Agents with Multiple Monies: Can the Monetary Policy Trilemma Be Solved?

Abstract

An Agent-based Model (ABM) of an economy and/or economies was implemented to investigate whether the Monetary Policy Trilemma can be solved. Money-demanding agents with heterogeneous preferences over interest rates and exchange rates can freely use multiple monies (with each money corresponding to money-supplying agents with heterogeneous preferences). There were no restrictions on capital movement so the hypothesis tests primarily focused on simultaneous monetary autonomy and fixed exchange rates. The ABM uses an extended version of the interest-parity condition. Adaptive expectations w.r.t exchange/interest rates were used by money-demanding agents (using perfectly rational expectations would have been too computationally intensive) and decisions were made according to the money-suppliers’ and money-demanders’ rationality regarding the impact upon their heterogeneous utility functions. A key finding was that the Trilemma can be solved: monetary autonomy and fixed exchange rates are simultaneously possible in various circumstances. The success rate typically increases over time steps and settles/fluctuates around a certain level. Success generally sharply decreases when ‘Helicopter Money’ is implemented (a more ‘extreme’ monetary autonomy test). The monies with fixed exchange rates often also correspond with money-supply expansions (even when Helicopter Money is implemented). Ceteris paribus, more monies in the system decreases the success rate but increases economic growth.Higher economic growth is correlated with less inequality. Ceteris paribus, the number of money-demanders does not have a discernible impact upon economic growth. An alternative interpretation of all these findings is that the Trilemma should not even theoretically exist or be a problem if there are no capital restrictions – contemporary institutional restrictions are briefly discussed. However, the model is demand-driven and the trading mechanisms, protocol, rationality behind money-creation etc. could be more sophisticated still. Nevertheless, the results are still insightful; they offer food for thought in contemporary debates on macroeconomic and monetary policy. The ABM itself is also a rudimentary formalisation of the theory that “money is an instrument of expectations-management.” The ABM is generalisable – it could apply, for example, to cities, conurbations, regions, countries, multiple countries, industries, continents, the world, transnational communities etc. It was implemented using Python and MESA (ABM framework).