The Hyperliquid vault leaderboard, and why the official list isn't enough

Hyperliquid's official vault list ranks by a flattering APR and sweeps the graveyard off the page. Here's the leaderboard I built instead — every number recomputed from public data, nothing taken on faith.

When you open Hyperliquid's vaults page, the first thing you see is a list sorted by APR. It looks like a leaderboard. The biggest number sits at the top, and your eye goes straight to it.

It took me a few months of watching that page every day to realize it isn't a leaderboard at all — at least not the kind that helps you decide where to put money. It's a list of what's running right now, ranked by the most flattering number we have on each one. Those are not the same thing.

So I built my own. It lives here: the Altcopy vault leaderboard. This post is about what it does that the official list doesn't — and why every one of those differences came from getting burned by the official one.

The problem with a headline APR

A vault's APR on the official page is an annualized return taken from a recent window. Annualize a good few weeks and you get a triple-digit number. It's not a lie, exactly. It's just the single most forgiving way to describe a vault, and it's the one printed in the largest font.

Three things that number quietly hides:

  • The drawdown. A vault can show a beautiful APR and still have lost 60% of its value at some point along the way. The APR doesn't carry the scar. The worst peak-to-trough fall — the drawdown — is the first thing I want to see, and the official list doesn't show it. (It's also the number that makes people sell at the worst possible moment.)
  • The base. A vault that started with a few hundred dollars and got lucky can post an APR of 2,000%. It's an artifact of a tiny base, not skill. The headline treats it exactly like a vault that compounded real money for two years. This is the same trap as a raw return percentage with no risk behind it — why raw percentages mislead is a whole post on its own.
  • The graveyard. This is the big one. The official list only shows you what's alive. It doesn't tell you that roughly two out of three Hyperliquid vaults ever created are already closed, drained, or abandoned. When you only see the survivors, every survivor looks like a good bet. I dug into that graveyard — and the fee now built on top of it — in The $10,000 toll.

Sorting the living by their best number, with the dead swept off the page, is how a list flatters you. I didn't want to be flattered. I wanted receipts.

The receipts principle

Every number on my leaderboard is recomputed from Hyperliquid's public API — nothing is taken on faith, including the APR itself.

The most important consequence: I rebuild returns from the profit-and-loss curve, not the equity curve. When someone deposits into a vault, its equity jumps — but that's not profit, it's just new money walking in. Read the equity curve naively and you'll mistake deposits for performance. I wrote a whole piece on that exact confusion — why a vault being "up" isn't the same as having made money. Working from realized P&L strips the deposits out. It's a boring, technical choice, and it's the difference between a real number and a marketing number.

Because everything is recomputed the same way for every vault, the board can do things the official list can't: rank by risk-adjusted return, show the real gross leverage a vault runs, surface how concentrated its depositors are, and flag the vaults whose gaudy numbers are just small-base mirages. What each of those means in plain English is the third post in this series.

Three questions before "what's the return?"

Here's the mental shift the leaderboard is built around. Before I ask how much does it make, I ask three questions the official list won't answer for you:

  1. Is it actually trading? A vault can be open and look fine while its manager has quietly stopped trading weeks ago — the equity curve keeps drawing a flat line and you'd never know. I check recent trading volume, not the equity line. No volume in the last month, and the vault gets a dormant flag. The official list will happily rank a frozen vault on the strength of a return it earned and then walked away from.
  2. Can you even get in? A surprising number of the best-looking vaults are closed to deposits — full, paused, or locked. Recommending one of those is like recommending a restaurant that stopped taking reservations. My board marks them and, in any recommendation view, hides them. You only get suggested what you can actually act on.
  3. Is it real, or a tiny-base mirage? Anything with a track shorter than two months, a near-zeroing blowup, or numbers so extreme they can only come from a tiny base gets a visible caution flag — dimmed rather than deleted, so you can see why a 2,000% APR isn't something to chase. The drkmttr blowup is what that flag is trying to keep you away from.

Alive, reachable, real. Only after a vault clears those three do its returns even mean anything.

So what is the leaderboard, exactly?

It's one page that takes every vault worth looking at, recomputes the honest version of its numbers, and lets you read the field three ways:

  • The Altcopy Index — a mechanical, rules-based basket of the strongest investable vaults, and the page's home view. How it picks that basket is the next post in this series.
  • By investor profile — Income, Balanced, or Aggressive. One click re-ranks the whole board by a 0–100 fit score for the kind of investor you are. There's a post on reading it through your own eyes, too.
  • The full field — every living vault, sorted by the one number that can't be flattered: how much real capital is actually in it.

Every metric has a plain-English explanation on hover, because a leaderboard you can't read is just another wall of numbers. If you're wondering why I bother being this careful, it's the same reason this blog stopped trusting centralized-exchange copy-trading leaderboards in the first place.

The honest part

Two things I won't pretend about. First, this is data, not advice — it's a sharper lens on a risky corner of the market, not a promise. Second, the numbers from Hyperliquid's API are coarse at the daily level, which is exactly why I've been taking my own snapshots every single day since June — the longer this runs, the finer the history nobody else has.

If you want to see the live board — the Index composition for today, and the full ranked field behind it — it's right here.

This is the first of a short series on how the leaderboard works. Next: how the Altcopy Index picks its basket.

Subscribe to Altcopy

The honest read on Hyperliquid vaults — leaderboard moves, Index rotations, and the traps to avoid. Free.
your@email...
Abone Olun