First-order effects

• The share of ether sitting in exchange-controlled wallets fell to 14.85%, the thinnest since summer 2016 — directly shrinking the sell-side order books traders rely on and tightening immediately available liquidity.
• Staking is named as the drain: coins moved out of exchange hot wallets into locked validator positions are no longer available for spot selling, so price discovery happens against a smaller float.

Second-order effects

• Exchanges lose a second revenue lever as coins leave custody — trading fees on thinner books plus the staking yield they would have captured had the ETH stayed parked with them — pushing them to compete harder on native staking products to retain balances.
• Validators and liquid-staking operators become the new custodial chokepoint, absorbing the market share exchanges are shedding; the later record ~36M ETH, or 30% of total supply, staked confirms this pull only deepened.

Third-order effects

• Crypto is structurally splitting its supply between two concentration models: ether increasingly locked at the protocol layer via staking, versus Bitcoin where Gemini and Glassnode research finds centralized treasuries — governments, ETFs, public companies — control about 30.9% (~$668B) of circulating supply; in both cases, less coin trades freely and on-chain data becomes the primary gauge of true market float.
• If staking continues absorbing supply, exchange-reported volumes become an ever-less-representative proxy for market activity, shifting how analysts, regulators, and index builders measure liquidity.