Reading the Hyperliquid Leaderboard: PnL Is Not ROI

A public leaderboard is one of the more honest artifacts in crypto: it shows realized results, not promises. Hyperliquid publishes a running ranking of its traders by profit and loss, and it is tempting to read the top row as "the best trader." It usually is not. To read a perp leaderboard well, you have to separate two numbers that get collapsed into one word — performance — and understand how a third number, account size, ties them together.

Three numbers, not one

Start with definitions, because the whole exercise depends on them.

PnL (profit and loss) is the dollar change in an account over a window — here, seven days of realized trading. It is an absolute number. A large PnL can come from a brilliant trade on a small account or a mediocre trade on an enormous one.

ROI (return on investment) is that PnL expressed as a percentage of the capital that produced it. It is a rate. ROI is what lets you compare a $30 million account with a $2 billion account on the same footing, because it normalizes for size.

Equity is the capital in the account right now — the base that ROI is measured against. On a perp venue equity is also, loosely, the collateral standing behind leveraged positions, so it tells you how much size a trader can carry.

The mistake is ranking by PnL and reading it as skill. PnL rewards size as much as edge. ROI isolates the edge. Neither is complete on its own, which is exactly why looking at all three together is informative.

The current top of the board

Here is the snapshot for 2026-07-10, top traders by seven-day realized PnL:

Read the first two columns alone and the ranking looks decisive. Read all four and it tells a very different story about where risk is being taken.

A worked example: same board, opposite bets

Compare the account at the top with the one sitting third.

0x24de…1b0f turned roughly $2.35 billion of equity into $298.6 million over the week — a 14.6% return. That is an extraordinary dollar figure and a strong weekly rate for capital of that magnitude. But the scale is the point: this is an account operating at a size where a modest, well-managed percentage move produces a nine-figure headline. The PnL is enormous because the base is enormous.

0x4e23…20c3 did something categorically different. On about $30.9 million of equity it booked $5.97 million — a 50.0% weekly ROI. In dollar terms it ranks third, more than fifty times smaller than the leader. In efficiency terms it is more than three times the leader's rate. That 50% is what aggressive, concentrated risk looks like when it works: high leverage, tight positioning, or both, on a small enough base that the percentage swings are violent.

Now the two middle-sized accounts. 0x393d…2109 carries nearly a billion in equity and made 0.7% — a large, cautiously deployed book that barely moved the needle in percentage terms. 0x8def…2dae, 0x8553…737f, and 0x4eb8…7819 cluster between 2% and 5% on nine-figure equity: meaningful dollars, restrained rates.

What the spread is telling you

The shape of this board — a handful of very large accounts posting low-to-mid single-digit ROIs, and one small account posting 50% — is a snapshot of how perp risk is distributed at the moment.

The dominant capital is not swinging for the fences. The billion-dollar accounts are compounding low percentages on large bases. That is consistent with size that must trade conservatively: at that scale, slippage and liquidation risk make aggressive leverage impractical, so returns come from being big, not from being reckless.

The outsized rate lives at the small end. The 50% ROI account is where concentrated conviction is being expressed. It is the most impressive number on the board and the least durable — a rate like that is as easy to reverse as it was to earn, because the same leverage that produced it cuts both ways in a bad week.

How to use a board like this

Treat it as a risk map, not a signal. Rank by PnL to see where the capital is. Rank by ROI to see where the aggression is. Cross-reference with equity to know which one you are actually looking at. A trader who posts 50% on $30 million and a trader who posts 15% on $2 billion are not competitors on the same axis — they are running different games, and conflating them is how copy-traders get hurt.

Compare traders on the axis that matters, with size and ROI side by side, in app.formion.ai.