Ethereum Whale Convergence Snapshot: April 22, 2026

This post is a timestamped snapshot of Ethereum whale buying activity. Nothing about it is forward-looking — it is a report of what happened in one 48-hour window, reproduced here so readers can see the raw data and draw their own conclusions. The cutoff is 2026-04-22 18:38 UTC; anything that happens after that is outside the scope of this post.

We have been asked frequently what "whale convergence" looks like in practice when the usual noise is removed. This post answers that question for one specific window of time. Everything below is descriptive.

Methodology

Convergence is defined here as the number of distinct wallets that bought a token in the observation window. Aggregate USD volume matters but it is not the headline metric, because one large buyer does not establish convergence — many buyers do.

The filters applied to the raw on-chain capture before counting:

• Buys only. Direction restricted to BUY, SWAP_BUY, and DEX_BUY.
• Window: 48 hours, anchored to the latest transaction in our database at the time of publication (2026-04-22 18:38 UTC).
• CEX and market-maker hot-wallet flow excluded. We classify wallets as likely exchange/MM when they trade 5+ distinct tokens per day at $1M+ gross volume over an extended period, or show high two-way churn across many tokens. This removes inventory rebalancing that can look superficially like directional accumulation. 119 wallets were flagged this way and removed from the count.
• Stablecoins and wrapped assets removed. USDT, USDC, DAI, FRAX, USDS, PYUSD, sDAI, USDe, WETH, WBTC, cbBTC, tBTC excluded from the token list.
• Threshold: at least 5 distinct directional wallets and $100,000 in total USD buys within the window.

The resulting list below is the complete set of tokens that met those criteria, sorted by number of distinct wallets.

Full data table

A note on interpretation

The most common way to misread a table like this is to treat a high wallet count as a directional signal. It is not. What the table shows is that in a specific 48-hour window, a particular token attracted buys from many distinct directional wallets at once. Why those wallets bought — whether it was a response to news, a listing, a narrative rotation, an airdrop farming strategy, passive rebalancing, or unrelated private decisions — is not something this data alone can answer. We have written elsewhere about why raw-transaction whale alerts often fail to produce tradeable signal on their own, and the same caution applies here.

Many of the wallets at the top of each list are not the ones you would find on a published "smart money" leaderboard. They are mid-rank active traders that do not have long public track records in our tracker. That is part of what we think is interesting about the result — the convergence is distributed across a wide set of mid-tier wallets, not concentrated in a handful of celebrated addresses. For context on the different wallet archetypes we observe, see our taxonomy of 8 behavioral whale types.

The $LINK concentration

197 distinct wallets bought $LINK in the 48 hours under review, totaling $20.9M — the highest convergence by wallet count in the entire universe of tokens we tracked in the window. The largest single buyer (ranked #105 in our tracker by lifetime USD volume) bought $5.1M across three transactions. Their lifetime footprint in our tracker is 29 trades. Four other wallets bought more than $1M each in the same window.

Again, this is a description, not an explanation. Many things can cause an event like this — from macro news to ecosystem announcements to simple technical patterns — and we make no claim about which of those, if any, drove this specific cluster.

Two gold-backed tokens, same window

A pattern worth flagging is that two separate gold-backed tokens — $PAXG and Tether Gold ($XAUt) — both appeared in the high-convergence list in the same 48-hour period. $PAXG drew buys from 97 distinct wallets for $11.0M, and $XAUt drew buys from 74 wallets for $5.1M. The two lists do not appear to be the same wallets buying both — the overlap between the top buyers on each token is limited.

We flag this purely because convergence on two independent gold-pegged assets in the same short window is not the default baseline. Whether it reflects a broader rotation, coincidence, or something else is not a question this snapshot can resolve. For context on how we think about on-chain signal strength more generally, see our writeup on the six on-chain signals that reveal what ETH whales are doing and our study of whale accumulation signals across 214 events. Readers who want to look into this specific window further can pull the raw data below.

Individual buyer data

For each token in the list above, the top five individual wallet buyers by USD in the window — including the rank in our tracker, lifetime volume, and lifetime trade count — are published in the CSV export below. Every wallet address is publicly verifiable on Etherscan.

Reproducibility

This snapshot was generated from our on-chain indexer at 2026-04-22 18:38 UTC. It is not regenerated automatically — the table in this post is static and refers to that specific window. If you want a live view of the same kind of data, visit the Trends dashboard or the Tokens page, both of which update block-by-block. The whale wallet leaderboard surfaces the addresses referenced in this snapshot with their full live trade history.

We intend to publish similar observational snapshots at irregular intervals. They will always be timestamped and will not include forward-looking claims. Earlier data-driven posts in this series include our analysis of aggregate vs per-wallet whale dry powder and our study of 37,099 whale-to-CEX deposits.

Limitations

• Our tracker covers Ethereum mainnet only. Whale activity on Solana, Base, Arbitrum, or other chains is not in this dataset.
• A wallet we have not yet added to our universe will not appear in these counts even if it is trading in size.
• The CEX/MM classifier is heuristic and not perfect — it is designed to exclude obvious hot-wallet behavior but may over- or under-reach on edge cases.
• Reported USD values are captured at the time of each swap and are not adjusted for subsequent price changes.