Scikit-learn: Machine Learning in Python
The Journal of Machine Learning Research
| Hi-index | 0.00 |
We present Themis, an agent-based simulation of a modern reserve system, along with the financial components of the public sector, and a proxy for the nonbank members of the private sector. The simulation primarily revolves around the key interest rate that benchmarks private lending, which arises from agents' trading within a market for bonds and loans, along with the financial implications of that interest rate. Several tunable parameters allow a user to experiment with different policy configurations and levels of private demand for loans. We both validate the simulation against real-world data, and show the results of three unconventional policy settings: one in which the central bank "loses its appetite" for public debt, a second in which the treasury runs deficits without corresponding debt issuance, and a third in which the central bank uses other policy tools to correct for some of the consequences of the second scenario.