How Ethereum ETF staking rewards work (and what Fidelity's FETH filing changes)

Fidelity is staking the ETH in its FETH Ethereum ETF. It filed a Form 8-K on August 10, 2026, disclosing a Sponsor Agreement that lets the trust stake up to 100% of its ETH through node operators. Under the arrangement, the trust retains 85% of the staking rewards it receives. The other 15% is paid to the Sponsor, then shared with custodians, node operators, or other third parties. The SEC declared the registration statement effective on August 21, 2026.

What Fidelity actually filed with the SEC

As per Fidelity Ethereum Fund’s Form 8-K, Fidelity’s Ethereum Fund entered into an Amended and Restated Sponsor Agreement with its custodians, setting the operating rules and the fee split for staking the trust’s ETH. The useful details are as follows:

• The fund: FETH, Fidelity’s spot Ethereum ETF, held 476,311 ETH, worth about $759 million, on June 30, 2026, according to its quarterly report. Under its prospectus, the trust may stake up to all of its ETH, holding some back for redemptions, expenses and liquidity.
• Who holds the keys: custodians keep the private keys and stake the trust’s ether through node operators. Fidelity does not run the validators. Shareholders do not either.
• Named custodians: Anchorage Digital Bank NA, BitGo Bank & Trust, N.A., and Fidelity Digital Assets, N.A., per the prospectus.
• A separate, existing fee: FETH already charges a 0.25% annual sponsor fee on its ETH. The new 15% staking fee is added on top. It does not replace the sponsor fee.
• What’s still open: the SEC declared the registration statement effective on August 21, 2026, and Fidelity says FETH “participates in staking activity and expects to receive staking rewards.” No shareholder distribution had been paid as of October 2026. The open questions are how the quarterly cash payments work in practice, how network rewards and expenses change the net yield, and how much of its ETH Fidelity stakes while keeping enough unstaked for redemptions.

Staking is underway; the first payout to shareholders is what has not happened yet.

How a validator reward actually reaches a shareholder

Now that staking is live, a reward does not move straight from a validator to a brokerage account. It passes through several parties first:

Step	Who	What happens
1. Staking	Custodian, via node operators	The custodian has a node operator run validators with the trust’s ETH, and keeps the private keys
2. Reward received	The trust	Validators earn ETH for helping secure Ethereum
3. Fee split	Sponsor, custodians, and node operators	15% of the staking rewards received is paid to the Sponsor, then shared with the custodians, the node operators, or other third parties
4. Net reward retained	The trust	85% of the staking rewards received stays in the fund
5. Distribution	Shareholders	What remains after the 15% fee and fund expenses is paid in cash, once a quarter, not as it is earned and not reinvested into more ETH. The cash can come from selling staking rewards, a portion of the trust’s ETH, or both

Both the 85/15 split and the quarterly cash payout of what remains after the 15% fee and fund expenses are in the 8-K. Compared with staking ETH yourself, the fee comes out before you see any yield, and the cash arrives on a quarter-end schedule rather than accruing as you go.

Why this wasn’t legally possible until now

Spot Ethereum ETFs are grantor trusts. That tax status lets income pass straight through to shareholders instead of being taxed at the fund first. Staking raised a real question: if the trust puts its ETH to work, rather than simply holding it, does it lose that status?

The IRS resolved this in November 2025 with Revenue Procedure 2025-31. According to legal analysis from Willkie Farr & Gallagher, it creates a safe harbor: a trust that holds one type of digital asset from a permissionless proof-of-stake network can stake it and remain a grantor trust, as long as it passes the staking rewards on to shareholders. On October 6, 2026, Revenue Procedure 2026-20 clarified and superseded it. The rule now reads: net staking rewards are distributed in kind or in cash no more than 60 days after the end of the quarter in which the trust gains control of them. Fidelity’s quarterly cash payout is how it meets that rule, not a product choice.

Before this safe harbor existed, issuers had a real incentive to leave ETF ether unstaked rather than risk the trust’s tax treatment. Fidelity’s filing is a direct downstream effect of that IRS guidance landing nine months earlier.

Fidelity isn’t first: Grayscale already pays staking rewards

Fidelity did not invent this structure. The Block reported that Grayscale was the first US issuer to activate staking on a spot Ethereum ETP, turning it on for ETHE and the Grayscale Ethereum Staking Mini ETF (ETH) in October 2025.

Grayscale announced ETHE’s first shareholder distribution on January 5, 2026: $9.4 million in total, or $0.083178 per share, for staking from October 6, 2025 through December 31, 2025. The cash was paid on January 6, 2026. Grayscale then moved to monthly payouts: ETHE paid about $21.2 million in staking distributions in the first half of 2026, according to its quarterly report.

Those payouts are the useful comparison because they already happened. Validators staked, rewards accrued, and cash reached Grayscale shareholders on a regular schedule. Fidelity is on the same path, but it has not yet made its first distribution. The Grayscale amounts were real, and small next to the fund, because a slice of the reward stays in the fee layer.

What the ETF wrapper costs you that direct exposure doesn’t

This is not an argument against ETF staking. It is a way to add yield to a product many people already hold in a brokerage account. But the wrapper has trade-offs worth highlighting:

• A fee between the validator and you. 15% of the staking rewards is paid to the Sponsor, then shared with the custodians, the node operators, or other third parties, before you receive anything. That fee sits on top of the 0.25% annual sponsor fee. Both are the price of not running the infrastructure yourself.
• No say over the validator. You do not pick the node operator, see its uptime history, or move the stake if it underperforms. The custodian and Sponsor make those calls.
• Regulatory dependency. The SEC has cleared FETH’s staking, but the product exists inside rules that regulators and the IRS can still change, as Revenue Procedure 2026-20 showed. You have no say in that.
• Quarterly cash, not a live balance. Rewards accrue continuously on Ethereum, but they reach you every three months, in cash. They do not automatically compound into more ETH.
• Compressed base yield. The reward being split is already falling. As of early August 2026, Ethereum staking participation had reached roughly 34% of circulating supply, and the 7-day staking APR was about 2.66%, close to half its 2023 peak, because more ETH is sharing the same reward pool. (Ethereum’s own protocol changes are part of why that pool is under pressure: a proposal called Tapered Issuance Burn is still a draft, not approved or scheduled for any upgrade.) A 15% fee comes out of an already-thinning number.

These trade-offs mirror something we’ve written about honestly elsewhere: every yield structure has a cost the marketing does not lead with. Our own breakdown of RUNEBond’s lock-ups and operator risk applies the same standard to bonding: the drawbacks are named, not buried.

The alternative: earning yield without an ETF or a validator

There is a third path if you want neither the ETF’s fees and quarterly cash cycle, nor a validator of your own. It is not on Ethereum.

THORChain is a protocol for swapping assets such as BTC, ETH, and RUNE directly between blockchains, without a centralized exchange in the middle. Node operators secure it by posting RUNE, its native asset, as a security deposit called a bond. Most operators cannot fund that bond alone. RUNEBond is the interface that connects them with bond providers: people who add RUNE to an operator’s bond and earn a share of the yield, without running any infrastructure themselves.

The yield comes from a different place than ETF staking. It is paid largely from THORChain’s swap fees, which are protocol revenue from actual trades, rather than from a validator reward that still has to pass a custodian, a Sponsor, and a fee split. There is no fund wrapper, no SEC approval to wait on, and no quarterly batch. RUNEBond never holds your bond: it moves from your wallet to the node. It comes with its own trade-offs, covered in our guide to RUNEBond’s risks and lock-ups: bonded RUNE cannot be withdrawn until the node operator opens an unbonding window, and RUNE has its own price risk. It is not a stand-in for holding ETH. It is a different tool aimed at the same job, which is removing layers between you and the yield.

If you’re comparing how THORChain’s model stacks up against holding and staking ETH directly, our ETH staking rewards breakdown and the side-by-side THORChain vs. Ethereum comparison go deeper on the mechanics.

Curious what that looks like in practice? Explore active node operators on RUNEBond and see the current yield range for yourself.