Ethereum is falling for one of five reasons, and the whole job is telling them apart.
Three of the five have nothing to do with Ethereum at all.
ETH drops because the entire crypto market is dropping, and it usually drops further than Bitcoin for a reason that is closer to plumbing than to news.
This article gives you the five causes, the tell for each one, and an honest reading of what Ethereum's crash history actually shows — which is not the reassuring thing everyone repeats.
Key Takeaways
Ethereum crashes for five reasons — a market-wide selloff, Bitcoin leading lower, a leverage cascade, ETH's supply math since Dencun, and an unresolved identity — and three of the five have nothing to do with Ethereum.
ETH falls harder than Bitcoin in the same selloff largely because its order book is thinner at almost every level, so the same forced selling walks its price further down.
A liquidation is not a decision to sell but a forced disposal with no reservation price, which is why cascades accelerate instead of finding a floor.
The Dencun upgrade of 13 March 2024 moved Layer-2 data into a separate fee market that floors near 1 wei, cutting how much ETH gets burned.
Ethereum kept shipping through every drawdown — Dencun in March 2024, Pectra in May 2025, and Fusaka on 3 December 2025.
Every Ethereum crash so far has been followed by a recovery to a new record, but the last one took 38 months and beat the old record by 1.4%.
| Reason | The tell | How to check it in five minutes | Time signature |
1 | The whole market is falling | Bitcoin, Solana and XRP are all red, in roughly the same shape | Open any two other large caps. Similar drawdowns means this is not an Ethereum story | As long as the macro shock lasts — days to months |
2 | Bitcoin moved first | BTC fell first, ETH fell further, same shape | Measure both drawdowns from the same starting point. Deeper but identical = beta, not news | ETH usually stops when Bitcoin stops |
3 | Leverage cascaded | Vertical candles, enormous volume, a partial bounce soon after | Check whether the drop took minutes or hours. Minutes means forced selling | Hours. The fastest of the five to resolve |
4 | ETH's supply math | No event at all — just months of ETH losing ground to BTC | Read the ETH/BTC ratio over quarters, not days | Structural. Quarters to years |
5 | Nobody agrees what ETH is | ETH lags on good news it should have won | Watch whether ETH rallies when its own fundamentals improve | Structural. Only a narrative resolves it |
Two readers can walk this table honestly and reach opposite conclusions about what to do. That is the point.
Most of the time, the answer to why is Ethereum crashing has nothing to do with Ethereum.
A macro shock drains liquidity out of every risk asset at once, and crypto sits at the far end of that pipe — first to be sold, last to be bought back.
Neither had anything to say about Ethereum.
The tell is the cheapest check in this article: pull up two other large caps, and if they are all red by similar amounts, stop looking for an Ethereum explanation.
Ethereum almost never crashes alone.
When Bitcoin falls, ETH gets sold as a leveraged version of the same trade — same direction, more of it.
The trap here is that ETH's deeper drawdown looks like information, so people go hunting for the Ethereum-specific bad news that explains it, and they always find something, because there is always something.
The tell: measure both drawdowns from the same starting point, and if ETH is deeper but the shape is identical, that is beta and not news.
ETH usually stops falling when Bitcoin stops falling — the section below explains why the gap between them is so consistent.
Leverage does not cause the fall. It decides how far the fall travels.
When a leveraged position's collateral slips below its required ratio, the position is not sold — the position is not sold, it is disposed of — on an exchange as a market order, on a lending protocol as a collateral auction — and in neither case is anyone holding out for a better price.
That lower price pushes the next tier of positions under water, and the loop runs again.
The tell: cascades happen in minutes rather than hours, and the volume spike is far larger than the news would justify.
This one is genuinely about Ethereum, and people miss it because it has no news event attached to it.
The Dencun upgrade of 13 March 2024 gave Layer-2 rollups a dedicated data lane — blobs — with a fee market entirely separate from the main one, and when that lane is not full the blob base fee floors near 1 wei. Issuance to validators never changed, which means that when network activity is low the protocol mints more ETH than it destroys and total supply grows — the burn rule is untouched, it simply stopped being fed.
The tell: no event, no candle — just quarters of ETH losing ground to Bitcoin.
Bitcoin is digital gold. ETH is… harder to explain.
Is it a tech platform? DeFi infrastructure? A store of value? An internet bond?
A story that takes four questions to tell is a story that cannot be defended during a panic, and assets without a one-line story get sold first and bought back last.
The tell for this one is the most frustrating of the five: ETH lags on news it should have won.
When Ethereum's own fundamentals improve and the price does not respond, you are not looking at a crash — you are looking at a market that has not decided what it is pricing.
The usual explanation is that ETH falls harder because it is more volatile, which is not an explanation at all — it says ETH falls harder because it falls harder.
Here is the mechanical version, from the venue side.
Every crash sends the same instrument into every order book: a market sell with no price limit. What decides how far the price travels is not how much gets sold — it is how much resting liquidity sits underneath the current price. ETH's book is thinner than Bitcoin's at almost every distance from the mid, so the same notional of forced selling walks ETH further down its own book than it walks Bitcoin down its. That is arithmetic, not sentiment, and it is the same arithmetic in every single crash. It is why ETH's drawdown is so reliably the deeper of the two.
Two things make it worse for ETH specifically.
The collateral is the asset. A Bitcoin position margined in stablecoins loses on one side. An ETH position collateralised in ETH loses on both at once — the position and the collateral fall together, which is exactly what happened on Sky in April 2025.
And liquidation prices cluster. A liquidation price is arithmetic, not opinion — it falls out of the entry price, the leverage and the maintenance margin, and people enter at round numbers and pick round leverage, so the levels stack up in predictable places. Once the price reaches one shelf, it reaches the next.
None of that is a forecast. It is the shape of the machine, and it does not change between crashes.
Cycle peak | Peak date | Bear low | Low date | Drawdown | Months low → next record | The next record |
| Jan 2018 | ~$83 | Dec 2018 | ~94% | ~35 | $4,878 — 10 Nov 2021 |
| 10 Nov 2021 | | 18 Jun 2022 | ~82% | ~38 | $4,946 — 24 Aug 2025 |
| 24 Aug 2025 | — | — | — | — | — |
All price figures per CoinGecko, the provider used across MEXC Learn.
Yes — every major Ethereum crash so far has been followed by a recovery to a new record. That sentence is true, and it is the sentence everybody repeats.
Now do the subtraction.
Ethereum's November 2021 record was $4,878. Its August 2025 record was $4,946. Forty-five months apart, for a gain of 1.4%. Both recoveries took roughly three years from the bottom — 35 months after 2018, 38 months after 2022 — and the most recent one cleared the previous record by less than the spread on a bad trade.
"Ethereum always recovers" and "you will be fine" are not the same sentence, and the distance between them is the part nobody quotes.
One thing has held through every drawdown: the roadmap did not stop.
Whether any of that maps to price is a question this article is deliberately not answering — Ethereum's record is that development and price have run on separate clocks for a decade, and the clocks have never asked each other's permission.
Q: Why is Ethereum crashing?
A: Almost always because the whole crypto market is falling and ETH falls further than Bitcoin does, with leverage deciding how fast — only ETH's post-Dencun supply math and its unresolved identity are genuinely Ethereum-specific.
Q: Why is Ethereum dropping today?
A: A same-day drop is nearly always reason 1, 2 or 3 — check whether other large caps are down too, and whether the move took minutes or hours.
Q: Why did Ethereum crash?
A: Every major ETH crash on record has been a macro or Bitcoin-led selloff that leverage then amplified, not an Ethereum failure.
Q: Why does ETH fall harder than Bitcoin?
A: Largely because ETH's order book is thinner than Bitcoin's, so the same amount of forced selling walks its price further down.
Q: Did Ethereum crash?
A: Yes, repeatedly — roughly 94% through 2018 and roughly 82% into June 2022, and each was followed by a recovery that took about three years.
Q: Is Ethereum crashing because of the Dencun upgrade?
A: Dencun did not cause any crash, but by moving Layer-2 data into a near-free fee lane it cut how much ETH gets burned, which weakens ETH's supply story over quarters rather than days.
Q: What is a liquidation cascade?
A: It is when forced selling pushes the price low enough to force more selling, and it accelerates instead of stopping because a forced sale has no reservation price.
Q: How long do Ethereum crashes last?
A: The leverage part resolves in hours, the macro part lasts as long as the shock does, and the structural reasons run for quarters — the table above tells you which clock you are on.
Q: Should I buy Ethereum after a crash?
A: That depends on your time horizon and why you bought ETH in the first place, and it is worth knowing that Ethereum's last recovery took 38 months to beat its old record by 1.4%.
Ethereum crashes are miserable, but they are not mysterious.
Five reasons, and the first three are not about Ethereum at all — they are about the market Ethereum trades in, and about a liquidation engine that does exactly what it was built to do.
The two that are about Ethereum move over quarters, not afternoons, so if you are reading this during a red candle, it is almost certainly not one of them.
What history tells you is that ETH has come back every time and that coming back has taken about three years and, last time, bought a 1.4% gain over the previous record.
Both halves of that sentence are true, and you need both.