The field of prediction markets and time series modeling is moving towards the development of more sophisticated and standardized tools for quoting, hedging, and transferring risk. Researchers are working on creating unified stochastic kernels and market maker's handbooks to improve the efficiency and transparency of prediction markets. Additionally, there is a growing interest in understanding the implicit biases of design choices for time series foundation models and their impact on model quality. The use of foundation models in finance is also being explored, with a focus on disentangling systematic and idiosyncratic risk. Notable papers in this area include:

• Toward Black Scholes for Prediction Markets: A Unified Kernel and Market Maker's Handbook, which proposes a logit jump-diffusion kernel for prediction markets.
• Trading with the Devil: Risk and Return in Foundation Model Strategies, which introduces an extension to the Capital Asset Pricing Model to disentangle systematic and idiosyncratic risk in foundation model strategies.