Ethereum price prediction: Lighthouse flags a Glamsterdam bug as ETH falls below $1,900

Ether is trading at around $1,896 on Saturday morning, roughly €1,720, its lowest level since mid-August. The same week brought a technical announcement that carries more weight for the medium-term outlook than the price suggests: the developer team behind the Ethereum client Lighthouse has submitted a fix for a vulnerability that would have become exploitable around the next major upgrade, Glamsterdam.

What is confirmed: the fix exists as a pull request and is still awaiting review, and the price briefly fell below $1,890 overnight into Saturday. Our assessment follows further down. In short, we see ETF flows as the stronger lever over the coming weeks, but the code as the factor that determines when Glamsterdam arrives at all.

Where the price could head over a horizon of months is covered in our Ethereum price prediction. This piece sticks to the short term: what lies behind the Lighthouse vulnerability, which levels have broken or held since Friday, and which steps investors can sensibly take now.

Lighthouse bug ahead of Glamsterdam: what the pubkey cache is

Lighthouse is one of several consensus clients for Ethereum. A consensus client is the software validators use to check and attest to blocks and to agree on which version of the blockchain is valid. Alongside Lighthouse, the network runs Prysm, Teku, Nimbus and Lodestar, among others. This diversity is deliberate: if one program has a bug, the whole network does not depend on it.

The vulnerability concerns the so-called pubkey cache. This is a buffer in which a node keeps the public keys of all validators so that it can check their signatures quickly instead of reading each key afresh from the state of the blockchain. According to the description in pull request 10569 in the Lighthouse repository, a key could enter this cache before the associated deposit was finally confirmed. On a side branch of the blockchain, this could have led to conflicting mappings between validator and key.

The trade site U.Today describes the consequence as follows: affected nodes would have rejected valid signatures from a validator and thereby split off from the valid branch of the blockchain. Exploiting it would take effort. Before the upgrade, an attacker would need three pending deposits, namely a large top-up for an existing validator and two new keys, one of them with builder permissions, plus a penalty (slashing) at the right moment and a side branch. A chance hit is therefore considered unlikely.

The fix comes from Lighthouse developer Michael Sproul. Under it, keys only enter the global cache once the epoch from which the validator may be activated has been finally confirmed. A new structure called HeadPubkeys links the keys from the current state with the global cache, so that no key is missing if the blockchain goes for a while without finalising new blocks. The pull request carries the labels “gloas” and “v9.0.0”, so the change is meant to land in a new major version before the upgrade. Sproul himself points out that rolling back to an older version afterwards will no longer work smoothly.

One detail matters because conflicting reports are circulating: U.Today reported the fix as merged on Friday. On GitHub, the pull request was still open when we checked on Saturday morning, and approval from a second developer was missing. Whether other consensus clients have a similar weakness has not been confirmed, according to U.Today.

Gloas and Glamsterdam: the upgrade is still stuck on testnet

Glamsterdam is the umbrella name for the next Ethereum upgrade. It combines Gloas, the consensus-layer part, and Amsterdam, the execution-layer part. The Lighthouse vulnerability belongs to Gloas, because it would only have become exploitable with the changes to deposits and builder roles planned there.

The timetable had already slipped before this announcement. Activation on the Hoodi testnet, originally planned for October 26, was postponed; crypto.news cites eight affected invariants, meaning rules the code must always uphold. On November 6, according to the same report, the Ethereum Foundation wrote that it was working with the audit firms ChainSecurity and Dedaub on the proposals EIP-8037 and EIP-8038. There is no new date for Hoodi yet, let alone for mainnet.

For the price outlook, this means an upgrade that traders could price in as a catalyst is not getting any closer. Every vulnerability found before launch, however, is one that can do no damage after launch. That is exactly the tension ether is caught in right now.

Ether since Friday: $1,890 briefly broken, $1,875 held

On Friday evening, in our analysis of the Foundation wallet's sales, we named three levels: $2,000 as resistance, and $1,890 and $1,875 below it as support. Here is what has happened since. At the start of Friday (UTC), ether still stood at around $2,017 according to CoinGecko. The 24-hour high was $1,981.82, just below the round number, and the low was $1,880.84.

Support at $1,890 thus broke briefly overnight, while the lower one at $1,875 held. On Saturday morning ether is trading just above it again. Against bitcoin, one ether stands at around 0.0283 BTC. Over seven days the decline adds up to just over 16 percent, according to market data on crypto.news.

What is new compared with Friday is above all the distance to the summer: according to CoinGecko's daily prices, ether last closed below $1,900 in mid-August, at around $1,874 on August 17. The setback has therefore wiped out the entire autumn recovery.

Ether ETFs: six straight days of outflows to November 12

We continue to see the greatest selling pressure in the US exchange-traded ether funds. According to data from the analytics firm SoSoValue, the spot ETFs lost a net $70.57 million on November 12, the sixth consecutive day of outflows. In November, outflows up to that point total $695.9 million. Friday's figures were not yet complete on Saturday morning.

The mechanism is simple. A spot ETF holds real ether. When investors redeem shares, the issuer has to shrink its holdings and hand over ether, either by selling on the market or by transferring it to the trading partners returning the shares. Either way, supply enters a market where few buyers are standing by. For comparison: the Foundation wallet we reported on Friday sold 2,000 ETH for around $3.87 million.

Fees and burn: cheap transactions let the ETH supply grow

A second reason for the weak sentiment lies within Ethereum itself. Since the EIP-1559 proposal, the base fee of every transaction is burned, meaning permanently destroyed. crypto.news shows how strong the effect is with a simple example: a transfer uses 21,000 gas. At a base fee of 10 gwei, one million such transfers burn 210 ETH; at 1 gwei, only 21 ETH. To burn the same amount at 1 gwei would take ten million transfers.

Because Ethereum has deliberately lowered fees with its latest upgrades, the burn is small. If issuance to validators exceeds the amount burned, the supply of ether grows. Co-founder Vitalik Buterin nonetheless stood by the model on November 8, saying EIP-1559 including the burn was “good” and should stay. Markus Hämmerli, head of asset management at the crypto bank Sygnum, told crypto.news that institutions holding ETH for the long term increasingly think in ETH rather than dollars. For them, the staking yield matters more than whether supply is shrinking.

Ether price levels: where $1,841 and $1,982 come from

Three levels for the coming days emerge from the price data; none of them is a recommendation. On the downside, the overnight low of $1,880.84 comes first, with the August 17 daily close at around $1,874 directly below it. If this area gives way, the next documented zone is the summer low: on July 18, ether closed at around $1,841 according to CoinGecko, its lowest daily close of the past 200 days.

On the upside lies the 24-hour high at $1,981.82 and above it the round $2,000 mark, where ether has failed several times since Thursday. Only a daily close above $2,017, Friday morning's level, would partly make up for this week's setback.

Our view: the code fix reassures, ETFs set the price

From the newsroom's perspective, the Lighthouse announcement is more of a good sign. The vulnerability was found before Gloas goes live, it has been publicly described and the fix is available. That speaks for the review process currently delaying Glamsterdam. On the other hand, every further finding extends the wait, and as long as there is no date for Hoodi, ether lacks an event buyers could bet on.

For the next two weeks we therefore consider two scenarios supported by the data. In the first, ETF flows turn positive, and a return above $1,982 and $2,000 becomes realistic because the largest seller drops out. In the second, outflows continue at their current pace, and the summer low at $1,841 comes within reach. Both scenarios are readings of the data, not price targets. Anyone holding or planning to buy ether should size every position so that even a total loss remains bearable.

Ether staking: what separates solo stakers from exchange customers

If you run your own validator with Lighthouse, you now have a concrete task: wait for the release of version 9.0.0, read the release notes and update before the upgrade. Given the warning about rolling back to older versions, it is worth backing up your configuration beforehand. If you stake through an exchange or a staking provider, you do not need to do anything; the provider runs the software. What makes sense there is to ask which clients it uses and whether it spreads them.

Buying ether under MiCA: licence, fees and custody

Since the transition period ended in late 2025, full MiCA obligations apply to purchases in the EU: anyone brokering crypto-assets to customers in the EU needs a licence as a crypto-asset service provider. Which platforms hold this authorisation is shown in our overview of regulated crypto exchanges. With a falling price, fees deserve particular attention: a spread of one percent costs almost $19 per ETH at $1,896 before the price has even risen.

For custody the old rule applies: what is not being traded is better kept in your own wallet over the long run than in an exchange account. If you want to stake, though, you give up the convenient exchange option and have to organise it yourself.

Ether below $1,900: without ETF inflows there is no path above $2,000

1. Compare where to buy: Fees and spreads differ markedly between providers. Our crypto exchange comparison sets out the terms for ether side by side.
2. Check staking providers: Yield, lock-up periods and the clients used are the criteria that matter after the Lighthouse announcement. The overview of staking platforms puts them in context.
3. Decide on custody: For ether you hold for longer, a hardware wallet is the safest solution. Which devices support ether and staking is shown in our hardware wallet comparison.

(As of November 14, 2026. This article is not investment advice. Prices and fee structures change; check the terms with the provider before you buy.)