How to Track Crypto Whales by Identity and Track Record

Whale vs Smart Money and Why Track Record Beats Size

A crypto whale is any wallet whose transfers get flagged by size alone. Smart money is a performance label given to wallets with a tracked record. Neither label tells you whose wallet it is, and that gap decides whether the signal is usable.

Learning how to track crypto whales starts with the fact that whale labels are size-based, and no threshold is universal. Whale Alert fires on transfer value alone, with a $100,000 minimum on its API and $20M to $200M per coin on its official social feed. The feed's bar is lower when the sender or receiver is a known address, such as an exchange: $20M instead of $50M for most ERC-20 tokens.

Smart money is a different animal. Performance labels are dynamic and get retired when the record fades. A smart-money label is a track record, not a balance.

Nansen runs the reference version. Its Smart DEX Trader (All Time) label requires at least $1.5 million in cumulative realized profits with no size test, so the label is earned by trading record alone. Arkham answers who owns an address and Nansen answers how well it trades, and a whale routine needs both answers.

Arkham sits in a third camp. Its intel guide separates entities (who), labels (a name for an address) and tags (what the wallet has done), and publishes neither a size nor a performance criterion for its whale tag.

Deal activity shows the same gap on the fund side. When we sorted every investor in the DropsTab database by deal count on October 25, 2026, the top spot went to Coinbase Ventures with 358 deals. It led just 6.1% of those rounds, yet it is the top co-investor for five of the eight major funds we profile in the fund holdings section.

Size, labels and rank all flag attention; they do not explain a move. This guide tracks whales by identity and track record, and blind copying fails even when the wallet has a name.

Types of Crypto Whales and Which Public Portfolios to Track

Public crypto portfolios split into five types, each proven to a different degree. Corporate treasuries file with the SEC and country treasuries publish official addresses, while funds self-disclose deals, public figures claim addresses and anonymous whales rest on estimates. Match the tracking method to how well the whale is proven.

Crypto whales come in five trackable types, and the type sets how well its holdings are proven. Filed covers corporate treasuries that disclose holdings in SEC filings. Official covers country treasuries that publish addresses, with El Salvador the contested case. Self-attested covers funds, market makers and public figures who claim addresses or disclose deals but not wallets. Researcher-attributed covers wallets linked by named researchers. Estimated covers pattern attribution, the weakest level of proof, where Satoshi sits. The better the proof, the less guessing.

The list below gives each whale's primary source and date.

El Salvador is the contested row. The donation-versus-purchase dispute over its holdings sits in the Mislead section below.

Can you follow someone's crypto portfolio? On DropsTab, yes for named holders, no for funds. Public portfolios give each named holder a read-only tracker showing Followers, Assets, 24h and 7d changes, Total PNL and Activity, plus upcoming unlocks of the portfolio's assets. Follow works here and only here. Funds are browsed, sorted and ranked on the investors pages.

A tracker is only as reliable as the source behind its wallet label. For building and publishing a portfolio of your own, our DropsTab portfolio tracking guide covers the feature end to end.

How to Track Whale Wallets Back to a Verified Owner

Wallet ownership is proven with different reliability. A signed name tag or a regulatory filing is the strongest proof, entity clustering by heuristics or AI predictions the weakest. Each level has documented errors, so check an ownership claim against the level it came from.

Etherscan name tags are the strongest public proof. Anyone can suggest a tag, but individuals are only tagged once they have publicly declared ownership, and falsely claimed ownership is Etherscan's documented failure mode. Removal needs a signature from the tagged address.

Arkham builds entities from four sources, AI clustering, off-chain records, a paid Intel Marketplace reviewed before posting, and in-house analysts, and says it covers 800K+ entities and 3.5B+ labels. It publishes no accuracy metrics and no dispute process, and both gaps have bitten. In the April 2025 OM crash, wallets carrying an AI-predicted "Laser Digital" label fed an insider-dump narrative, and Laser Digital denied ownership the next day.

Research accounts usually work one level lower, with researcher attribution: a wallet "linked to" an entity with no signed tag. Lookonchain flagged a suspected Ethereum Foundation sale that way on X.

A wallet linked to the #EthereumFoundation sold 2,785 $ETH ($6.42M) at $2,304 over the past hour.

"Linked to" is a research attribution. Treat the sale as the Foundation's only once a name tag or the Foundation itself confirms the wallet.

For a worked example, vitalik.eth carries a self-attested ENS name, so the address reads as a single person. On October 25, 2026, we pulled the address's 12 most recent ERC-20 transfers through the Etherscan API. All 12 were inbound, and 11 of 12 were unknown tokens, three of them minted straight from the zero address. The only recognizable deposit was 1.00 USDC. The ENS name tells you whose address it is, not that these transfers were his doing: other people sent them.

With funds, a name can mislead the other way, because one name can cover several organizations. DropsTab's investor pages list three separate profiles named Jump (Jump Trading, Jump Crypto and Jump Capital), and Jump Trading and Jump Crypto share just one portfolio coin, Solana. Names on Hyperliquid are weaker still, because wallet owners pick and register .hl names themselves. On October 25, 2026, we searched for "Jump" and "Wintermute" on the Hypurrscan explorer. "Jump" returned two such names, jump.hl and jumper.hl, and neither is verified as the fund's wallet; "Wintermute" returned nothing. That is why our Hyperliquid smart money guide does not treat a trader's handle alone as proof of who they are.

Set Crypto Whale Alerts That Filter Out the Noise

Alert noise is aimed at exactly the wallets people watch. Poisoning and spam target rich, active addresses. Four filter families cut most of it, direction, a value floor, labeled destinations and excluded mints.

The noise is industrial, and it is aimed at the wallets people watch. A Carnegie Mellon team (Tsuchiya, Dong, Soska and Christin) measured it in a peer-reviewed USENIX Security 2025 paper on address poisoning. Across Ethereum and BNB Smart Chain transfers from July 2022 to June 2024, they found 270 million poisoning transfers sent to 17 million addresses, with attackers favouring rich, active wallets. Only 6,633 attacks succeeded, yet they cost victims at least $83.8M. We cite it because it is the largest independent count of this spam, and it explains why a watched wallet's raw feed fills with junk.

Lookonchain, the account behind the Ethereum Foundation post above, framed three HYPE deposits to Coinbase as the moves of one probable whale.

Whales are dumping $HYPE! 3 wallets, possibly belonging to the same whale, deposited 494,310 $HYPE ($37.83M) into #Coinbase over the past 24 hours.

Four filter families do most of the work, and all four are documented in primary docs.

The Satoshi genesis address is the extreme case. When we checked it on mempool.space on October 25, 2026, it had about 67,100 confirmed transactions and 0 BTC ever sent, so every event on it is an unsolicited deposit and none is a move by the owner.

Apply that to a named target. For vitalik.eth in Drops Bot, set direction to OUT, set a transaction-value floor, note in the wallet label how well its owner is proven, and hide the addresses that keep sending it spam.

The bot tracks 20 wallets free across 30 networks and needs only a public address, with no wallet connection. Its wallet filters set the direction (IN, OUT or ALL), a minimum transaction value, the event types, such as transfers, DEX buys and sells or token approvals, and allowed or hidden addresses. Alerts arrive in Telegram, and our guide to Drops Bot and its token unlock alerts covers the full setup.

If you keep IN alerts on, coin filters deal with the spam tokens. They mute alerts for tokens below the market cap, 24-hour volume or holder count you set, on EVM networks, Solana and Tron, and the holder filter works on EVM only. The direction filter alone would have suppressed all 12 transfers in our October 25 sample.

How to Track Crypto Fund Holdings Through the Rounds They Lead

Funds do not disclose wallets, they disclose deals. The visible track record is the share of rounds a fund leads and how recently it led one. A fund that co-invests in many rounds is a popular partner, which does not mean it picks the deals.

Wallet tracking does not work for funds: they disclose deals themselves and keep their addresses private. What you can see is the rounds a fund led, meaning it picked the project and set the terms instead of joining someone else's round. We went through each fund's DropsTab investor profile and counted that share. Paradigm led 61.5% of its 109 rounds and Polychain 57.0% of 214. Coinbase Ventures led 6.1% of 358 and Wintermute 3.7% of 81, roughly 16 times less often than Paradigm.

The date of the last led round shows which funds are investing now and which have paused. Polychain last led a round 10 days before our snapshot, Jump Crypto 17 days and Wintermute 19, while DWF Labs sat at 255 days and Jump Trading at 326.

The major funds on the list are rarely in the same deals. Across the eight major funds profiled, 775 of 934 portfolio entries are distinct coins and only Solana sits in five of eight portfolios. Even a16z and Polychain, matched at 215 versus 214 deals, share just 25 portfolio coins. There is no consensus deal flow to copy.

An unlock is not a sale until tokens reach an exchange deposit. Worldcoin's vesting schedule on DropsTab shows daily unlocks across seven allocations, and a standard OpenZeppelin vesting wallet releases linearly to a beneficiary, but the sell-intent step is the post-unlock transfer to a labeled exchange deposit.

Analyst Dylan Hart put the same rule plainly on X.

The loudest whale transfer is not automatically a sell signal. Coins can move between custody, market makers and internal wallets without reaching an exchange. Until the flow meets order-book pressure or confirmed deposits, the alert is incomplete evidence.

The counter-cases matter just as much. Movement's 66M MOVE "dump" in December 2024 came from a market maker liquidating loaned inventory. Hyperliquid's Core Contributors allocation, 23.8% of supply, unlocked in monthly tranches starting January 6, 2026, as our HYPE vesting page shows. A tranche date only says when tokens can move. A sale shows up later, as a deposit to an exchange.

Here is the workflow on one fund, Polychain. The goal is to see whether the fund is investing now, what it holds and when its coins unlock.

• Is the fund active? Polychain's investor profile shows it led 122 of its 215 rounds. The Funding Rounds tab lists every round with its date, stage and investors. The latest is dated October 15, 2026, so the fund is still writing checks.
• What does it hold? The Portfolio tab of the same profile lists the fund's coins with the private-round price and the private and retail ROI. The Investments Allocation chart on the Funding Rounds tab splits its deals by category and stage.
• When do its coins unlock? Each coin has a vesting page with its unlock schedule, like Celestia's, one of the coins in Polychain's portfolio.
• How not to miss the unlock. Drops Bot sends Telegram alerts for Seed and Private round unlocks, the rounds funds usually buy into.
• New deals. The funding-rounds feed lists fresh rounds and who led them, so a new Polychain deal does not slip by.

When Whale Signals Mislead Even With a Name Attached

A named wallet is a verified identity, not a verified signal. Custody, trustee and donation flows look like whale moves on chain. An unlock only becomes a sale at an exchange deposit, and fame is not a track record.

• Custody flows. BlackRock-linked wallets moved about $300M to Coinbase Prime on January 13, 2026, and a second move above $430M followed on January 22, both framed as ETF creation, redemption and settlement flows. Grayscale's 4,000 BTC to Coinbase Prime in January 2024 did coincide with redemption selling. Only knowing whose wallet it is and what that owner does tells the two apart.
• Donations read as purchases. El Salvador's holdings kept growing during 2025 even as its IMF program's Bitcoin-accumulation targets went unmet. The IMF's October 1, 2026 review granted waivers for the unmet Bitcoin-accumulation criteria, and in September the IMF stated that documentation verified the accumulation since the first review "reflects private donations", with no further accumulation expected beyond them.
• Trustee reshuffles. The Mt. Gox trustee moved about 30,000 BTC, some $2.4B, on November 11, 2024, an estate our crypto bankruptcy tracker follows alongside FTX and Celsius.

Tracing a seller back can also end at a custodian. Nazo_ku followed an AAVE seller's funds on X.

Whale wallet (0xb38) dumped a total of 77,000 $AAVE worth $11.173m at avg. $145 over 2 days. … Tracing the funds led to a custody wallet at Anchorage Digital that received the tokens over a year ago.

• Fame is not a track record. Murad, whose 11 wallets ZachXBT attributed in October 2024, ran up nearly $60M in unrealized losses, mostly on SPX6900. Machi Big Brother, one of the most-watched on-chain traders, is the same lesson at larger scale, a public wallet that turned from signal into a cautionary tale.

Copy-trading research points the same way. A 2019 study in Finance Research Letters by Jin, Zhu and Huang used data from a popular Chinese social trading platform. People who followed others' trades began trading more often and drifted into more volatile stocks, while the original leading trades did better than the copies that came after them. A 2025 paper in the Journal of Behavioral and Experimental Finance by Klocke, Müller-Okesson, Hasso and Pelster analysed trading records from a social trading brokerage. Investors trade more when their peers post good returns, and their own returns then fall and swing more. Both studies come from traditional markets, and neither covers crypto. We cite them because they are the closest measured evidence on what happens to people who copy a visible trader, which is exactly what a public wallet invites.

The best-known study of a single large flow cuts the other way. In the Journal of Finance (2020), Griffin and Shams traced Tether flows on the blockchain. Tether purchases timed after market drops during the 2017 boom, attributable to one entity, were followed by sizable Bitcoin price increases. The study shows that one large flow can move price, but it documents manipulation in a single episode and gives no rule for trading whale alerts.

A Whale-Tracking Routine You Can Run Every Week

One hour a week is enough to check every whale on the list. Check how well each whale is proven, re-verify one address, clear alert noise, check fresh lead rounds, and cross one unlock against an exchange deposit before calling it a sale.

• Open your whale list and check how well each whale is proven. Filings for treasuries, the IMF review for El Salvador, public portfolios for named figures, investor profiles for funds.
• Re-check one address owner level by level, from name tag to entity page to the deals behind the entity.
• Clear the alert queue with the alert filters, direction first, then the value floor, then mints. Drops Bot holds 20 wallets on the free plan.
• Check the funding-rounds feed for rounds led this week, then open the lead fund's investors profile.
• Cross one upcoming unlock on the vesting calendar against actual exchange deposits before reading it as a sale.

No published study measures copier latency on crypto venues; the lag evidence comes from social trading on traditional brokerages, so treat every copy window as unmeasured until your own fills say otherwise. Our guide to crypto copy trading covers how copy setups work.

Where to Run the Routine on DropsTab

Every check in the routine can be done free on DropsTab. Named treasuries, funds and public figures sit on public portfolios with Followers, Assets, 24h and 7d changes, Total PNL and Activity columns, plus the upcoming unlocks of each tracked asset.

The investors list sorts funds by lead investments and shows the last round date on every profile, and Drops Bot delivers the wallet alerts in Telegram with IN and OUT filters, a value floor and 20 tracked wallets on the free plan. Follow named whales free on DropsTab public portfolios to keep every whale on your list on one screen.

Check every alert against three questions: whose wallet it is, how we know it belongs to that owner, and what the owner has said or filed. If you cannot answer them, the alert only tells you that a large transfer happened, not who made it or why.