What is Ethereum (ETH)?

Ethereum is a computer that no one owns and everyone can write to. Where Bitcoin records who paid whom, Ethereum records the state of arbitrary programs — a lending market, an exchange, a registry of names — and executes them exactly as written. ETH is the asset you pay in to have that computation performed, and since 2022 it is also what secures the network.

That second role is worth stating precisely because it is what makes ETH different from a company's shares. Validators lock ETH as a bond. Propose an invalid block or equivocate and the protocol destroys part of it. Security is bought with capital at risk rather than with electricity, which is why the Merge cut Ethereum's energy use by more than 99% and why you can no longer mine it.

You cannot mine Ethereum

People still search for how. Since 15 September 2022 there is no mining — GPUs were made irrelevant overnight. Anything advertising Ethereum mining today is either mining something else, most often Ethereum Classic, or is a scam. The equivalent activity is staking, which needs capital rather than hardware.

What Fusaka changed

Fusaka activated on 3 December 2025, twelve EIPs led by PeerDAS. Before it, every node had to download all the data rollups posted. PeerDAS means each node samples a portion and verifies the rest statistically, so total blob capacity rises sharply without raising the cost of running a node. Layer-2 fees fell accordingly. The upgrade also raised the block gas limit and introduced a per-transaction gas cap.

One correction, because it circulates: Fusaka did not change staking economics. That was Pectra, the upgrade before it.

• Gas: every transaction and contract call on the base layer is paid in ETH, and a portion of that fee is burned.
• Staking collateral: ~38.9m ETH is bonded to secure the chain, which removes it from the tradeable float.
• The reserve asset of DeFi — the collateral lending markets are denominated in and the base of most trading pairs.
• Rollup settlement: Arbitrum, Base, Optimism and the rest post their data to Ethereum and pay in ETH. This is the largest growth area and the source of the problem below.

Ethereum versus Solana

The honest version of this comparison is that they made opposite bets. Solana put execution on one fast chain and accepted higher hardware requirements for validators. Ethereum kept the base layer cheap to verify and pushed execution to rollups. Solana feels better to use for a single application. Ethereum has more value settled on it and more institutions willing to hold the asset. The bet Ethereum has to win is that the rollups it hosts eventually pay it enough — which brings us to the supply.

There is no supply cap and never was. Issuance pays validators; EIP-1559 burns a portion of every fee. Net supply is the difference, so Ethereum's monetary policy is a function of how busy the chain is — the design that produced the "ultrasound money" claim when the chain was congested.

The L2 paradox

The scaling roadmap worked, and that is precisely the problem for the asset. Activity moved to rollups, which pay for cheap blob space instead of expensive base-layer gas, so the amount of ETH burned collapsed. Through 2026 ETH has been mildly inflationary, roughly 0.2% to 0.8% net annual supply growth. The engineering succeeded and the monetary narrative it had been paired with did not survive it.

The counter-argument, which is reasonable and unproven: about 32% of supply is staked and effectively illiquid, so the tradeable float is much smaller than the headline number, and as rollup volume grows blob fees could become a large enough share of the burn to turn the balance negative again. That is a projection. The current figure is inflationary, and an investment case that depends on the projection should say so.

What is not in doubt is the bond: no team allocation unlocking, no vesting cliff, no foundation tranche hitting the market on a schedule. Ethereum's supply risks are structural, not calendar-driven.

Staking ETH means locking it as a validator bond and being paid to propose and attest to blocks. Around 38.9m ETH — roughly 32% of supply — is staked across approximately 897,000 validators.

The native APR has compressed to about 2.8%, down from 3–3.5% after Pectra. The reason is simple arithmetic: rewards are shared among validators, so the more ETH is staked, the less each stake earns. Institutional demand from yield-distributing ETFs and corporate treasuries has pushed the entry queue to multi-year highs, which will compress it further.

The four ways to do it, and what each costs you

• Solo staking: 32 ETH and a machine you keep online. Full rewards, no fee to anyone, and slashing risk that is entirely yours.
• Liquid staking: deposit with Lido or similar and hold a receipt token you can use elsewhere. Convenient, a fee, and the receipt can trade below par in stress — which is exactly when you would want to sell it.
• Exchange staking: simplest and most expensive, and it makes the exchange your counterparty.
• A staking ETF: since 12 March 2026 BlackRock's iShares Staked Ethereum Trust, following the first US staking ETF from REX Shares and Osprey in September 2025. You get the yield net of fees inside a brokerage account, and you never hold the asset.

Exits are queued, not instant. In calm markets that is a formality; in a rush it is the constraint that matters, and it is worth knowing before you need to know it.

Value capture, not technology

Ethereum's central risk in 2026 is not that the chain fails. It is that the chain succeeds and the asset does not capture it. Rollups now carry most activity and pay comparatively little for it; the burn fell; supply turned mildly inflationary. If blob demand does not eventually grow into the gap, ETH is the equity of a settlement layer whose customers found a cheap plan.

Competition from a different design

Solana took the applications that need low latency and high throughput in one place. Ethereum's answer is a fragmented rollup landscape with its own bridging friction and user confusion. The modular thesis may still win, but it has been losing the user-experience argument for two years.

Staking concentration

A large share of staked ETH sits with a handful of liquid-staking providers and exchanges. That is a governance and censorship concern rather than a solvency one, and it is the criticism Ethereum has answered least convincingly.

Drawdown behaviour

ETH fell hard in the 2026 market decline, as it has in every cycle — it is a higher-beta expression of the same risk appetite that moves Bitcoin. Anyone sizing a position on a 2.8% staking yield should model a 50% price move against it, because the yield does not change the volatility and has never offset it.

Complexity

Ethereum ships a substantial upgrade roughly every year — Merge, Dencun, Pectra, Fusaka. Each has landed without a consensus failure, a genuinely impressive record. Each also adds surface area, and the record of no failures is a reason for confidence rather than a guarantee.