How to Capture Funding Rates on Hyperliquid Perps Profitably

Funding-rate capture on perpetual swaps is one of the few genuinely repeatable edges available in on-chain derivatives — not because it's a secret, but because most participants either ignore the carry or actively push it in your favor. Hyperliquid's architecture makes running this strategy in production materially cleaner than on centralised venues: one API, sub-second fills, and no withdrawal friction between your hedge legs. Here is exactly how we build and operate these positions.

How Hyperliquid's Funding Mechanism Works

Hyperliquid settles funding every hour on the hour. The rate is computed as the premium of the mark price over the index (a volume-weighted composite of spot oracles) divided by a 24-hour normalisation window, clamped to ±0.05% per hour. In practice the clamp rarely bites except during violent trend episodes.

The key structural detail: funding is paid continuously and accrues to open interest, not to settled positions. If you open a short five minutes before the hourly mark and close it five minutes after, you capture roughly 10 minutes of that hour's rate — not the full hour. Strategies that try to snipe individual settlements typically pay more in slippage than they collect. Run the position continuously and let the carry compound.

Pair Selection: Which Markets Pay the Most

Not all perps are equal. Across the twelve months ending Q1 2026, the following markets consistently showed positive average funding on the short side (meaning longs were paying shorts):

• SOL-PERP — average 8-hour rate ~0.008%, spikes to 0.05% during momentum runs
• WIF-PERP and other meme-cap tokens — high variance but median positive; short funding paid ~60% of hours
• BTC-PERP — lower average (~0.004% per 8h equivalent) but the tightest spreads, best for large notional
• HYPE-PERP — structurally elevated since the token launch, though liquidity depth is shallower

The pattern is consistent with perpetual perp markets: retail speculation is net long, so longs overpay to maintain exposure. Filter pairs by: (a) 30-day median funding rate > 0.005% per hour, (b) average hourly volume > $5M, (c) bid-ask spread < 0.05% at $50k depth. Running on anything outside those bounds erodes carry faster than it accrues.

Constructing the Delta-Neutral Leg

The textbook structure is short the perp, long the equivalent notional in spot. On Hyperliquid this means your short perp position lives on the exchange while your spot hedge lives either in a Solana wallet (for SOL, JLP, etc.) or on a spot venue with a reliable oracle feed.

A few things that matter in production:

• Sizing the hedge precisely. Perp notional is marked-to-market continuously; spot is not. Delta drifts as price moves. If SOL moves 5%, a $100k position is $5k offsides in delta before you rebalance. We rebalance when delta exceeds ±2% of gross notional, which typically means 1-4 rebalances per day on a volatile pair.
• Margin buffers. The perp leg needs enough cross-margin headroom to survive a 20-30% adverse move without liquidation. Targeting 4-5x initial leverage on the perp gives adequate room while keeping capital efficiency reasonable.
• Spot custody. Holding raw spot on-chain incurs no counterparty risk but adds execution latency and gas overhead for rebalancing. Using a venue with instant settlement (like Hyperliquid's own spot layer for supported pairs) closes the loop significantly.

Our trading bot infrastructure handles delta monitoring, rebalance execution, and risk circuit-breakers as a unified loop rather than separate scripts — that matters more than it sounds when rebalances need to fire in under 500ms.

Threshold Filters and When Not to Run

Funding rate capture is not always worth running. The carry needs to clear three hurdles before a position is worth opening:

1. Net carry after fees > 0.01% per hour. Taker fees on Hyperliquid are 2.5 bps for non-VIP tiers. If you're rebalancing with taker orders four times a day on a $100k position, that's ~$10/day in rebalance costs alone. At 0.008% hourly funding on $100k, gross carry is $8/hour or $192/day — the math works. At 0.003%, it does not.
2. Funding rate has been positive for at least 6 of the last 12 hours. A single-hour spike is noise; structural carry requires persistence. We use an exponential weighted moving average of the last 12 hourly rates with a decay factor of 0.85.
3. OI imbalance on the same side as the carry. If open interest is 70%+ long and funding is negative (longs paying), that's a structurally supported trade. If funding flipped because of a large short attack that will likely unwind, you're picking up pennies in front of a steamroller.

When conditions flip — funding goes negative for the short, or OI balance reverses — positions unwind within one rebalance cycle, not gradually.

Real Costs and Realistic Returns

Annualised gross yield from a well-selected portfolio of 3-5 pairs on Hyperliquid has ranged from 18% to 55% in observed periods. After subtracting maker/taker fees, rebalancing slippage (~1.5 bps per rebalance round trip), and the occasional funding reversal drawdown (budget 5-10 days of flat-to-negative carry per quarter), net annualised returns in the 15-40% range are realistic on $50k-$500k notional. Beyond $1M per pair, market impact on rebalances starts to eat the edge.

The strategy is not risk-free. Tail risks include: a violent unwind where spot and perp diverge faster than rebalancing can track, oracle manipulation on smaller pairs, and exchange-level smart contract or liquidity risk. Position sizing should reflect these, not the average-day carry math.

---

If you want this running autonomously with proper risk controls rather than managing it manually, talk to us — we build and operate these systems for clients already trading on Hyperliquid.