Perpetual Basis Arbitrage Between CEX and DEX
Perp basis arbitrage CEX DEX is a delta-neutral strategy that exploits funding rate differences between perpetual swap markets on centralized exchanges (CEX) like Binance and decentralized exchanges (DEX) like dYdX. The core idea is straightforward: when the same perpetual contract trades with divergent funding rates across venues – one paying longs and the other paying shorts – an arbitrageur can capture that differential with minimal directional risk.
Unlike classical basis trades (spot vs futures), perp basis arbitrage uses two perpetuals with different funding dynamics. CEX funding rates are driven by order-book premium, while DEX methods vary (orderbook-based, AMM-driven, or synthetic). This creates persistent, though narrowing, opportunities for advanced traders who can manage execution, margin, and funding timing.
- Perp basis arbitrage between CEX and DEX captures funding rate differentials with delta-neutral positions.
- Success requires real-time monitoring of funding rates, low leverage (2–3x), and use of limit orders to reduce costs.
- CEXs offer deep liquidity and reliable funding; DEXs provide pseudonymity and sometimes higher differentials.
- GMX, dYdX, and Hyperliquid are the most liquid DEXs for this strategy; Binance and Bybit are top CEXs.
- Risk management is paramount: monitor rates hourly, use alerts, and never allocate more than 20% of capital.
- Fees, slippage, and margin requirements must be computed for each leg individually to ensure positive expected return.
Understanding Perpetual Basis and Funding Rates
A perpetual swap tracks an underlying index via a funding rate mechanism. The basis is the difference between the perpetual price and the index. When the basis is positive, longs pay shorts; when negative, shorts pay longs. This payment, typically every 8 hours, aligns the perpetual price with the spot index.
On CEXs like Binance, Bybit, and OKX, funding rates are calculated from a premium index (the difference between the mark price and the global index). On DEXs, the implementation varies:
- dYdX (v4) – uses a similar orderbook-based premium index, settled every hour for some markets, every 8 hours for legacy.
- Hyperliquid – orderbook DEX running on its own L1 chain, hourly funding based on premium, no KYC.
- GMX – synthetic AMM; funding rate is dynamic, paid on each position size change, not on a fixed schedule; based on long/short open interest imbalance.
- Perpetual Protocol (Curie) – vAMM with funding dependent on deviation from index.
Understanding these nuances is critical because the timing and magnitude of funding payments affect arbitrage profitability.
Why Funding Rates Diverge Between CEX and DEX
Several structural factors cause persistent funding rate differentials:
- User Demographics: CEXs have a mix of retail and institutional traders; DEXs often attract more crypto-native, yield-seeking users who may bias one side (e.g., perpetual longs on GMX due to farming incentives).
- Liquidity Depth: CEX liquidity is deeper, but DEX liquidity can be thin, leading to larger premium/discount deviations before funding catches up.
- Fee Structures: CEXs have maker/taker fees (e.g., 0.02%/0.04%); DEXs may have lower or higher fees (dYdX tiered fees; GMX has 0.1% per trade). These affect net profitability.
- Leverage Limits: CEXs allow up to 125x, DEXs typically cap at 20x (dYdX) or 50x (Hyperliquid). Differences in leverage appetite can push funding rates apart.
- Smart Contract Risk: DEXs carry extra risk that can lead to a funding premium as compensation.
For example, during a sharp altcoin rally, a CEX's ETH perp funding can spike strongly positive (longs paying shorts) while a DEX's ETH perp funding lags near zero, creating an opportunity to short the CEX perp and long the DEX perp.
Mechanics of a Delta-Neutral Perp Basis Trade
The trade is executed by opening equal notional positions in the same underlying asset: short the perp with higher funding rate and long the perp with lower (or negative) funding rate. For example:
- If Binance BTC perp funding is +0.02% (longs pay shorts) and dYdX BTC perp funding is -0.01% (shorts pay longs), you go short Binance (you receive +0.02%) and long dYdX (you receive +0.01%). Net funding received: 0.03% per 8 hours relative to a single leg's notional (equivalently ~0.015% of the combined two-leg notional).
- If both are positive but one is higher, you go short the higher (receive more) and long the lower (pay less), net profit = difference.
To be delta-neutral, the notional values must be equal. Use leverage (e.g., 2–3x) to reduce capital requirements while keeping margin levels safe. Your P&L from price movements will be near zero if both positions track the same index – but small divergences (tracking error) can occur due to different funding history or oracle slippage. The primary source of profit is the cumulative funding flow.
Example with numbers (illustrative only):
| Exchange | Position | Notional | Funding Rate (8h) | Funding Received (8h) |
|---|---|---|---|---|
| Binance | Short ETH-PERP | $50,000 | +0.02% | $10 |
| dYdX | Long ETH-PERP | $50,000 | −0.01% | $5 (short pays long) |
| $100,000 total | $15 net per 8h |
At $15 per 8h, that's $45/day, or ~0.045% daily on combined notional. Annualized ~16.4% – but rates change.
Step-by-Step Execution Guide
Follow these steps to execute a perp basis arbitrage trade:
- Monitor funding rates across CEX and DEX using tools like Coinglass, Laevitas, or custom scripts via API (Binance, dYdX, Hyperliquid). Look for persistent differences ≥0.005% per 8h after fees.
- Select the same asset (e.g., BTC, ETH, SOL) with active perpetual markets on both venues. Ensure the index is similar (e.g., both use global weighted spot index).
- Choose direction: If CEX funding > DEX funding, go short CEX, long DEX. If DEX > CEX, do the opposite.
- Calculate equal notional. For example, decide a capital allocation per leg (say $25,000). Set leverage (2x) so margin required is $12,500 per side.
- Enter positions with limit orders to reduce slippage. For DEX, use limit orders on dYdX or Hyperliquid. For CEX, use post-only to earn rebates where possible.
- Monitor funding intervals (typically every 8 hours). You can hold for several intervals if the disparity persists. Close before the funding payment if rates are about to converge, or simply collect and re-evaluate.
- Exit synchronously to avoid temporary price exposure. Use market orders if the spread is acceptable, or cancel exess and follow price.
Tools and Platforms for Finding Opportunities
Real-time funding rate aggregation is essential. Key tools:
- Coinglass Funding Tracker – shows current and historical funding rates for Binance, Bybit, OKX, dYdX, and more. Sortable by rate.
- Laevitas – advanced analytics, including funding rate heatmaps and open interest data.
- TradingView – can display funding rate as a custom indicator if you have data.
- DEX-specific explorers: dYdX stats page, Hyperliquid L1 data.
Platforms most suitable for this strategy:
| Platform | Type | Funding Model | Fees (Taker/Maker) | Max Leverage | KYC |
|---|---|---|---|---|---|
| Binance Futures | CEX | Premium index, 8h | 0.04%/0.02% | 125x | Required |
| Bybit | CEX | Premium index, 8h | 0.04%/0.02% | 100x | Required |
| dYdX v4 | DEX (orderbook) | Premium index, 1h or 8h | 0.05%/0.02% (tiered) | 20x | None |
| Hyperliquid | DEX (orderbook) | Premium index, 1h | 0.025%/0.01% | 50x | None |
| GMX | DEX (synthetic AMM) | OI-based, continuous | 0.1% per trade | 30x | None |
Note: GMX's continuous funding makes timing different; you earn from the spread as soon as you open.
Risk Management Essentials
Primary risks and mitigations:
- Funding rate change: Rates can flip within one interval due to large trades or market moves. Set alerts and be ready to close if the differential narrows below your profitability threshold (e.g., 0.003%).
- Liquidation: Even with delta-neutral positions, if one leg uses higher leverage and price moves sharply, you can get liquidated on that side. Use leverage ≤3x and maintain comfortable margin buffers (e.g., 3x safety factor).
- Tracking error: CEX and DEX may reference different index oracles, causing temporary non-zero delta. Monitor cumulative funding and exit if one side creeps.
- Liquidity risk: DEX orderbooks can be thin; entering/exiting large positions may cause slippage. Use limit orders and break trades into smaller chunks.
- Smart contract/custody risk: DEXs are not immune to hacks and contract bugs. Use platforms with audited contracts and limit exposure to a portion of capital.
A common rule is to allocate no more than 20% of your trading capital to this strategy and maintain at least 2x the required margin on each leg.
Advanced Considerations: Cross-Exchange Basis with Options or Futures
For sophisticated traders, the same concept can be extended using options, futures, or spot-perp pairs to further reduce risk or enhance yields:
- Spot-perp basis: Go long spot (or futures) on one venue and short perp on another. This captures the funding from the short perp while the spot hedges price. Requires spot liquidity and borrowing costs.
- Options-perp hedge: Use at-the-money options to hedge tail risk from extreme price moves that could cause liquidation on one leg before you can close.
- Cross-margin optimisation: If you use a portfolio margin account on CEX that allows cross-margin with futures, you can reduce collateral requirements.
- Arbitrage across multi-asset perp pairs: Sometimes funding discrepancies appear on correlated assets (e.g., ETH and BTC), allowing you to use a second pair to capture funding while hedging beta.
These strategies require deeper liquidity and more complex risk management, but can smooth returns when pure perp-perp spread narrows.
Comparing CEX vs DEX: Which Is Better for This Strategy?
| Aspect | CEX (e.g., Binance) | DEX (e.g., dYdX) |
|---|---|---|
| Funding rate precision | Good (high liquidity leads to smooth rates) | Good but can be noisier; more prone to spikes |
| Funding schedule | 8h fixed (00:00, 08:00, 16:00 UTC) | dYdX: 1h/8h depending on market; Hyperliquid: 1h |
| Fees | Low (0.04% taker) | Moderate (0.05% taker on dYdX); rebates available |
| Liquidity | Deep (best for large size) | Thinner (use limit orders; slippage risk) |
| KYC | Required | Not required (pseudonymous) |
| Custody | Exchange holds funds (risk of freeze) | Self-custody (smart contract risk) |
| Capital efficiency | Up to 125x leverage (use low) | Up to 20x (dYdX) or 50x (Hyperliquid) |
There is no clear winner; many traders use a mix: the stable funding on CEX combined with the lower trigger rates on DEX. The key is to run the math on fees and margin costs before each trade.
Alternative Strategy: Using Cross-Perpetual Spreads with One Leg on a DEX AMM
While orderbook DEXs like dYdX mimic CEX mechanics, AMM-based perps (GMX, Perpetual Protocol) have funding determined by pool imbalance. This creates a different opportunity: you can short the CEX perp and long the synthetic AMM perp when the AMM's long bias is high (paying shorts). For example, if GMX's BTC pool is heavily long, the funding rate paid to shorts might be high. Meanwhile, Binance's BTC funding may be moderate. This asymmetry can be captured by going long GMX (as shorts pay you) and short Binance (receive funding). However, AMM perps have higher fees and imperfect hedging (synthetic prices may deviate from index). They are best for smaller, nimble trades.
Step-by-step
- Monitor funding rates on Coinglass or Laevitas; identify pairs with persistent differences >0.005% per 8h after fees.
- Select the same underlying asset (e.g., BTCUSD PERP) on a CEX (Binance) and a DEX (dYdX or Hyperliquid).
- Calculate equal notional sizes for both legs (e.g., $50,000 each) with low leverage (2–3x) to minimize liquidation risk.
- Go short on the exchange with higher funding rate, go long on the exchange with lower (or negative) funding rate.
- Enter using limit orders to control slippage; on DEX use limit orders if available (dYdX, Hyperliquid).
- Hold for at least one funding interval (typically 8h); re-check funding rates before the next payment.
- Close both legs simultaneously using market or stop orders if the differential narrows below your profitability threshold (e.g., 0.003%).
- Record costs (fees, slippage) and adjust future trade size accordingly.
Common mistakes to avoid
- Overleveraging – using leverage >5x on one leg leads to liquidation during normal price swings, losing the arb.
- Ignoring fees – zero or negative net profit after taker fees and spreads on both exchanges.
- Not accounting for funding timing – holding through a funding payment when rates are about to converge, paying twice on one side.
- Assuming constant funding – rates can flip within minutes; not using alerts to exit when differential narrows.
- Neglecting tracking error – CEX and DEX may reference different indexes (e.g., .BTCUSD vs .BTCUSD) causing P&L deviation.
- Trading illiquid pairs – thin orderbooks on DEX cause huge slippage erasing profits.
Frequently asked questions
What is a typical profit per trade for perp basis arbitrage?
Profits are thin, usually 0.01% to 0.05% per 8-hour interval on combined notional after fees. Annualized returns range roughly 5% to 20% in favorable conditions, but rates are variable.
Can I do this without KYC on a CEX?
No, most regulated CEXs (Binance, Bybit, OKX) require KYC for futures trading. However, you can use other CEXs like MEXC or KuCoin that offer perps with reduced KYC, but liquidity may be lower. Alternatively stick to DEXs only.
Why not just use spot-perp basis trading instead?
Spot-perp basis requires holding spot (capital intensive) or borrowing (interest costs). Perp-perp arbitrage is more capital efficient as you only need margin on both sides, and you can use leverage. But it carries counterparty and funding timing risk.
Which DEX has the most reliable funding rates for this strategy?
dYdX and Hyperliquid both use orderbook-based premium indexes similar to CEX, with predictable hourly funding. GMX's continuous funding can be exploited but timing is different and fees are higher.
Related reading
Track the entities behind the concepts
DeFi Intel maps 11,000+ protocols, tokens and companies to a typed knowledge graph — with live data, incidents and regulation.