What would a crash actually cost you?

What a stress test tells you that a portfolio tracker does not

A tracker answers “what do I have?”. It will happily tell you that your largest position is 47% of the portfolio, that 68% of it sits on one exchange, and that your annualised volatility is 74%. All three are facts, and none of them is a decision.

A stress test converts those facts into a number you can act on: if this happens, you lose this much. That reframing matters most in the cases people systematically misjudge - a portfolio spread across a dozen tokens that all fall together, or one whose real risk is not the market at all but the single venue holding most of it.

The scenarios

Historical scenarios replay what assets actually did, measured from daily closing prices in the same price store the product runs on. Figures are close-to-close and measured from the day before the event to the worst close inside it, so a window that happens to include the bounce still reports the fall rather than the recovery.

10 Oct 2025 flash crash bitcoin -9%

The largest deleveraging event in crypto history: about $19B liquidated in a day. BTC touched roughly $102k from $123k intraday and closed about 9% down; the long tail closed two to four times lower than that.

FTX collapse (Nov 2022) bitcoin -24.5%

An exchange holding customer funds went from solvent to bankrupt in under two weeks. The market fell with it, and everything still on the venue was gone.

LUNA / UST collapse (May 2022) bitcoin -18%

An algorithmic stablecoin lost its peg and its backing token went to essentially zero in five days, taking the wider market down with it.

USDC depeg (Mar 2023) bitcoin +18.4%

A bank failure stranded part of the reserves behind USDC and it traded near $0.87 for a weekend. The lesson was that "cash" in a portfolio is still somebody's liability.

Covid crash (Mar 2020) bitcoin -39.5%

A global liquidity panic took bitcoin down about 40% between closes - more than half of it intraday - and correlations across every asset class went to one.

Tariff shock (Apr 2025) bitcoin -8.6%

Crypto traded as a risk asset on a macro headline, falling with equities rather than acting as a hedge against them.

Market scenarios

These are not historical. You set what bitcoin does and every other holding responds according to how it has actually behaved against bitcoin, so the long tail falls further than the majors - which is what every real crash has done.

• BTC -30%. An ordinary bear-market leg. Not a crisis; this happens.
• BTC -50%. A cycle drawdown. Every past cycle has had at least one.
• BTC -70%. A full bear market, peak to trough. The 2018 and 2022 bottoms both got here.
• Alt season. The upside case: BTC grinds higher and the long tail runs much harder. Useful for seeing how much of your result now depends on it.

Exchange failure

The scenario no equities tool has an equivalent for, and the one this product is unusually well placed to answer. Nothing about the market changes: prices stay exactly where they were, and the balances sitting at one venue simply stop being yours. If that number surprises you, it is the most useful thing on this page.

How each holding's response is worked out

In order, most reliable first, and the tool labels every line with which one it used - so an estimate never reads as a measurement:

1. Its own measured history. What that asset actually did over that window, from daily closes.
2. Its measured beta. How much it has historically moved for each 1% bitcoin moved, over the trailing year.
3. Its sector and size. A documented estimate, because an unranked memecoin and a mega-cap layer 1 do not fall together.
4. Nothing. If we cannot honestly estimate a holding, it is left untouched and counted as unmeasured rather than quietly scored as a 0% move.

Stablecoins are held at their peg and only move under the depeg scenario. Wrapped and staked tokens follow the asset they are a claim on, so wrapped bitcoin falls with bitcoin instead of counting as a second, independent position.

Measured history covers 254 assets, built on 2026-09-11. Anything outside that set falls to the estimate rungs above, and the tool says so.

What this page cannot do

It only knows the allocation you typed. Connect your exchanges and wallets to the app and the same engine runs against your real positions - measuring your own holdings' history rather than estimating from a sector, scoring the pillars that need price history, and letting you draft hypothetical trades to see whether they would actually have helped.