Philadelphia Fed: Bitcoin wallets react faster to whale alerts

The Federal Reserve Bank of Philadelphia published a working paper in September 2026 examining wallet behavior after public notifications of large on-chain transfers, commonly called whale alerts. The paper compares responses on the Bitcoin and Ethereum networks and reports short-term patterns in trading activity and realized volatility.

The study uses an event-study design. Researchers defined a whale wallet as any address that had made at least one transfer worth more than $50 million, excluding addresses tied to exchanges and smart contracts. To isolate single events, the sample kept only transfers with no other whale transaction within two hours on either side. That produced 6,645 Bitcoin and 5,075 Ethereum whale-transfer events for analysis.

On Bitcoin, the clearest response occurred within the first 15 minutes after an alert. For whale buys, buy participation among small non-whale wallets rose by 14.81 percentage points, by 23.72 points for medium wallets and by 3.50 points for large wallets. After whale sells, sell participation rose by 12.95 points for small wallets, 29.52 points for medium wallets and 2.95 points for large wallets. The elevated same-direction activity declined toward baseline within about an hour.

Ethereum showed a different pattern. Post-alert participation among small, medium and large non-whale Ethereum wallets remained broadly stable. The most consistent same-direction response appeared only in the largest non-whale cohort after whale sells. Medium Ethereum sellers reached a weaker 10% statistical significance in the authors’ tests. The sample included wrapped Bitcoin (WBTC) transfers on Ethereum; the response to WBTC alerts was statistically indistinguishable from zero.

The paper also reports short-term realized volatility around alerts. Bitcoin alerts were associated with a temporary rise in realized volatility at short horizons, but by 24 hours the net effect on Bitcoin volatility had reversed. Ethereum realized volatility tended to be lower after alerts. Responses to WBTC alerts did not show a clear volatility effect.

The authors note differences in market structure as a possible factor: Ethereum activity often routes through exchanges, smart contracts and layer-2 venues where user transactions can be aggregated into larger balance transfers. The contrast in results persisted across Ethereum’s September 2022 transition to proof of stake.

The working paper lists several limitations. Wallet-size groups are transaction-based proxies and a single owner can control multiple addresses. Exchange flows were excluded from the analyzed wallet activity. The event study documents statistical associations around public alerts but does not establish that the alerts caused every observed trade or volatility change.