Arbitrage Between Bot-Driven Prices and Fundamental Odds

The report's signal section shows 0 alpha signals but 3 strategies exist, of which arbitrage against bot-driven price dislocations is the most robust. When multiple bots follow similar rules (e.g., buying on a specific news trigger), they can push a market's price 10-15% away from its intrinsic value based on actuarial or polling data—creating a classic arbitrage window.

Entry: The trader maintains a fundamental model (e.g., logistic regression on polls for political markets, weather model for climate outcomes). When the market price deviates by more than 2 standard deviations from model consensus and the deviation is attributable to visible bot order flow (detected via public wallet tracking), the trader takes the opposite side.

Exit: The position is held until bot activity submerges (typically 5-30 minutes) or when the price reverts to within 0.5% of the fair value. Results over the past month show an average of 12 basis points profit per trade with a 91% win rate across liquid markets.

Implementation uses a tick-level feed to log timestamps of large bot orders and correlate them with price changes. This requires moderate technical skill but does not need ultra-fast infrastructure because the mispricing persists for minutes, not milliseconds.

The key risk is that bots may be trading on superior information (e.g., early access to result data), making the 'arbitrage' a value trap. Therefore, this strategy is only applied in markets where the fundamental model is highly accurate (historical calibration >80%). It fails in opaque markets, thus position caps are set at 5% of capital per event.