Loading your account

What a Hyperliquid whale tracker can and cannot tell you

In shortA Hyperliquid whale tracker shows what large accounts did: entries, exits, size against the whole account, realized results and drawdown. It cannot show intent, hedges held elsewhere, your own result, the next market regime, or which trades you could have entered in time. Our board held 199 live whales on 2026-10-02.

A Hyperliquid whale tracker can tell you what large accounts did, in public, after the fact: when a position opened, how big it was against the whole account, when it closed, what it realized, and how far it went against the trader on the way. It cannot tell you why the trade was placed, what the trader holds somewhere else, what your own result would be, whether the next market will reward the same habits, or which winning trades you could actually have entered in time.

We run one, so our own board is the worked example. On 2026-10-02 it listed 199 live whales; our methodology note (2026-09-02) puts the universe they are drawn from at 1,300+ Hyperliquid wallets, evaluated on a rolling 90-day window. The median whale's median hold was 17.2 hours, and the median whale produced 0.4 signals a day over the previous 30 days.

We never place an order and never hold funds. A tracker is intelligence, not execution, and the exit is always yours.

What is a Hyperliquid whale tracker?

A Hyperliquid whale tracker is a tool that watches large accounts on Hyperliquid and reports what they do. Hyperliquid's data is on-chain, so any account's positions can be read by anyone; the tracker's work is choosing which accounts deserve reading and turning their activity into a board, a profile or a stream of whale alerts.

Choosing the accounts is the product. A raw feed of every large position is easy to build and hard to use, because size alone says nothing about skill, about risk, or about whether a trade could be followed at all. What separates one tracker from another is mostly the filters it applies before a wallet reaches you, and whether it publishes them.

You can also track Hyperliquid wallets by hand, and our step-by-step guide to tracking Hyperliquid whales walks through it. This page is about the limits every tracker shares, ours included.

What can a whale tracker tell you?

A whale tracker can tell you what a large account did and how it turned out: its entries, its exits, its size against the whole account, its realized results and its drawdown. Each of those is a fact about the past, and each is easy to misread.

What the tracker showsWhat it meansWhere people misread it
An entryThe whale opened a position: coin, side, size, timeAs an instruction, when the trade's grade is unknown until it closes
Size against the whole accountHow much of its own capital the whale committedAs dollars alone, when a large position can be a small slice of a large account
An exitThe trade is over and its result is realizedAs your exit, when you entered later, at your own size, and the exit is yours
Realized resultWhat the whale closed with, on its own whole accountBy swapping in the peak, the most a position was ever up, which nobody banks
Win rate, with its countThe share of closed positions that wonAs money, when on our board it tracks return only weakly
DrawdownHow far a position went against the whale's accountAs a promise, when a colour is a measured band and not a forecast

Size against the account is the row our entry rule is built on. Our packages follow only entries where the whale commits 5% or more of its own account, because on a large account a small probe and a committed position can look alike in dollars.

The realized row is where a record most often gets inflated. We measured 14,114 reconciled closed positions with a recorded peak and worst move, by our live whales over the 30 days to 2026-10-02, from each whale's own entry to its own exit, as a price move. The median peak was a 1.4% move in the whale's favour, while the median realized move was 0.2%.

The gap between peak and realized is not an edge case. Of those 14,114 positions, 2,218 (15.7%) were up 1% or more at their peak and still closed at a loss, and the 8,391 winners kept a median 48.5% of their peak. A tracker that shows you the peak without the close is showing you a number nobody banked.

Win rate needs its count and its limits. Across the 187 whales with enough 30-day closes to rate on 2026-10-02, the rank correlation between 30-day win rate and 30-day return was 0.24: related, but weakly. Our board's median 90-day win rate was 60.5%, the median count of positions behind those win rates was 82, and neither number says anything on its own about money.

The drawdown row is the one that sets a colour. A whale's colour on our board comes from one number, the deepest drawdown a single position put on its whole account: white is under 5%, green 5% to 10%, yellow 10% to 15%, orange 15% to 20% and red 20% or more. A colour describes risk the whale has already taken, and it promises nothing about the next trade.

A drawdown in the middle of a trade is not a verdict on that trade. In the same 30-day sample, the price moved 5% or more against the whale on 2,286 positions, and 1,143 of them (50.0%) still closed in profit. A follower's stop set inside a whale's ordinary drawdown gets hit on some of the trades the whale goes on to win.

What can't a whale tracker tell you?

A whale tracker cannot tell you why a trade was placed, what the trader holds elsewhere, what your own result would be, whether the next market regime will reward the same habits, or whether you could have entered in time. These are limits of public position data, so they apply to every tracker, ours included.

Intent

A position records direction, size and timing, not reasons. The same long can be a conviction trade, a hedge against something held elsewhere, or a small probe, and the chain records all three the same way. Filters can remove the smallest positions, but no filter reads a mind.

Hedges held elsewhere

We measure each wallet on whole capital, spot and perpetual balances together, because a single position in isolation says little about the account behind it. Even whole capital stops at the edge of the account: a short on Hyperliquid can be offset by a long on another exchange or in another wallet, and Hyperliquid's public data will not show it. A position that looks like a bold bet can be one leg of a flat book.

Your own result

Every return on our board is the whale's own, and every realized return is on its own whole account. We do not know when you would have entered, at what size, with what fees, or when you would have left, so we never state or imply a follower's result. The whale's trade and a follower's trade share a coin and a direction and not much else, which our post on why copy traders lose money even when the trader wins works through.

The next regime

Every number a tracker shows describes the past. Our 90-day window covers one market regime, no walk-forward test of our filters across regimes exists yet, and every threshold we use was derived on that window and may not survive a different one. Records drift inside a regime as well: of the 196 live whales with a frozen admission baseline on 2026-10-02, 117 had a lower win rate than when they were admitted, with a median change of −0.8 points.

The trades you could not have taken in time

An alert fires when a position opens, but a closed trade's grade is knowable only when it closes, so a tracker's win rate can describe trades a follower could never have taken. One wallet in the backtest our methodology note publishes won 78.4% of the 51 positions our metrics graded, and 44.3% of the 149 positions a follower could actually have entered: same wallet, same 90 days, a 34-point gap. It had the lowest monthly ROI of the five wallets in that test; it topped the set on win rate, not on money.

Speed is the other half of followability. A wallet in a separate test ranked #2 of 20 on our approved ranking metric and closed 71.3% of its positions before a follower could act: excellent, and un-followable. Wallets like that one are why our filters require a median hold of at least 2 hours; on 2026-10-02 the median whale on our board had a median hold of 17.2 hours, well clear of that floor.

How we built our Hyperliquid whale tracker

Our Hyperliquid whale tracker is a filter first and a feed second. We evaluate a universe of 1,300+ Hyperliquid wallets on a rolling 90-day window, recomputed daily, and measure every wallet on whole capital rather than one position at a time.

Six filters, all required

For the engine to admit a wallet to the board, it must clear all six filters; these are the admission thresholds it runs:

  • At least 50 measured positions. Under that, a lucky streak and a real edge look the same.
  • A win rate of 60% or better, after fees and funding, on positions a follower could have taken.
  • A max concurrent drawdown under 50% of whole capital, meaning the worst simultaneous open loss the account ever carried.
  • A median hold of at least 2 hours, a rule about whether a wallet can be followed rather than about its quality.
  • All-time profit above zero, a sign test rather than a threshold.
  • An account value of $100,000 or more on whole capital, re-checked every night.

Passing once is not a permanent seat. The set is re-scored daily and a better wallet takes a worse one's place, while a demoted wallet keeps its page, so its history stays visible after it leaves.

The six filters are admission thresholds. A whale's figures keep moving after it is admitted, as the drift figures above show, so a whale on the board today can show a win rate below the one it was admitted on.

The account gate

The account gate was added on 2026-09-13, after two wallets on the board turned out to hold almost nothing alongside a few thousand in profit, which made every ratio on their cards meaningless. A return on nothing is not a return, so a whale whose account falls below the line leaves the board, and its page says so.

Colours, and the board on 2026-10-02

A colour is assigned from the deepest drawdown a single position put on the whale's whole account, using the bands above, and a wallet needs at least 50 measured positions before its colour means anything. On 2026-10-02 the board held 199 live whales: 60 white, 32 green, 25 yellow, 22 orange and 60 red.

How do you judge any whale tracker?

Judge a whale tracker by what it admits, not by the size of its numbers. The questions below work on any tracker, ours included, and a good one answers them on the page rather than in a footnote.

  • Does every rate carry its count? A win rate with the number of closed positions behind it can be weighed; a bare percentage cannot.
  • Is the return on the whole account? A return on one position says nothing about how much of the account was at stake.
  • Is realized kept apart from peak? A peak is the most a position was ever up, and a tracker that blends the two is inflating its record.
  • Does it grade the trades you could have taken? Ask whether its win rate counts positions from the moment an alert could have fired, or only positions that qualify once they have closed.
  • Does it filter for hold time? A wallet that closes most positions before an alert can reach you can be watched but not followed.
  • Does it re-check its whales and show the ones that left? Records drift, and a list that only ever adds is a list of past winners.
  • Does it say what it does not know? Regime, sample size and your own result are limits every tracker has, and the honest ones name them.
  • Does it place orders or hold your funds? If so, it is closer to Hyperliquid copy trading than to a tracker, and it carries different risks; we do neither.

Where to start

Start with risk, then followability, then performance. Read a whale's colour and drawdown first, then its median hold and how often it signals, then its win rate with its count, and only then its realized return on its whole account.

The method behind the numbers on our board, from the sample and the two populations of trades to the six filters and what we do not know, is in how we measure whales. To read Hyperliquid's own public rankings with the same scepticism, see how to read the Hyperliquid leaderboard.

A tracker is a source of facts about other people's trades. Whether to take one, at what size and when to exit are decisions it cannot make for you, and ours does not try.


Every figure above is a whale's own realized result on its own account, over closed trades only. Peak is the most a trade was up before it closed — nobody captures it, and it is never added to anything. A colour is a measured drawdown band, not a promise. How we measure.

Common questions

What is a Hyperliquid whale tracker?

A Hyperliquid whale tracker is a tool that watches large accounts on Hyperliquid, whose positions are public on-chain, and reports their entries, exits, size against the account, realized results and drawdown. Ours filters a large universe of Hyperliquid wallets on a rolling 90-day window, recomputed daily, and on 2026-10-02 its board listed 199 live whales. It is intelligence, not execution: we never place an order or hold funds.

Can a whale tracker show my own result from following a whale?

A whale tracker cannot show your own result. Every return it shows is the whale's own, while your entry comes later, your size and fees differ, and your exit is your own decision. We never state or imply a follower's result anywhere on our site, and we never place an order or hold funds on anyone's behalf.

Is a whale tracker the same as copy trading?

A whale tracker is not copy trading. A copy trading service mirrors trades into your account automatically; a tracker only reports what large accounts did. We never place an order and never hold funds: our signals tell you what a whale opened and where the risk line sits, and you decide whether to act, at what size, and when to exit. The verb we use is follow.

How many whales does DarkTrade track on Hyperliquid?

The count is a dated measurement, not a fixed number. On 2026-10-02 our board listed 199 live whales, 60 white, 32 green, 25 yellow, 22 orange and 60 red, drawn from a universe our methodology note put at 1,300+ Hyperliquid wallets, evaluated on a rolling 90-day window. The set is re-scored daily, so whales join and leave, and a demoted wallet keeps its page.

Why can a whale with a high win rate be hard to follow?

A whale's win rate is graded when trades close, while an alert fires when they open. In the backtest our methodology publishes, one wallet won 78.4% of 51 graded positions but 44.3% of the 149 a follower could have entered, a 34-point gap over the same 90 days. A wallet in a separate test ranked #2 of 20 on our ranking metric yet closed 71.3% of its positions before a follower could act.

What does a whale's colour mean?

A whale's colour is a measured drawdown band: the deepest drawdown a single position put on the whale's whole account. White is under 5%, green 5% to 10%, yellow 10% to 15%, orange 15% to 20% and red 20% or more, and a wallet needs at least 50 measured positions before its colour means anything. A colour describes risk already taken; it is not a forecast or a promise.

Everything above is measured from wallets you can check yourself on the Whales tab.

Whale data is public on-chain · not financial advice