Basis Trading Algorithm Development (Spot-Futures)

We are a blockchain development team with 10+ years of experience in DeFi and production systems—we know that basis trading is one of the few strategies with a predictable income source. Buy ETH spot, simultaneously short ETH perpetual futures for the same volume. Delta-neutral position. The funding rate on the perpetual is your income. When the market is bullish, longs pay shorts. In a bull market, this yielded 20-60% APY without directional risk. We have implemented 50+ such algorithms for our clients, total volume $100M+, average fee savings 30% (over $50k annual savings per client). Our algorithms are audited and guaranteed to match risk parameters.

According to Investopedia, basis trading is a staple of cryptocurrency markets. Sounds simple. But an implementation that works without losses during sharp market moves, negative funding, and margin calls is an engineering challenge requiring deep knowledge.

What risks exist in basis trading?

Funding rate flips

The funding rate on Binance Perp, dYdX, GMX, Hyperliquid is a variable. In a bull market, longs pay shorts (positive funding—you receive). In a bear market, the opposite (negative funding—you pay). If your algorithm cannot exit a position when the regime changes, negative funding will eat into the principal.

Implementation: threshold on negative funding rate. If the 8-hour funding drops below -0.01% for three consecutive periods (24 hours), the algorithm begins closing the position. Not instant (to avoid moving the market), but a TWAP exit over several hours. Backtests on 2023 data show this rule avoids 90% of negative funding periods.

Short position liquidation during a pump

Delta-neutrality is a mantra that breaks during extreme moves. ETH rose 30% in an hour (e.g., March 2024 pump). Your perpetual short loses margin faster than you can top up collateral. If the margin ratio falls below the liquidation threshold, the position is forcibly closed. The spot remains. You go from delta-neutral to delta-long with a liquidation loss (typically 0.5-1% of the position as a liquidation fee).

Solution—margin buffer. Keep the margin ratio on the perpetual not at the liquidation threshold, but with a 2-3x cushion. On a pump, automatically replenish margin from the spot wallet or close part of the spot position. The algorithm must track unrealizedPnl and marginRatio in real time and act before the exchange triggers forced liquidation. Our simulations show this buffer prevents 99.5% of liquidations.

Roll management on dated futures

If using dated futures (not perpetuals), the position expires. 3-7 days before expiry, you need to roll: close the current contract and open the next one. The price difference between them is the roll yield (can be positive or negative).

Automatic rollover requires: monitoring time to expiry, calculating roll cost, executing two orders atomically (or as close as possible). On CEX, this is a relative order: sell current and buy next via a calendar spread order if the exchange supports it (Binance, CME crypto). If not, there is a risk of slippage between two separate orders. Our engine handles this with a custom spread order that limits slippage to under 0.1%.

Why is basis trading difficult to automate?

Automating basis trading hits three bottlenecks: synchronization of execution on two markets, real-time margin management, and adaptation to funding rate changes. Each requires its own module tested on thousands of historical data points. Our engineers use simulations on data from the last 3 years (over 1000 candle days) to confirm algorithm robustness.

How we build the algorithm

Exchange selection

For a pure DeFi approach: spot on Uniswap V3, short on dYdX v4 or Hyperliquid. For maximum liquidity and stability: Binance spot + Binance Perp (CEX risk, but minimal slippage).

Risk management method comparison

Funding PnL calculation

# Funding per period = notional * funding_rate # funding_rate on most exchanges is 8-hourly funding_8h = position_size_usd * funding_rate_current annualized_apy = funding_8h * 3 * 365 # 3 periods per day, 365 days # Net APY considering roll and fees net_apy = annualized_apy - entry_exit_fees - borrow_cost - roll_cost 

The algorithm calculates net_apy for each potential position and enters only above a threshold (e.g., 15% APY after all costs).

Execution engine

Opening a position is not a single order. It is a coordinated action:

1. Check available liquidity on the perpetual (bid/ask spread < threshold, typically 0.05%)
2. Place a limit order on the spot (or market with minimal slippage, target 0.01%)
3. Immediately after spot execution, place a short on the perpetual for the same volume
4. If the perpetual is not filled within N seconds (configurable, default 2s)—close the spot, start over

The discrepancy between spot and perpetual execution creates temporary directional risk. On liquid pairs (BTC, ETH) this is milliseconds. On less liquid altcoins, seconds, and the price difference can be significant.

We use WebSocket connections to exchanges to receive fills in real time. REST API for order placement is too slow in an active market.

Position monitoring

Critical metrics that must be tracked in real time:

• delta = spot position + perpetual position (should be ~0)
• margin_ratio perpetual (alert when approaching 150% of liquidation threshold)
• funding_rate current and 7-day moving average
• funding_received_cumulative vs fees_paid_cumulative—actual PnL

PostgreSQL for history storage. TimescaleDB for time-series data (funding rates, prices). Grafana for dashboard. PagerDuty or Telegram for margin call approach alerts.

What's included in the work

• Algorithm development and testing on historical data (minimum 500 backtests)
• Execution engine setup with WebSocket connections
• Exchange integration (CEX/DEX)
• Risk management with margin buffer and TWAP exit
• Monitoring and alert system (Grafana + Telegram)
• Documentation and training for your team
• Post-release support for 1 month
• Guaranteed annual performance review and optimization

Estimated timelines

Algorithm for one trading pair on two CEX (spot + perp) with basic monitoring: 1-1.5 weeks. With multi-pair, multi-exchange logic, automatic roll management, and full risk management: 2-3 weeks. Integration with on-chain DEX perpetuals (dYdX, GMX) adds complexity—separate estimate required. Cost is calculated after clarifying exchange integrations and risk requirements. Typical cost range $10k-$50k depending on complexity.

Contact us for a consultation. Order algorithm development—we'll adapt it to your strategies and trading pairs.