Hyperliquid Spot-Perp Funding Loop Using On-Chain HYPE Spot

The cleanest funding capture on Hyperliquid doesn't route through Binance at all. You buy HYPE spot on Hyperliquid's own on-chain order book, short the HYPE perp on the same exchange, and collect funding while both legs sit inside one margin account and one settlement layer. That last part changes almost everything about how the trade behaves — no bridge risk between your spot and hedge, no separate CEX withdrawal to babysit, and margin math that treats the two legs very differently than a cross-venue cash-and-carry does.

If you've read the cross-exchange cash-and-carry breakdown mental model, this is the same delta-neutral idea with the seams removed. It also introduces its own gotchas that will quietly eat your yield if you don't design for them.

Why the native spot leg matters

In a classic carry you hold spot on Venue A and short a perp on Venue B. Your capital is split, your PnL is split, and a funding-driven margin call on the short can force a liquidation while your fully-paid spot sits stranded on the other venue, unable to help. You end up over-collateralizing the short just to survive variance.

On Hyperliquid, spot and perps live in the same account under one clearinghouse. HYPE bought on the spot book counts toward your account value. When the perp short loses money because price ripped up, the spot HYPE it's hedging gained the same amount, and the exchange's cross-margin accounting sees both. Net account equity barely moves. That's the structural win: you're not funding two independent margin buffers, you're funding one net-flat position.

The tradeoff is concentration. Both legs, your collateral, and your yield are all exposed to one exchange's solvency and one order book's depth. There is no venue diversification left. If you're sizing this beyond hobby capital, an independent review of the execution and risk logic is money well spent before you route real size through it.

The mechanics, leg by leg

The position is:

• Long: N units of HYPE on the spot book (@HYPE or the spot asset index), fully paid.
• Short: N units of HYPE perp, margined against the same account.

Funding on Hyperliquid pays hourly, not every 8 hours like most CEXs. That's 24 settlements a day. When funding is positive — longs pay shorts, the usual state for a token in an uptrend — your short collects every hour. Annualized, a steady +0.01%/hr is roughly 87% APR before costs, which is why these loops exist and why they compress the moment everyone piles in.

Your realized yield is:

net_apr = funding_apr
        - 2 * taker_or_maker_fees        // entry + exit, both legs
        - spot_perp_basis_drift          // spot and perp don't track 1:1
        - rebalance_slippage             // from re-hedging drift

The basis term is the one people forget. Hyperliquid's perp mark is an oracle-anchored index, while your spot fill is whatever the on-chain book gave you. Entering, you might pay a few bps of premium on the perp relative to your spot cost. That basis is a real, one-time drag you eat on entry and recover (or lose more of) on exit.

Sizing the hedge so it actually stays neutral

Naive delta-neutral means equal units: long 100 HYPE spot, short 100 HYPE perp. But the perp short requires margin, and as funding accrues and price moves, your effective delta drifts. Two things break neutrality:

1. Margin haircut on spot. Spot HYPE used as collateral gets a weight below 1.0. You can't assume every dollar of spot backs a dollar of perp notional.
2. Funding and fees change unit counts. Funding settles in USDC to your perp balance; it doesn't change your HYPE unit count, but it does change your account equity and therefore your liquidation buffer.

A worked example. Say HYPE is at $40 and you have $10,000.

• Buy 150 HYPE spot ≈ $6,000.
• Post the remaining $4,000 as perp margin, short 150 HYPE perp ($6,000 notional). That's 1.5x on the short leg against its own margin, but roughly delta-flat at the account level because the spot covers the short's directional loss.
• Net delta ≈ 0. If HYPE goes to $50, the perp short is down $1,500 and the spot is up $1,500. Account equity holds; funding keeps dripping.

The liquidation price on the short is what you actually monitor. With cross-margin and the spot counted as equity, your effective liquidation on HYPE is far away — but "far" is not "never." A vertical move plus a funding flip plus thinning spot depth can converge faster than a backtest suggests. Keep the perp leverage low; the yield comes from funding, not from leverage.

The re-hedge loop

Delta drifts because the spot and perp legs revalue slightly differently and because you may want to compound funding back into position size. A tight bot loop looks like this:

target_delta = 0.0
band = 0.02  # re-hedge only outside +/-2% of notional

while running:
    spot_qty = get_spot_balance("HYPE")
    perp_pos = get_perp_position("HYPE")   # negative = short
    net_delta = spot_qty + perp_pos.size    # signed

    if abs(net_delta) > band * spot_qty:
        # adjust the perp leg, not spot -- perp fills are cheaper
        adjust_perp_short(target=-spot_qty)

    if funding_rate("HYPE") < min_funding_apr:
        unwind_both_legs()   # yield gone; don't hold for free
    sleep(POLL_SECONDS)

Two opinions baked in there. First, re-hedge on the perp leg, not the spot leg. Perp taker fees are lower and fills are deeper than the on-chain HYPE book, which can be thin at size. Second, the exit trigger is funding, not price. The whole reason to hold is the funding rate; when it decays below your fee-and-slippage breakeven, the trade is over even if nothing looks wrong. A dashboard that surfaces live funding, net delta, and the liquidation buffer in one view is worth building early — this is the same discipline behind a proper strategy monitoring dashboard. The DLMM auto-rebalance crowd learned the same lesson about band-based re-hedging in Meteora fee farming; the trigger logic transfers directly.

Where it bites you

• On-chain spot depth. The HYPE spot book is shallower than the perp. Entering or exiting 200+ HYPE at once walks the book. Split spot fills; be patient on the leg that has less liquidity, and consider passive maker orders to earn the rebate instead of paying the spread. The JIT liquidity people obsess over exactly this kind of book-depth timing in just-in-time LP on Meteora.
• Funding flips negative. In a hard sell-off, funding can invert and now your short pays. Your exit condition must fire on that, not wait for a scheduled rebalance.
• Spot leg can't be shorted to unwind fast. If you need to flatten in a hurry, you sell spot into a thin book while covering the perp. The perp covers instantly; the spot doesn't. Model that asymmetry.
• Fee tier reality. At retail taker rates, four fills (two legs, in and out) can cost 15–25 bps round trip. Against an 80% funding APR held for weeks that's noise; against a two-day scalp it's most of your edge.

The native-venue version of this loop is genuinely cleaner than the cross-exchange one, but "cleaner" means the failure modes are concentrated, not absent. Get the hedge sizing, the re-hedge band, and the funding exit right and it runs quietly for weeks; get the liquidation buffer wrong and one candle undoes a month of drips.

If you'd rather have the loop, the Hyperliquid perps hedging bot built and battle-tested before you fund it, that's the sensible place to start — or talk through whether the funding regime even justifies it in a strategy consultation.