lb trading

ntarsis got his HRT return offer and has started full time. he works on the core prediction-market market-making team. at a high level, he quotes a fixed spread around a fair probability on a per-market basis.

ntarsis notices something unusual as time goes on. end users seem to buy YES contracts more frequently than they sell them. e.g., when someone with YES contracts for trump to win the election changes their mind (or wants to hedge), they buy YES contracts for kamala instead of selling their YES contracts for trump. these are economically equivalent because trump and kamala are complimentary.

ntarsis runs a small backtest and confirms his suspicion: there is an exchange-wide, systematic bias towards market orders buying YES contracts.

ntarsis immediately pages his team and proposes a new method of quoting. what does he propose?