depositor, and it is summed only over the books this measure could reach.
44 of 44 breakable books have an instrument with enough history to measure. Counts and the equity behind them; which book is which is below.
| Distance | What that means | Books | Equity | Share |
|---|---|---|---|---|
| Under 1σ | An ordinary day for this instrument reaches the break. | 1 | $22.2k | |
| 1σ to 3σ | A bad day or a bad week reaches it. | 0 | $0.00 | |
| 3σ to 6σ | It takes a move this market makes occasionally. | 3 | $147.9k | |
| Over 6σ | Nothing this instrument has done recently comes close. | 40 | $25.03M |
The distribution above is a fact about the market and stays free. Which vault sits inside a single ordinary day of its own instrument is the number somebody acts on.
See what unlocksAlready a customer? Sign inNo probability is attached and none can be. A six-sigma day is one in a trillion under a normal distribution and this asset class produces them, so a probability here would describe the distribution rather than the market. The volatility is realised, hourly-bucketed, and needs five days of span before it is published at all. And the distance it divides is itself a ceiling: one instrument moves at a time, so anything market-wide arrives sooner.
The same computation is at GET /api/v1/stress, where ?scale=sigma places the rungs by how far out they are rather than at round numbers. The API also has the cascade, what the forced selling from these liquidations does to everyone else.