Why down is sharper than up
Start with a structural fact: forced liquidation only triggers in one direction.
If you are long with leverage and the price falls to a certain level, the system sells your position for you — and that sale pushes the price lower, which reaches the next trader's liquidation level. Falling manufactures more falling. It is a positive feedback loop.
The same loop exists on the way up (liquidating a short requires buying), but the two sides are rarely symmetric in size: in most periods leveraged long positioning exceeds short positioning, so there is more fuel below. Add a behavioural asymmetry — people react faster to fear than to excitement — and you get the widely observed “it falls faster than it rises”.
What each layer looks like on the way down
Layer 1 — supply: a grind, not a crash
Increasing supply-side pressure (large unlocks maturing, long-term holders distributing steadily, exchange balances rising) produces weeks of slow decline: a little each day, feeble bounces, no volume expansion. Searching today's news for that shape will find you nothing.
Layer 2 — macro: falling with everything else
Same identifying move: check equity indices, the dollar, gold. If risk assets are falling together, this isn't about crypto. The extreme version is a liquidity shock — when institutions urgently need cash they sell what they're able to sell rather than what they ought to sell, and even havens fall alongside.
Layer 3 — leverage: the minute-scale vertical drop
The most common immediate cause of an intraday crash. For what this shape has looked like historically, browse the entries tagged “leverage” in the volatility timeline. Identify it by shape plus data: a vertical fall inside minutes, a burst of volume, a visible bounce afterwards — with liquidation data for the same window showing long liquidations far exceeding shorts.
For a sense of scale: in August 2026 one public liquidation dashboard showed roughly $300 million liquidated across the market in 24 hours and on the order of 80,000 accounts closed out. The ratio matters more than the total — in a decline the overwhelming majority of those forced out are positioned long, and that asymmetry is the tell. Such figures are aggregated from platforms' public endpoints with site-specific methodologies, so read them as orders of magnitude.
Layer 4 — structural events: traceable to a primary source
Withdrawals halted, an abrupt regulatory measure, a protocol exploit, a large institution in trouble. The signature is crypto falling alone (equities show no matching reaction) combined with a specific notice or filing you can open.
The cascade: one decline becoming three
Broken into stages it's clear enough:
-
An initial push
It can come from any layer, or simply be a large market sell hitting a thin book. The size is usually small — one or two percent is plenty.
-
The first tier is liquidated
The highest-leverage positions breach maintenance margin first and the system starts selling. Those sell orders do not care about price; they have to fill.
-
Price moves lower, reaching the next tier
Forced sales consume the bids beneath, the traded price steps down, and positions at the next leverage tier enter their liquidation range. The loop is running.
-
Liquidity withdraws, then the bounce
In violent conditions market makers widen quotes or pull orders to control risk, so the book thins and the same size produces a bigger impact. When the fuel runs out the price usually recovers part of the move — that “wick down and back” shape is the cascade's signature.
The full mechanism, and why it favours the thinnest hours, is in how liquidation cascades amplify moves.
“It fell because of liquidations” is incomplete. Leverage is an amplifier, not an engine. The complete version: an initial decline triggered forced selling, which multiplied the decline several times over. Remember only the first half and you'll apply this explanation to declines of a completely different character, in periods when leverage was never elevated in the first place.
Two misreadings specific to declines
One: treating “someone is dumping” as an explanation. On-chain you can see large transfers, but not intent; you can see assets moving to an exchange, but not whether, when or to whom they were sold. “Whales are dumping” is usually an unverifiable narrative; see are whales really dumping.
Two: reading falling market cap as money leaving. Market cap falling is arithmetic — circulating supply times the last price. The money you received when you sold went to a buyer; it didn't disappear.
Common questions
Where will it stop falling?
This site doesn't answer that and doesn't believe a reliable answer exists. This article explains what drove a decline that already happened; it does not extend forward. “Support levels” sit at different places on different people's charts and enforce nothing.
What should I do during a sell-off?
We give no timing advice. The one thing worth saying has nothing to do with the market: if a decline produces a strong emotional reaction in you, that usually indicates the position is large relative to what you can carry. That's information about you, not about the market.
Why do crashes happen at night and on weekends?
Because liquidity varies by hour. Books are thinner when volume is light, so the same sell order produces a larger price impact and more easily triggers chained liquidations. It's an ordinary feature of market microstructure and needs no conspiracy.