Combining Polymarket Order Book Data With Chainlink TWAP

Combining Polymarket order-book prices with Chainlink Time-Weighted Average Price (TWAP) data can help estimate fair probability in short-term Bitcoin (BTC) markets. Polymarket markets may trade at prices that differ from Bitcoin's spot market prices. The article highlights the importance of separating two signals: market signal (Polymarket order book) and underlying-price signal (Chainlink BTC/USD price).

Using Black-Scholes model can provide a probability-related quantity, but it should be treated as a research model rather than an oracle. Combining Polymarket order-book data with Chainlink TWAP data can be a more useful framework. The model edge is the difference between the model probability and market probability. Evaluating the model edge can help determine if there is an edge that persists after accounting for spread, execution, changing volatility, and the difference between the price model and the Chainlink resolution mechanism.

TWAP can reduce sensitivity to instantaneous noise, providing a more stable signal. However, aggressively smoothing the price can result in losing important information. The order book adds valuable information about market positioning and liquidity, which the Chainlink price alone cannot provide. This mental model emphasizes the importance of combining both data sources to gain a better understanding of the market.

Some common trader mistakes include assuming BTC price equals the resolution price, equating market price with objective probability, and relying solely on Black-Scholes model assumptions for short-term crypto events. Additionally, assuming that a TWAP is a better signal automatically is a misconception. Aggressive smoothing can introduce lag.

The article suggests a practical research experiment, which involves collecting synchronized observations: timestamps, Chainlink reference price, TWAP, estimated volatility, Polymarket bid, and Polymarket ask. This approach can help in understanding the relationship between the different data sources and the fairness of probability estimates in short-term BTC markets.