Hyperliquid io:OAI perps: holding rounds, cost lines and real liquidation records for 0 mark-to-market profitable addresses. As of 2026-09-22 the pool holds $2M long and $1M directionally short (plus $0M spot-hedged, which bets on nothing).
The 0 wallets below have the highest mark-to-market PnL on io:OAI, with market-making bots and high-frequency accounts removed. These are things that already happened, not a forecast of what comes next. Click a code for that wallet's full profile. None of them were liquidated in this window. Window 2026-09-02 – 2026-09-22.
| Code | Mark-to-market PnL | Rounds | Peak exposure | Liquidated | Last round ended |
|---|
The chart above overlays up to 3 wallets: candles, the whole pool's exposure on io:OAI, and each wallet's own holding rounds and cost line. Method and known limits are under “About”, top right.
Pick a round: every fill is overlaid on the chart above — buys joined by a gold line, sells by a cyan one.
The gold line joins his buy prices in order, the cyan one his sells — so you can see whether he sized in at once or scaled in, and whether he exited decisively or dragged it out.
⚠️ The x-axis is daily, so several fills on the same day land on the same position: that is why the gold line has vertical segments — he added at several prices that day. For fill-by-fill ordering, use the full unfold link.
Hyperliquid io:OAI perps: holding rounds, cost lines and real liquidation records for 0 mark-to-market profitable addresses. As of 2026-09-22 the pool holds $2M long and $1M directionally short (plus $0M spot-hedged, which bets on nothing).0 of these wallets have been liquidated. Window 2026-09-02 – 2026-09-22.
On Hyperliquid perps: the addresses that are genuinely profitable when marked to market (open positions included, plus funding, minus fees) — and when they opened and closed on this coin.
| Top | Daily candles for the coin |
| Middle | The whole pool's exposure on this coin (free tier) |
| Bottom | Holding rounds of the selected wallets, one row each |
All three share one time axis — drag or scroll to zoom and they move together, with a linked crosshair.
One coin held from first entry until fully flat. Adding to or trimming a position does not start a new round; flipping long to short counts as two. So a single round can contain thousands of fills — click a block and the detail bar gives you both numbers at once.
Some traders buy spot and short the perp in equal size. Price moves cancel out; they earn funding and are not betting on direction at all. The perp ledger only shows the short leg — without splitting it out, you would read someone's hedge as a bearish signal.
⚠️ This is still a lower bound: the spot ledger is monthly (not daily), and the spot-index → perp-ticker mapping covers about 63% by volume. Hedges we cannot identify fall into "directional short".
Three places: the coloured line in the middle panel (that wallet's own exposure, on its own right-hand scale — the pool runs at $1,000M while a single wallet is $0.3–2M, so a shared axis would flatten it against the floor); hover the candles for that day's position size (in coin), exposure (in $), average cost, and the share of his own peak; and click a ribbon block to see it in the bottom bar.
Six actions, matching the categories a trading terminal uses — but every one of them derived from our own fill-level ledger:
| ▲ filled | Open (built up from flat) |
| △ hollow | Add |
| ▽ hollow | Trim |
| ▼ filled | Close (position back to zero) |
| ◆ | Flip (long to short, or short to long) |
| ✕ | Liquidated |
The first five come straight from the tape's dir field and the change in position.
Markers are drawn below the day's low / above the day's high, never at the fill price —
drawing them at the fill price buries them in the candle body and wicks.
| Filled | That round ended profitable |
| Hollow (outline only) | That round ended in a loss |
| Green / red | Long / short |
Colour encodes direction only, fill encodes profit or loss only — wallets are told apart by row, never by colour. That way "did they enter at the same time?" is answered by lining up rows.
For each coin we take the addresses with the highest P&L on that coin's leg, after removing market makers and high-frequency accounts (low taker share with ≥50 fills a day, or ≥100 fills a day outright) — without that filter the top of the list is all market makers, which is useless to a normal reader.
① The daily candles are synthesised from hourly VWAP (open = first hour, high/low = day's extremes,
close = last hour) — not true OHLC, because the price panel only stores hourly VWAP.
Real candles are needed before this is a finished product.
② Exposure is an end-of-day snapshot; positions opened and closed within a day do not appear.
③ Spot hedging can only be judged monthly, so it is not "is he neutral right now".