(1/10) During the 2024 US Presidential election, @Polymarket distributed over $25K daily in market maker rewards. Here's how I became a top 5 recipient by doing something radical – measuring volatility. I'm also open-sourcing my code.
@Polymarket (2/10) When new markets get retail-dominated, easy opportunities emerge. The opportunities stay longer when the consensus is that the market is unique. But anything with a price that moves is the same. It can be modeled, its volatility calculated, and strategies backtested
@Polymarket (3/10) First, I decoded the polymarket reward formula. Despite the scary looking docs, it's simple: rewards multiply if you post on both sides (bid/ask), and the closer you put the bid/ask, the exponentially better rewards you get.
@Polymarket (4/10) My system retrieved market data and calculated live volatility. When I frequently requested historical data, I crashed Polymarket's API server. It was a super bullish sign as it meant no one else was systematically analyzing this data live
@Polymarket (5/10) I upgraded to websockets and selected markets daily with the best risk-reward ratio: low volatility with high expected rewards. It was easy because people were providing the same liquidity in a 500% volatility market as they did in a 100% for the same reward.
@Polymarket (6/10) My bot targeted markets below volatility thresholds, placing YES and NO orders one tick above the nearest 200-share level. Risk management included position exits, dynamic stop-losses, continuous trade flow monitoring, and orderbook imbalance checks to protect capital
@Polymarket (7/10) I relied entirely on backtests for these parameters and it revealed something powerful: market signals were embedded in orderbook dynamics and cumulative trade volumes before showing up completely in price action.
@Polymarket (8/10) And, Instead of optimizing for each market, I classified params into four types based on liquidity patterns and behavior. Each type shared the same parameters for volatility thresholds and stop-losses, making the system surprisingly scalable.
@Polymarket (9/10) This strategy shows that while most makers focus on news-based MMing, simpler and well-accepted market-making strategies worked well in Polymarket. I'm open-sourcing my repo at http://github.com/warproxxx/po... and believe the protocol can multiply its liquidity with such vaults.
@Polymarket (10/10) The repo includes websocket handling, position merging, and performance tracking, which the official SDK lacked. Follow me @defiance_cr for more insights on cryptos, stocks, commodities, exchange operations, and market making.
@defiance_cr @Polymarket This is a great thread btw! Our team are fans of Polymarket and have also developed custom tools for analysis and trades previously. Would love to know how we can make your experience just as great or even better on Foresight!
@foresightnow @Polymarket Thank you. I just went thru your platform. I love that you guys have lending, too, allowing leverage on your bets. Perps are nearly impossible in prediction markets, at least after the price crosses >0.85. P2P lending is a good call around it, where the market decides the rate
@defiance_cr @Polymarket always edge somewhere!