Documentation Index
Fetch the complete documentation index at: /llms.txt
Use this file to discover all available pages before exploring further.
Tune an optimisation policy’s risk, trade and holding aversions by back-testing: a grid of values side by side (risk-return frontier of realized results) or a greedy search that keeps moving the values while Sharpe ratio, return or growth improves.
mpo_tuning: 2 tools.
| Tool | What it does | Notes |
|---|---|---|
mpo_gamma_sweep | Back-tests the same optimisation policy for every combination (at most 24) of the given risk, trade and holding aversions, over the same window and costs. | |
mpo_hyperparameter_search | Searches the risk, trade and/or holding aversions that maximise a back-test metric (Sharpe by default) with the simulator’s greedy method: from the objective’s values it back-tests one move up and down for each tuned aversion, takes the best while the metric improves and stops when no move does. |
fincept_describe_tool.