BNY and BlackRock funnel billions through one infrastructure provider, exposing the fragile illusion of crypto diversification

BNY's Digital Asset Custody platform plans to provide institutional crypto staking support through Galaxy's infrastructure, the two firms said Aug. 4.

BNY touches roughly 20% of the world's investable assets, with $62.6 trillion in assets under custody and administration as of June 30.

Galaxy is one of three validator firms approved to stake Ethereum for BlackRock's iShares Staked Ethereum Trust (ETHB). The prospectus says the fund can stake 70% to 95% of its holdings under normal conditions.

Two of Wall Street's largest names now route institutional crypto staking through the same infrastructure provider. Galaxy also runs staking for Solana and other proof-of-stake networks, extending that overlap across multiple chains.

How institutional crypto staking separates ownership from control

ETHB owns the ETH and collects the crypto staking rewards, and its custodian holds the private keys and controls withdrawals.

Galaxy and the other approved validators hold the validator keys and perform the validation work, but the prospectus is explicit that they never gain the keys needed to move the trust's staked ETH themselves.

That structure is safer than handing tokens to a validator, but the shareholder who owns the economic exposure still has no say in how the validator behaves once it is running.

The investor supplies the economic stake and collects the yield, while the product sponsor, an ETF issuer or a bank, decides staking allocation and disclosure, and the custodian holds keys and controls withdrawal authority.

The crypto staking provider runs the validator itself, and its choices of cloud infrastructure, client software, and compliance policy also become the network's exposure.

Validators receive no token-weighted votes on Ethereum improvement proposals, but their power still lies in block production, transaction inclusion, and finality.

Ethereum's documentation says that validators controlling more than 33% of staked ETH can prevent the chain from finalizing blocks if they go offline or attest incorrectly. A share above 66% can finalize a preferred version of the chain outright.

Exchanges, bridges, and DeFi protocols all lean on finality to decide when a transaction is safe to treat as settled.

Solana labels the smallest group that can control roughly 33% of delegated stake a superminority. Nakaflow reporting put the Nakamoto coefficient at 10 as of Aug. 5, the minimum number of validators needed to reach that share.

A coordinated failure within such a small group can stop the network from voting on new blocks in real time.

The Invesco Galaxy Solana ETF filing lists Coinbase Custody as the crypto staking provider and node operator for the fund's SOL, with BNY Mellon acting as administrator.

Why total supply is the wrong number to watch

The important number is the share of active stake a provider controls, a very different figure from its share of total token supply.

About 33% of ETH's total supply is currently staked, meaning that routing roughly 11% of all ETH through a single provider would already put that provider near the one-third threshold for currently staked ETH.

Solana's staking ratio is much higher, at around 68% of supply, so reaching that same one-third share of active stake there would require about 22.7% of total SOL supply.

Figment's report for the second quarter puts its Ethereum validators at 6.26% of all staked ETH and its Solana validators at 6.96% of all staked SOL. Both numbers show how much active stake a single mid-size institutional operator can already carry.

The custodian controls the withdrawal route and private keys, so its failure or compromise can freeze customer funds even when the validator behaves correctly.

ETHB's prospectus warns that slashing, inactivity penalties, and correlated penalties across many validators can cause losses that the trust may never recover from, particularly if those validators share one staking provider.

Many institutional validators may end up using the same client software, cloud region, or key management vendor. When that happens, a single bug or outage can spread across every validator that shares the same setup.

The prospectus cites Ethereum's May 2023 finality disruption as an example of how quickly that can happen.

A single staking provider running validators for several banks and funds can apply one sanctions or transaction-filtering policy across all of them, producing a coordinated inclusion policy without anyone formally colluding to create one.

How dangerous institutional crypto staking can be

BNY's institutional crypto staking service still needs regulatory approval before it exists, while Galaxy is only one of three approved validators inside ETHB. Institutional staking can also improve operational discipline compared with token holders running validators on their own hardware.

The more dangerous version needs no bad actor at all, just ordinary institutional habits. Banks favor approved vendors, funds minimize operational risk by choosing the same infrastructure, and custody products simplify customer choice until validator selection and voting rights quietly disappear.

Ethereum's own community is already arguing about a version of this problem. EIP-8361 would burn a larger share of validator rewards as the staking ratio rises, aiming to reduce the incentive to keep piling ETH into staking.

Its authors cite custodial concentration as one of their reasons for proposing it.

A 2025 paper on Ethereum's staking market found that solo stakers respond to changes in rewards more than centralized exchanges or liquid-staking providers do.

Cutting issuance could push smaller, independent validators out first, leaving the remaining stake even more concentrated among the institutions the proposal is trying to rein in.

How this could still go either way

The bull case has disclosure catching up before concentration does. Products start publishing which validators hold their customers' stakes, cap how much of a single provider's book comes from any one client, and diversify the clients, clouds, and compliance policies underlying them.

Wall Street adds real stake to Ethereum and Solana without creating a single operational chokepoint, and staking products get safer as a result.

The bear case has yield-chasing outrunning disclosure. Staking becomes a default checkbox inside custody accounts and ETFs, and investors never see which validator holds their stake.

A handful of approved providers end up running a large share of active validators across several major networks at once. Product brands keep multiplying while the operators underneath them keep consolidating.

Investors who assumed five institutional brands meant five independent risks discover they were exposed to the same two or three operators the entire time.

The next fight blockchains will have to endure will be over who operates the stake behind them.