Risks and Costs of Funding Rate Arbitrage
This material is for educational purposes only and does not constitute financial advice, an investment recommendation, or a guide to action. Trading cryptocurrency derivatives carries a high risk of loss. All decisions are made by you independently and at your own risk.
Funding rate arbitrage looks like a risk-free strategy — long on one exchange, short on another, and you earn the rate difference. In practice, there are several factors that affect the real outcome.
Funding Rate — The Main Source of Income
The strategy's core profit is the rate difference between two exchanges. The rate is not fixed: it changes dynamically depending on the market and can narrow or disappear at any moment.
The system shows the last received value — accurate to within the update delay of the specific exchange.
Rate Update Delays by Exchange
| Exchange | Funding rate delay | Accrual interval |
|---|---|---|
| AsterDex | 3 sec | 1h / 2h / 4h / 8h (depends on instrument) |
| Lighter | ~10 sec | 1 hour |
| Extended | ~10 sec | 1 hour |
| Nado | ~20 sec | 1 hour |
| Omni | 15–30 sec | 1h / 2h / 4h / 8h (depends on instrument) |
| EdgeX | ~30 sec | 4 hours |
| Hyperliquid | 30–60 sec | 1 hour |
| Paradex | ~40–45 sec | continuous (every second) |
On Paradex, data arrives only when the rate has changed. If it is stable — the value is current. If the rate just changed — the system may learn about it instantly (if the change was in the nearest batch) or with a delay of up to 40–45 seconds (if the batch just went out).
How Much Accrues at Different Rates
Position of $10,000 per side ($20,000 total deployed). Accrual happens on one side — the one where the rate is in your favor (short receives at a positive rate).
Income = $10,000 × spread APR / periods per year
| Spread APR | Per 1 hour | Per 2 hours | Per 4 hours | Per 8 hours | Per day | Per week | Per month |
|---|---|---|---|---|---|---|---|
| 10% | $0.11 | $0.23 | $0.46 | $0.91 | $2.74 | $19.18 | $82.19 |
| 50% | $0.57 | $1.14 | $2.28 | $4.57 | $13.70 | $95.89 | $410.96 |
| 100% | $1.14 | $2.28 | $4.57 | $9.13 | $27.40 | $191.78 | $821.92 |
| 200% | $2.28 | $4.57 | $9.13 | $18.26 | $54.79 | $383.56 | $1,643.84 |
The table shows theoretical income assuming the rate remains unchanged for the entire period. In practice, the rate changes with each accrual.
Entry and Exit Costs
In addition to funding income, each position opening and closing incurs costs: exchange fees, bid/ask spread, and possible slippage.
Trading Fees
Position $10,000 per side. Four operations: long entry, short entry, long exit, short exit.
| Exchange | Taker | Cost for $10,000 (4 operations) |
|---|---|---|
| Paradex | 0% | $0 |
| Lighter | 0% | $0 |
| Omni | 0% | $0 |
| Extended | 0.025% | $10 |
| AsterDex | 0.04% | $16 |
| Hyperliquid | 0.045% | $18 |
Example: you open a Paradex + Hyperliquid pair. Paradex fee = $0, Hyperliquid = $18. Total $18 needs to be recovered from funding rate before turning profitable.
Bid/Ask Spread
The spread is the difference between the buy price (ask) and sell price (bid). When entering, you always buy at ask and sell at bid — meaning you pay half the spread on each operation.
With a 0.1% spread on each exchange and a position of $10,000:
| Operation | Cost |
|---|---|
| Entry (2 exchanges) | $10,000 × 0.1% × 2 = $20 |
| Exit (2 exchanges) | $10,000 × 0.1% × 2 = $20 |
| Total | $40 |
Adding slippage of 0.05% per exchange (order consumed several order book levels):
| Cost | Amount |
|---|---|
| Spread (entry + exit) | $40 |
| Slippage (entry + exit) | $10,000 × 0.05% × 2 × 2 = $20 |
| Total | $60 |
At a spread of 100% APR, the first hour accrues ~$1.14. To recover just the entry/exit costs ($60), the position must run for ~53 hours (~2 days). At 50% APR — already ~4 days.
Impact of Price Movement on Results
Funding rate arbitrage is a delta-neutral strategy: the long on one exchange offsets the short on another. But the exit execution price depends on the bid/ask at that moment — and if prices on two exchanges have diverged, the result will differ.
Scenario A — prices unchanged (ideal)
| Exchange 1 (long) | Exchange 2 (short) | |
|---|---|---|
| Entry | ask $100 | bid $100 |
| Exit | bid $100 | ask $100 |
| Price PnL | $0 | $0 |
All income is from funding rate only. Costs are only spread and fees.
Scenario B — prices diverged, then converged
Entry: Exchange 1 bid/ask = $99/$100, Exchange 2 bid/ask = $102/$103. At exit: Exchange 2 moved to $100/$101.
| Exchange 1 (long) | Exchange 2 (short) | |
|---|---|---|
| Entry | ask $100 | bid $102 |
| Exit | bid $100 | ask $101 |
| Price PnL | $0 | −$1 per $10,000 = −$100 |
The short on Exchange 2 was closed at a higher price than it was opened — a $100 loss from price movement alone, regardless of funding.
This is not a strategy bug, but market risk: if prices on two exchanges diverged and did not converge by the time of exit — that is a loss. The longer the position is open, the higher the probability that prices will equalize — but there is no guarantee they will return to an acceptable level at the right time.
Funding Rate Scenarios
Scenario 1 — both exchanges pay you ✅
Paradex: −100% APR (negative rate → shorts receive). Hyperliquid: +50% APR (positive rate → shorts receive).
You opened: short on Paradex, long on Hyperliquid.
| Exchange | Your position | Rate | Result |
|---|---|---|---|
| Paradex | short | −100% APR | receive |
| Hyperliquid | long | +50% APR | pay |
| Total spread | 50% APR | receive |
Per day at $10,000: +$13.70. Both exchanges work in your favor through different mechanics.
Scenario 2 — you pay on one exchange ⚠️
At the next accrual, the Hyperliquid rate became negative: −20% APR.
| Exchange | Your position | Rate | Result |
|---|---|---|---|
| Paradex | short | −100% APR | receive |
| Hyperliquid | long | −20% APR | receive |
| Total spread | 120% APR | receive more |
In this case it got better. But the reverse can also happen:
Scenario 3 — spread collapsed ❌
Paradex: −100% APR → became +20% APR. Hyperliquid: +50% APR → remained +50% APR.
| Exchange | Your position | Rate | Result |
|---|---|---|---|
| Paradex | short | +20% APR | pay |
| Hyperliquid | long | +50% APR | pay |
| Total | −70% APR | loss |
Both legs became unprofitable. The position needs to be closed — otherwise funding consumes capital.
The rate is fixed at the moment of each accrual, not at the moment of entry. The spread at opening is only a starting point. Monitoring current rates and closing the position in time when the market changes is a mandatory part of the strategy.
The Rate Is Fixed at Accrual Time, Not at Entry
This is a key risk that is easy to miss.
You open a position when you see:
- Paradex (short): −200% APR → short receives
- Hyperliquid (long): −100% APR → long pays
- Spread: 100% APR — looks great
But accrual happens not at the moment of opening, but at the end of the current hourly period. During this time, the market rate recalculates. By the time of the first accrual, rates may have changed — for example, the spread may narrow to 10% or go to zero altogether.
You will earn exactly as much as the spread was at the moment of each specific accrual — not at the moment of entry.
A high spread at opening is a signal of opportunity, not a guarantee of income. The longer you hold the position, the more accruals happen at current (changed) rates.
How to Manage Rate Change Risk
A delay of tens of seconds is not a critical problem in itself. The critical problem is concentrating all volume in a single token.
The right approach: trade a portfolio of tokens simultaneously. If the rate on one instrument changed sharply — the other positions continue generating income. Diversification across instruments and exchanges reduces dependence on a single event.
Bid/Ask Spread and Slippage
When opening and closing a position, you pay the spread twice — at entry and at exit. The system shows only the best price (best bid/ask) — without volume at the level.
Bid/Ask Delay by Exchange
| Exchange | Bid/ask delay |
|---|---|
| Lighter | ~50 ms (WebSocket) |
| Extended | real-time (WebSocket) |
| Nado | real-time (WebSocket) |
| AsterDex | ~10 sec (REST) |
| Omni | 15–30 sec (REST) |
| EdgeX | ~30 sec (REST) |
| Paradex | ~40–45 sec (WebSocket batch) |
| Hyperliquid | 30–60 sec (REST) |
On Omni, bid/ask are calculated market-maker quotes at fixed sizes ($1K, $100K, $1M), not a real order book. Actual execution may differ.
Slippage on Large Positions
Real slippage is determined by order book depth — how much volume sits at the best levels. The larger the position, the higher the probability of execution across multiple levels.
This is why the right strategy is to distribute capital across multiple tokens, not open one large position. A smaller size per instrument means less slippage per entry.
Trading Fees
A fee is charged on opening and closing each of the two legs of a position.
| Exchange | Maker | Taker |
|---|---|---|
| Paradex | 0% | 0% |
| Lighter | 0% | 0% |
| Omni | 0% | 0% |
| Extended | 0% | 0.025% |
| Hyperliquid | 0% | 0.045% |
| AsterDex | 0.005% | 0.04% |
| EdgeX | to clarify | to clarify |
| Nado | to clarify | to clarify |
For market orders, the taker rate applies. On a $10,000 position at Hyperliquid (0.045%), the fee is $4.50 at entry and $4.50 at exit — total $9 per leg. Both legs = $18.
Minimum spread for breakeven depends on fees and holding period:
Minimum APR ≈ total fee (both legs, entry + exit) / holding period (in years)
Example: total fee 0.18%, holding 30 days:
0.18% / (30/365) ≈ 2.2% APR
A spread below this threshold means a loss after fees.
Additional Portfolio Risks
Beyond rate changes, when working with multiple exchanges simultaneously, it is important to consider:
- Exchange hack or technical failure — funds on one exchange may become inaccessible while the position on the second remains open
- Loss of liquidity — in a sharp volume drop, closing a position may cost more than calculated
- Liquidation when using leverage — a sharp price move may liquidate one leg while the other remains open, turning the hedge into a directional position
Distributing across multiple exchanges and tokens reduces the concentration of each of these risks.
Points Accrual
Several exchanges award points for trading volume — they convert to tokens at listing or are distributed as a retrodrop.
| Exchange | Program |
|---|---|
| Paradex | Paradex Points for trading volume |
| Hyperliquid | HYPE token (retrodrop already happened) |
| Lighter | Lighter Points for volume |
| Extended | XTN Points for volume and liquidity |
| Nado | Nado Points for volume |
| EdgeX | EX Points for activity |
| AsterDex | ASTER token program |
| Omni | check on the exchange website |
Points are additional income on top of the funding rate. All else being equal, exchanges with an active points program are preferable. Check current terms on each exchange's website — programs change.
Order Execution Risks
Limit Orders: Two Legs Don't Execute Simultaneously
When opening an arbitrage position, two legs need to be executed — a long on one exchange and a short on another. Using limit orders, they are placed sequentially or in parallel, but execute independently. This creates several risks.
Risk 1 — partial fill
A limit order for $10,000 may only take $6,000 from the order book — if there wasn't enough liquidity at the needed level. As a result, one leg is open at $10,000, the second at $6,000. The position is unbalanced: $4,000 is an open directional position without a hedge.
Risk 2 — one leg didn't fill at all
Price moved away from the limit before the order hit the order book. The first leg is open, the second is not. You hold a naked long or short at $10,000 with no insurance on the second exchange. Funding accrues from only one side — and not in your favor.
Risk 3 — funding changed between first and second leg execution
While the first leg was being executed and you were placing the second — time passed. In seconds on a volatile market, the rate may have changed. You entered at an 80% APR spread, but by the time the second leg executed, the spread became 15%.
The longer the gap between first and second leg execution — the higher the risk of entering at a stale spread or ending up with an unbalanced position.
Closing a Position: Risks Are Critically Higher
An unbalanced leg at opening is a nuisance. At closing, it can be a disaster.
Scenario: one leg didn't close
You are closing the position with limit orders. The short closed, the long did not (price moved above the limit). Now you have:
- Short closed
- Long $10,000 open without a hedge
If price continues rising — the loss grows every second. Funding is now also from only one side.
Scenario: partial close
The long closed at $7,000 out of $10,000. $3,000 of long remains without a short. A small directional move against you produces real losses.
Scenario: market moves while you close
You were closing both legs in sequence. Closed the first at $100, while closing the second — price moved to $103. The final exit price is different on the two exchanges, the hedge is not ideal.
Closing a position with limit orders requires active monitoring. If an order stalls — a decision must be made quickly: raise the limit, replace the order, or close at market. An unclosed leg with the second closed is an open directional position, not arbitrage.
Market Orders at Closing
Closing with a market order guarantees execution, but you pay the current spread and possible slippage. With sufficient liquidity — this is preferable to a stalled limit at a critical moment.
| Order type | Execution price | Guaranteed fill | When to use |
|---|---|---|---|
| Limit | Controlled | No | Entry in a calm market |
| Market | Current ask/bid | Yes | Closing, urgent exit |
Funding Rate Reversal
One of the most common scenarios in arbitrage: you opened a position at a good spread, and the rate reversed.
What Happens at Reversal
Entered at:
- Exchange A (short): +150% APR → short receives
- Exchange B (long): +20% APR → long pays
- Spread: 130% APR ✅
Several hours later the rate on Exchange A fell and went negative:
- Exchange A (short): −30% APR → short pays
- Exchange B (long): +20% APR → long pays
- Total: −50% APR ❌ — both legs against you
Funding is now being deducted from the position every hour. Each accrual is a loss.
Wait or Close
There is no universal rule — the decision depends on how quickly and deeply the rate reversed, and what the history says.
When to wait:
- The reversal is small (−10–20% APR) and the rate is historically unstable on this instrument — high probability it will return
- Closing costs (spread + fee) are comparable to losses from several hours of negative funding
- The position is small relative to the portfolio — risk is manageable
When to close immediately:
- The rate reversed sharply and continues going negative
- Negative funding per day exceeds closing costs
- No understanding of why it reversed — better to exit and analyze
How to Use Rate History
The system has funding rate history available for each instrument. Before making a decision, it is useful to look at:
- How long a high rate typically holds on this exchange for the given token
- Whether similar reversals happened before — and whether the rate returned
- Average rate level over the last few days — this is a reference for what is normal
History does not guarantee recurrence. If the rate reversed 9 times out of 10 — that is an argument for waiting, but not a guarantee. The decision is always yours.
Calculating the Breakeven Point at Reversal
Before deciding — calculate how long you can hold the unprofitable position before losses from negative funding exceed closing costs.
Hours until close = Closing costs ($) / (|Negative APR| / 8760 × position size ($))
Example: closing costs $40, negative spread −50% APR, position $10,000:
Loss per hour = 50% / 8760 × $10,000 = $0.57/hour
Hours until close = $40 / $0.57 ≈ 70 hours (~3 days)
If the rate does not return within 3 days — closing was more beneficial than waiting.
All examples and calculations in this section are for educational purposes only. Real results depend on many factors: exchange liquidity, market volatility, order execution, and other conditions. Nothing here constitutes financial advice. Trade consciously and only with funds you can afford to lose.