Who Is Paying to Stay Long on Hyperliquid? Funding Rate Data Across 204 Pairs on Binance
You are watching a token move on Hyperliquid. The price is elevated relative to Binance. You want to know: is this a real dislocation or just HL longs active? More specifically how does this one compare to the rest of the market right now?
That question is hard to answer manually. Hyperliquid runs 230 perpetual markets. Binance runs 681. Checking the funding rate on even a handful of pairs across both venues, then normalizing for the different settlement windows, takes long enough that the opportunity has already moved.
The monitor below does it in one run.
Repository: hl-funding-arb
Why These Two Venues
Hyperliquid is an on-chain perpetuals exchange where you trade tokens on leverage directly from a wallet, with no intermediary holding your funds. Put up a fraction of the position size as collateral, go long or short, and the exchange settles everything on-chain. It runs on its own L1 blockchain (HyperBFT consensus), so the full orderbook, every trade, and every funding rate calculation is verifiable on-chain in real time.
The reason we compare against Binance is that Binance runs the largest perpetuals market in crypto by volume. It is the closest thing to a neutral benchmark — its user base is broad enough (retail, HFT, market makers, institutional hedgers) that its funding rates tend to reflect a balanced market. When Hyperliquid's funding on a pair diverges significantly from Binance's, it signals that the crowd composition on HL is skewed, not that the market as a whole has moved.
Both venues list many of the same pairs. Hyperliquid tends to list tokens earlier — often before Binance adds them — and with fewer leverage restrictions. That is what draws the on-chain retail crowd in first, and it is where the most extreme funding divergences show up.
The Problem With Cross-Venue Funding
Hyperliquid settles funding hourly. Binance settles every 8 hours. The rates are not directly comparable without normalization, and most charting tools do not show them side by side. So when HL funding on a pair is running at 90% annualized and Binance is at 5%, that 85-point spread — and the carry trade or positioning signal it represents — is functionally invisible unless you build something to surface it.
The same applies to reading crowd positioning. Funding rate is a real-time measure of who is paying whom to hold a position. A persistently elevated positive rate means longs are the consensus on that venue. But is it elevated for this token specifically, or for everything on HL right now? You cannot know without seeing all pairs simultaneously.
The Monitor
The script posts to Hyperliquid's public /info endpoint — no authentication required. Binance data comes from the USDT-margined premiumIndex endpoint. Both return the current funding rate and open interest. The script normalizes both to a 1-hour equivalent, annualizes, and computes the spread.
def fetch_hl_data():
resp = requests.post(HL_API, json={"type": "metaAndAssetCtxs"}, timeout=15)
meta, ctxs = resp.json()
results = []
for coin_meta, ctx in zip(meta["universe"], ctxs):
name = coin_meta["name"]
funding_1h = float(ctx["funding"])
mark_px = float(ctx["markPx"])
oi_usd = float(ctx["openInterest"]) * mark_px
results.append({
"symbol": name,
"funding_annual": funding_1h * 24 * 365,
"oi_usd": oi_usd,
})
return results
Annualization: rate_1h × 24 × 365 for HL, rate_8h × 3 × 365 for Binance. The spread column is HL_annual − BN_annual. Break-even days is how long a delta-neutral position — long on the lower-funding venue, short on the higher — needs to stay open to recover the round-trip taker fee.
Fee assumption: HL taker 0.035%, Binance USDT taker 0.050% — standard retail tier. Round-trip cost: 0.170%. Both venues have volume tiers; HL drops to 0.010% at higher tiers, Binance VIP 6 is 0.017%. At those rates the effective round-trip is ~0.054%, cutting break-even roughly 3×. The HL_TAKER_FEE and BN_TAKER_FEE constants in the script are adjustable for your tier.
Full source: hl_funding_arb.py
pip install requests
python hl_funding_arb.py --top 25 --min-oi 5
--min-oi sets a minimum OI filter in millions USD. Default is 0 (all markets). --top controls pairs displayed per section.
What It Shows: The Platform-Wide Pattern
The snapshot below is from 2026-05-09 03:07 UTC. The first thing the data shows is not a specific opportunity — it is a platform-wide pattern.
Across 204 perpetual pairs listed on both Hyperliquid and Binance:
- 141 pairs show HL funding higher than Binance (crowd is paying to be long on HL)
- 25 pairs show HL funding lower than Binance (short consensus on HL)
- Average absolute annualized funding across all HL crypto perps: 24.72%
That 85% directional skew is structural, not incidental. Hyperliquid's user base weights toward retail traders who want long exposure to small-cap tokens and new listings — assets they often cannot access on Binance at all, or where Binance has imposed leverage restrictions. Binance, with a more heterogeneous mix of retail, HFT, and hedgers, runs closer to neutral.
The practical implication: when you see a token moving up on HL, there is an 85% prior probability that the crowd on HL is net long and paying for it. The monitor tells you by how much, and whether that positioning is more extreme than the rest of the market right now.
The Spread Table
Here is where the discoverability starts to matter. These are the pairs with the largest spreads — the tokens where HL crowd positioning is most divorced from Binance.
| Symbol | HL Annual | BN Annual | Spread | OI | B/E Days | Direction |
|---|---|---|---|---|---|---|
| VVV | +556.68% | +34.79% | +521.89% | $30.8M | 0.1 | BN short / HL long |
| STBL | +254.09% | +5.48% | +248.62% | $4.3M | 0.2 | BN short / HL long |
| SAGA | +200.81% | +5.48% | +195.33% | $1.9M | 0.3 | BN short / HL long |
| XPL | +123.72% | +5.48% | +118.25% | $31.3M | 0.5 | BN short / HL long |
| ATOM | -103.67% | +10.95% | -114.62% | $3.6M | 0.5 | HL short / BN long |
| FET | +94.57% | -8.64% | +103.21% | $5.4M | 0.6 | BN short / HL long |
| CHIP | +108.46% | +5.48% | +102.98% | $16.8M | 0.6 | BN short / HL long |
| ENA | +97.88% | +5.48% | +92.41% | $42.3M | 0.7 | BN short / HL long |
| MON | +97.51% | +5.48% | +92.03% | $44.5M | 0.7 | BN short / HL long |
| MEGA | -171.98% | -84.80% | -87.18% | $12.9M | 0.7 | HL short / BN long |
| PENGU | +85.88% | -5.89% | +91.76% | $17.5M | 0.7 | BN short / HL long |
| ALGO | +95.17% | +7.56% | +87.61% | $6.3M | 0.7 | BN short / HL long |
| STRK | +92.73% | +5.48% | +87.26% | $15.0M | 0.7 | BN short / HL long |
| XMR | +97.90% | +11.39% | +86.51% | $37.5M | 0.7 | BN short / HL long |
| NEAR | +51.79% | -9.47% | +61.26% | $57.9M | 1.0 | BN short / HL long |
| BIO | +65.73% | +5.48% | +60.26% | $9.4M | 1.0 | BN short / HL long |
| ICP | +59.37% | +3.71% | +55.66% | $5.6M | 1.1 | BN short / HL long |
NEAR at $57.9M OI is the largest liquid market in this table. A 61% annualized spread with 1-day break-even. ENA and MON both above $40M OI with spreads over 90%. These are not thin markets.
The pairs going the other direction — ATOM and MEGA — are worth separate attention.
ATOM at -103% on HL vs +10% on Binance. A 114-point spread where the HL crowd is heavily short. Whoever holds that short is paying 103% per year in carry. That kind of cost implies a strong directional conviction.
MEGA at -171% on HL vs -84% on Binance. Both venues are short-biased, but HL is nearly twice as extreme. The short positioning is concentrated here specifically. Crowded shorts at these carry levels are worth tracking — not because the thesis is wrong, but because the cost of being wrong is compounding fast.
The VVV Case
VVV sits at the extreme: +556.68% annualized on HL against +34.79% on Binance. A 521-point spread. Break-even in 0.1 days.
At that level, a delta-neutral $100K position covers the round-trip fee in under 3 hours. The annualized carry is $521K per $100K deployed.
Why does the spread persist at all? Execution: running this trade requires margin accounts on both venues, real-time monitoring, and clean entry and exit on both legs simultaneously. Most retail participants are not doing that. Liquidation risk during entry — if the price moves sharply while you are filling both sides — is real and not priced into the carry math. And extreme funding on new listings tends to compress fast once the short side builds or the crowd gets squeezed. The window is real but short.
VVV's 556% funding is a new-listing dynamic: a concentrated long retail crowd with insufficient short interest to balance it yet. It shows up repeatedly near the top of the spread table in different tokens as listings roll through.
HIP-3 Equity Perps: What the Data Actually Shows
Hyperliquid launched HIP-3 to enable tokenized equity perpetuals — stock-price-referenced perps for TSLA, AAPL, NVDA, and others. The narrative was that this opens 24/7 equity trading on-chain at scale.
The monitor checks 40 equity tickers against HL's live universe. At the time of this snapshot:
| Symbol | 1h Rate | Annual | OI | Signal |
|---|---|---|---|---|
| SPX | +0.0013% | +10.95% | $7.4M | neutral |
One ticker. $7.4M OI. 0.1% of total Hyperliquid OI.
The individual stock perps — TSLA, AAPL, NVDA, MSFT, COIN, MSTR — did not appear in the live market data. Either they are not yet active, listed under different names, or sitting at zero OI.
SPX funding at 10.95% annualized is below the 24.72% crypto average. The positioning is modestly long but not aggressive. At $7.4M against a $6.94B platform, equity perps are not yet a market in any functional sense — the infrastructure exists, the adoption has not followed.
That is useful to know. The discoverability problem cuts both ways: the monitor tells you where the crowd is actually concentrated, not where the narrative says it should be.
The Full Snapshot
| Metric | Value |
|---|---|
| Total HL markets | 230 |
| Total HL OI | $6.94B |
| Crypto perp OI | $6.93B (99.9%) |
| HIP-3 equity perp OI | $0.01B (0.1%) |
| Cross-venue pairs (HL + BN) | 204 |
| HL-only crypto pairs | 25 |
| Equity perps detected | 1 (SPX) |
| Pairs: +spread > fees | 141 |
| Pairs: −spread > fees | 25 |
| Avg annual |funding| — crypto | 24.72% |
| Avg annual |funding| — equity | 10.95% |
Data: Hyperliquid public API + Binance USDT perpetuals API. Snapshot: 2026-05-09 03:07 UTC. Run the script for current values — funding rates change continuously.