Fidelity just filed to turn its spot Ethereum ETF into a yield-bearing product. The market cheered. I checked the code.
There is no code. Cocooned within the filing is a trust that will hold ETH, delegate staking to three custodians and three node operators, and distribute 85% of rewards as quarterly cash dividends. The other 15% goes to the intermediaries. This is not a protocol upgrade. It is a financial product engineering exercise โ a clever one, but engineering nonetheless.
Context
Fidelity Ethereum Fund (FETH) holds $903 million in ETH. Until now, it was a passive wrapper: buy shares, hold ETH, track price. The new amendment, filed with the SEC, allows the trust to stake up to 100% of its ETH. The staking will be executed by Anchorage Digital Bank, BitGo Bank & Trust, and Fidelity Digital Assets (custodians), who then delegate to Blockdaemon, Figment, and Galaxy (node operators). The mechanics: 15% of staking rewards are skimmed as a fee, split among the sponsor, custodians, and operators. The remaining 85% pays fund expenses first, then the surplus is converted to USD and distributed quarterly.
Why now? The IRS safe harbor rule from November 2025 cleared the tax uncertainty. It allows qualified crypto trusts to stake without losing their grantor trust status, provided they distribute net rewards at least quarterly. Fidelity, Grayscale (which started staking in October 2025), and 21Shares are all riding this wave. BlackRock chose a different path: a standalone staking ETH ETF launched in March 2026.
Core
Let me dissect the architecture. Three custodians, three node operators. This is not decentralization โ it is redundancy. The trust does not run validators. It outsources the technical overhead to Blockdaemon, Figment, and Galaxy. Each of these firms is a top-tier staking provider. But the trust's exposure to slashing risk is real. The filing warns: "Staked assets may be subject to slashing penalties. The custodian's liability for node operator actions is limited." In plain English: if a node operator screws up, the trust absorbs the loss, not the operator. The 15% fee covers this risk, but the magnitude of potential slashing is not quantified. Based on my audit experience, a single slashing event can wipe out months of rewards. The trust's multi-operator setup reduces the probability of a catastrophic failure, but it does not eliminate it.
Liquidity is another hidden cost. Staked ETH has an activation and exit queue. During unstaking periods, the trust cannot redeem ETH immediately. The filing reserves the right to delay redemptions or pay in cash instead of ETH. This is a technical compensation for the illiquidity of staking. In practice, it means investors who want to exit quickly may face friction. The "no minimum staking requirement" clause suggests the trust will dynamically adjust its staking ratio based on redemption expectations, further complicating the liquidity profile.
Now, the economics. The staking rewards come from ETH's consensus layer inflation and execution layer fees. This is real yield, not token emissions. At current ETH staking APR of around 3%, the $903 million trust generates roughly $27 million annually. After the 15% fee ($4 million) and the ETF management fee (0.25%, or $2.26 million), about $20.7 million remains for distribution. That is a 2.3% yield on top of ETH price appreciation. Compare this to self-staking via Lido or Rocket Pool, where users can earn 3-4% with full liquidity. The ETF offers convenience and compliance, but at a cost. The 15% staking fee is moderate, but the cumulative drag of management fees and the cash distribution mechanism (which creates taxable events) make it less efficient than direct staking for sophisticated investors.
Contrarian
The bulls will say this is a watershed moment: institutional capital finally accessing ETH staking through a regulated vehicle. They are not wrong. Fidelity has 53 trillion in assets under management and a massive retail distribution network via 401(k)s and IRAs. The staking feature makes FETH a "dividend stock" analog, attracting income-seeking investors. BlackRock's standalone staking ETF and Grayscale's earlier move validate the trend.
But here is the cold truth: this is a liquidity transformation, not a new capital source. Most of the money flowing into staking ETH will likely come from existing ETH holders switching from non-staking ETFs to staking ones. The net incremental demand for ETH is marginal. The total ETH staked may increase by 0.08% if FETH goes 100% staked, but that is a rounding error. The real winner is Fidelity, which locks in a recurring 15% fee stream from its own product. The node operators and custodians also benefit. The investor? They get a slightly better yield than a plain vanilla ETF, but they lose the flexibility of direct staking and bear the slashing tail risk.
Furthermore, the concentration of staking power through large ETF issuers poses a risk to the Ethereum network. If Fidelity, Grayscale, BlackRock, and 21Shares all funnel ETH into a handful of institutional node operators, the validator set becomes more centralized. This is not a hypothetical. The filing lists Blockdaemon, Figment, and Galaxy โ all of which are already major validators for Lido and other protocols. The feedback loop: ETF staking reinforces the dominance of these operators, which could undermine the network's decentralization ethos.
Takeaway
Fidelity's staking ETF is a well-engineered product that bridges traditional finance and crypto staking. It is not a technological breakthrough. It is a compliance arbitrage enabled by the IRS safe harbor rule. The code โ the actual Ethereum staking logic โ has been running for years. Fidelity just wrapped it in a familiar trust structure. Investors should understand the trade-offs: convenience and regulation come at the cost of efficiency, liquidity, and decentralization. The question is not whether this is good for Fidelity (it is), but whether it is good for the Ethereum ecosystem. I do not guess; I verify. On-chain, the flow of ETH into these vehicles will tell the story. Watch the validator sets, watch the redemption queues, and ignore the marketing hype. The code does not lie; only the auditors do.