What does this position actually risk?
Pick an amount and a leverage. We show the move that wipes the position, what your stop costs when it is hit, and — the part most calculators skip — how often that move actually happened in our own daily candles, next to the textbook estimate. Not a recommendation, and not a prediction.
$1000 at 2× on SOL
A 50.0% fall wipes the margin you posted. Over 30 days, that move happened in 6% of the windows we can measure or model — whichever of the two is less flattering. That is the number to sit with before the one below it.
$2000Position controlled
50.0%Move that wipes the margin
8.7%Stop distance (2 x ATR(14) = 8.7%)
$173.74Cost when the stop is hit
17.4%... of your margin
68.7%Realised volatility, annualised
| Barrier | Distance | Model (GBM) | Our candles | サンプル |
|---|---|---|---|---|
| Stop is hit | 8.7% | 64% | 71% | 2084 windows |
| Margin is wiped | 50.0% | <1% | 6% | 2084 windows |
History is harsher than the model here — by
6 points on the wipe-out barrier. The formula assumes returns are
well-behaved; this asset's were not. Use the right-hand column.
Where each number comes from
- Move that wipes the margin — 1 / leverage. This is the naive distance: real liquidation arrives earlier, because maintenance margin, funding and fees are not modelled here. Treat it as the optimistic end.
- Cost when the stop is hit — stop distance × leverage × your margin, capped at your margin. Slippage and fees are not included, so a real fill is worse.
- Model (GBM) — the textbook first-passage probability for driftless geometric Brownian motion, 2Φ(−b / σ√T) with b = −ln(1 − d). The barrier is applied in log space; using the raw fraction understates deep barriers.
- Our candles — the share of overlapping 30-day windows in our daily series for SOL in which the low fell that far below the window's open. Series: 2020-11-09 to 2026-08-23, 2114 daily candles.
- Realised volatility — standard deviation of daily log returns over the trailing 365 days, annualised by √365. Stop preset — 2 × ATR(14) of the same series.
What this cannot tell you
- It is not a forecast. Both columns describe the past: one through a formula fitted to it, one by counting it. Neither knows what happens next.
- Windows overlap. Neighbouring windows share most of their days, so the sample size is smaller than the count suggests. It is shown so you can discount it, not to look large.
- Drift is set to zero. In a sustained uptrend the downside barrier is less likely than modelled, and in a downtrend more likely. We do not assume a direction on your behalf.
- Leverage above 3× is not offered here. Our engine can do more; a public page that models 10× normalises it. What we publish about our own trading is capped at 3× too.
- It says nothing about whether to take the trade. This is arithmetic on price history, not advice, not a recommendation, and not a signal.
Statistics built from our own daily candles, generated
2026-08-23T11:00:17Z · 2114 candles for
SOL_USDT. Trading crypto derivatives can lose you money, up to and including
everything you put in. This page is arithmetic, not financial advice.