Using Parameter-Plane Heatmaps to Find Robust Strategy Settings
Summary
The article presents parameter-plane analysis as a way to visualize how combinations of strategy settings affect backtest results. Its example is a Chinese equities strategy combining valuation and float market capitalization ranks. It varies the number of holdings and holding period, runs backtests across the chosen grid, and places total returns in a heatmap. The central recommendation is to favor a contiguous region of strong results over a single standout parameter pair, since nearby settings that perform similarly may be less sensitive to tuning choices.
The example reports peaks at two holdings with holding periods of ten or twenty days, and identifies a broader region around two to four holdings and ten to twenty days as comparatively stable. It selects three holdings for fifteen days from that region. These are reported backtest figures, not evidence of future performance; the article itself warns that its assessment uses total return alone and recommends considering risk measures such as Sharpe ratio and maximum drawdown. It also notes that grid results depend on the chosen parameter ranges and step sizes, and do not identify an exact optimum.
Key ideas
- A parameter-plane heatmap compares backtest outcomes across a grid of strategy settings.
- The example varies portfolio size and holding period for a low-valuation and small-capitalization equities strategy.
- The article favors a nearby cluster of strong settings over the single highest-return point to reduce sensitivity to tuning.
- It selects three holdings and a fifteen-day holding period from the reported stable region.
- The reported returns are backtest results, and evaluating only total return omits risk and robustness checks.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.