Funding rates and copy trading on Hyperliquid: hidden drag or opportunity?

How funding works on a perpetual future

A perpetual future has no expiry date. Without an expiry, there is no settlement event to force its price back to the underlying, so the contract needs a different tether. Funding is that tether.

At regular intervals, one side of the market pays the other. When the perp trades above the index price, longs pay shorts, which makes being long marginally more expensive and being short marginally more attractive until the gap closes. When the perp trades below the index, the direction reverses.

Why funding exists and who pays whom

The critical property is that funding is a transfer between traders, not a fee paid to the venue. Nobody extracts it; it moves from the crowded side of the market to the other side.

That means funding is not automatically a cost. If you are positioned on the side receiving payment, funding is income. Whether it helps or hurts depends entirely on which side of a persistently skewed market you happen to be holding.

In practice, most copy trading followers experience funding as a cost, because most followed strategies are directional and most persistent skews occur in the direction the crowd — including the leaders — is already leaning.

Hyperliquid's funding mechanics at a practical level

Hyperliquid computes funding from the relationship between the perp's mark price and its oracle price, and applies it periodically to open positions. The rate is not fixed: it varies by market and by how far the perp has drifted from the index.

Two practical consequences follow. First, funding is charged against the notional size of an open position, not against your margin, so a leveraged position accrues funding on the full exposure rather than on the capital you posted. Second, funding accrues only while the position is open at the moment funding is applied — a position that opens and closes between intervals pays nothing.

This note deliberately quotes no rate figures. Rates move continuously and any number printed here would be wrong by the time you read it; the mechanics are what generalise.

• Funding is paid on notional exposure, not on posted margin.
• Leverage multiplies funding exposure exactly as it multiplies price exposure.
• A position closed before the interval pays nothing for that interval.

Funding when you copy one trader

When you mirror a single leader, you inherit their funding profile in full. Their market selection, their direction and their holding period become yours, and so does the funding those choices generate.

But you do not inherit it exactly, and the difference is where followers get surprised.

Holding period is the multiplier

Funding cost is approximately notional exposure multiplied by rate multiplied by the number of intervals held. Skill affects the first term. Strategy style affects the third, and the third is where the variation lives.

A scalper holding minutes at a time may pay funding on a minority of trades. A swing trader holding a directional position across days pays on every interval, on the full notional, regardless of whether the position is currently in profit.

This is why funding is invisible in the leaderboard-style evaluation most people do. A leader's realised PnL already includes their funding, so it looks handled. What it does not tell you is how much of their gross edge funding is consuming — and therefore how sensitive the strategy is if the funding regime shifts against it.

Timing divergence: you can pay an interval the leader avoided

Mirroring is never instantaneous. Your order is placed after the leader's is detected, and your fill lands at a different moment.

Most of the time that gap costs a small amount of slippage. Occasionally it costs a full funding interval: the leader closes just before funding is applied and you close just after, so they pay nothing and you pay on the whole notional.

The reverse also happens, and over many trades the two partially offset. But they do not offset symmetrically, because a system reacting to a leader is structurally late on both entry and exit — late to open means missing an interval you would have paid, late to close means catching one you would have avoided, and the closing side is the one that bites when funding is running against the position.

Funding across a score-weighted basket

Diversifying across several leaders changes the funding picture in two opposite directions at once, and it is worth being precise about both.

Same-direction leaders concentrate funding exposure

Diversification across leaders is not automatically diversification across funding exposure. If several qualified leaders are long the same major perp — which is common, because the same market conditions that make a strategy score well often push several strategies the same way — your basket holds one large directional exposure spread across several sub-accounts.

Every one of those sleeves pays funding independently, in the same direction, at the same time. From a funding perspective, the basket behaves like a single concentrated position.

The mitigating factor is that leaders with genuinely different styles hold different markets and different horizons, so the overlap is rarely total. But it is a real correlation, and it is worth monitoring rather than assuming away.

Opposing leaders in isolated sub-accounts both pay

Here the cost of isolation is explicit. With one Hyperliquid sub-account per mirrored leader, a long from one leader and a short from another in the same market coexist. Both positions are open, and both accrue funding — the long pays and the short receives, or the reverse.

In a single netted account, those two positions would have cancelled and the funding would have largely cancelled with them. Isolation deliberately gives that up.

This is a genuine trade-off rather than a flaw, and it should be stated as one. Netting would reduce funding on offsetting positions and would also destroy the per-leader attribution and containment that make the basket work — a long and a short cancelling means neither leader's thesis is expressed, and one leader's liquidation would consume margin backing all the others. The funding cost is the price of keeping each sleeve intact.

• Isolation preserves each leader's position and each leader's funding line.
• Netting would offset funding on opposing positions and cancel the trades themselves.
• Opposing exposure across leaders is usually partial and short-lived, but it is not free.

Strategy types by funding sensitivity

Funding sensitivity is a property of strategy style rather than of trader quality. The table below groups the common styles by how much funding they accumulate and in which direction it typically runs.

The pattern is consistent: exposure scales with time in position and with how closely the strategy aligns with the crowded side of the market. Trend followers riding an extended move are the classic case — the same crowding that makes the trend work also makes funding expensive for anyone holding it.

Mean-reversion strategies are the interesting inverse. Fading an extended move often means positioning against the crowd, which frequently means receiving funding rather than paying it. That is a structural tailwind, not a large one, and it does not compensate for being wrong on direction.

How to monitor funding exposure

Monitoring funding is straightforward and rarely done. Three habits cover most of the value.

First, look at aggregate direction rather than per-position. Sum the signed notional exposure across all sleeves per market. That number, not the number of open positions, determines your funding bill.

Second, watch holding period. If average time-in-position across the basket is lengthening, funding exposure is rising even if position sizes are unchanged.

Third, read funding separately from PnL. Most reporting folds funding into realised PnL, where it disappears. Tracking it as its own line is what turns funding from a mystery into a manageable cost.

• Aggregate signed notional per market across all sub-accounts.
• Track average holding period as a leading indicator of funding cost.
• Report funding as a separate line, not folded into PnL.

Practical implications for followers

Funding rarely decides whether a copy trading allocation works. It routinely decides how much of the gross edge survives to your account, particularly for slower strategies.

The practical stance is to treat funding as a variable cost that scales with time and leverage, and to be suspicious of any strategy whose edge is thin enough that a normal funding regime would erase it. A strategy that only works when funding is favourable is a funding bet wearing a directional costume.

It also argues for caring about leader style diversity rather than leader count. Ten leaders holding the same crowded direction give you one funding exposure and the illusion of diversification.

Limitations

Funding rates are not forecastable in any reliable way. They reflect current positioning, and positioning changes faster than most analysis of it.

Nothing here should be read as an argument that funding can be optimised into an edge by a follower. You do not control entry timing, market selection or holding period — the leader does. What you control is which styles you allocate to and how much leverage sits behind them.

Funding is also a second-order cost next to the first-order risk. Trading perpetual futures carries a substantial risk of loss, including the loss of your entire position, and a liquidation makes the funding line irrelevant.

Conclusion

Funding is neither hidden drag nor free opportunity. It is a transfer between traders that scales with notional and time, and it becomes material exactly when a basket concentrates in one direction and holds.

If you want the rest of the cost picture in one place, the note on the real costs of copy trading sets funding alongside exchange fees, spread, slippage and idle margin, and the isolation note explains why opposing sleeves are kept separate despite the funding cost.