Możliwości naśladowania agentów-handlarzy na giełdzie papierów wartościowych
[Prospects of imitating trading agents in the stock market]

Mateusz Wiliński (IDEAS Research Institute, Warsaw & Tampere University, Finland)


In this work we show how generative tools, which were successfully applied to limit 
order book data, can be utilized for the task of imitating trading agents. To this 
end, we propose a modified generative architecture based on a state-space model, 
and apply it to limit order book data with identified investors. The model is 
trained on synthetic data, generated from a heterogeneous agent-based model.