SHARKFin has several implementations of a financial market, and may have more in the future.
While the macro-agents do not currently have rational expectations of the market, they could potentially be given a learning function that learns those expectations based on Monte Carlo simulation.
Currently, the expectations of the macro-agents are set by the FinancialModel class. The current implementation is the backwards-looking, statistical inference process designed by John.
An alternative FinancialModel implementation might load expectations data from an external file, which had the data for a model that was trained via convergent Monte Carlo.
SHARKFin has several implementations of a financial market, and may have more in the future.
While the macro-agents do not currently have rational expectations of the market, they could potentially be given a learning function that learns those expectations based on Monte Carlo simulation.
Currently, the expectations of the macro-agents are set by the FinancialModel class. The current implementation is the backwards-looking, statistical inference process designed by John.
An alternative FinancialModel implementation might load expectations data from an external file, which had the data for a model that was trained via convergent Monte Carlo.