Evology: an Empirically-Calibrated Market Ecology Agent-Based Model for Trading Strategy Search
Our paper, Evology: an Empirically-Calibrated Market Ecology Agent-Based Model for Trading Strategy Search, was presented at the AI for Agent-Based Modelling workshop at ICML 2022.
It is joint work with Aymeric Vié and J. Doyne Farmer. Evology views the market as an ecosystem of diverse, interacting, evolving trading strategies. Because it is empirically calibrated, it doubles as a training environment: rather than assume a strategy and test it in isolation, you let the strategy compete against the live ecology and search for what survives. It is a companion to Towards Evology, which calibrates the same ecology to US equity mutual funds.
The model and the search experiments are in the paper.