Arbitrage Trading

Arbitrage Trading

Identify executable price differences while assuming that every apparent spread may compensate for latency, liquidity, settlement, or venue risk.

Note: Agent can identify arbitrage opportunities via price comparison but cannot execute cross-exchange trades directly. Use this skill to detect and report opportunities.

Arbitrage Types

Profitability Check

Cross-Exchange:

• Gross P&L = executable sell proceeds − executable buy cost for the same quantity
• Net P&L = Gross P&L − fees − slippage − funding/borrow − transfer/gas − hedge and rebalancing costs

Triangular:

• Expected cross rate = Price_A / Price_B
• Actual cross rate = observed market rate
• Spread % = (Actual - Expected) / Expected x 100

Require a positive buffer above estimated costs and model error; Net Profit > 0 on stale top-of-book quotes is not sufficient.

Workflow

1. Fetch synchronized executable quotes from every leg. Candle closes are a coarse screening proxy only; do not call the result actionable without bid/ask, depth, size, timestamp, and venue status.

get_candles(symbol="BTC/USD", exchange="binance", interval="5m", count=1)
get_candles(symbol="ETH/USD", exchange="binance", interval="5m", count=1)
get_candles(symbol="ETH/BTC", exchange="binance", interval="5m", count=1)

2. Simulate the exact path using the correct bid or ask at each leg, lot/tick rules, rounding, and available depth.

3. Check carry and convergence terms from the venue's contract specification, funding history, margin rules, and settlement index—not a news summary alone:

get_financial_news(topic="BTC funding rate perpetual futures premium")

4. Stress all costs and failures: one-leg fill, reject, latency, transfer halt, depeg, funding flip, liquidation, borrow recall, and exchange default.

5. Report opportunity with: pair(s), spread %, estimated fees, net profit, and time sensitivity.

Triangular Arbitrage Example

BTC/USDT: 50,000 | ETH/USDT: 2,000 | ETH/BTC: 0.041
Expected ETH/BTC: 2000/50000 = 0.040
Actual: 0.041 → ~2.5% spread (before fees)
Path: USDT → BTC → ETH → USDT

Evidence and Validation

• Treat the setup as a testable hypothesis, not a prediction. Define thresholds, entry, invalidation, and exit before evaluating outcomes.
• Calibrate on the same instrument, venue, session, and timeframe. Use closed candles and a held-out or walk-forward sample; record every variant tried.
• Include spread, fees, slippage, borrow or funding, partial fills, and latency. Reject the setup when net expectancy is not positive or depends on one narrow parameter.
• Return observed inputs, missing data, cost assumptions, entry, invalidation, exit, and a valid, watch, or no-trade status.
• Research basis: John, Li & Liu and Vidal-Tomás show that observed crypto spreads reflect liquidity, settlement, fees, and exchange-default risk—not risk-free profit.

Key Rules

• NEVER report an opportunity without deducting all fees (trading, transfer, gas, slippage)
• NEVER assume execution is instant -- note that cross-exchange opportunities decay in seconds
• Pre-funded accounts on both sides are required for cross-exchange arb; transfers kill the edge
• Use a venue-, size-, and latency-specific minimum edge; no universal spread floor is valid
• Check order book depth: thin liquidity means slippage will eat the spread

Related Skills

• funding-rate-trading -- funding rate arb (spot + perp) is a specific delta-neutral strategy
• altcoin-rotation -- cross-pair price analysis overlaps with rotation screening