Searching Stock Factor Weights with Sharpe-Based Backtests
Summary
This example combines three cross-sectional stock measures into a composite score: market capitalization rank, turnover rank, and the rank of the latest close-to-prior-close ratio. It filters out risk-warning stocks, recently listed stocks, certain listing sectors, and firms with nonpositive trailing earnings. A grid search varies the weights assigned to the three ranks, runs a portfolio for each combination, and selects the combination with the highest estimated Sharpe ratio. The code also collects portfolio value series and plots them together for visual comparison.
The material illustrates a workflow for factor-weight exploration and performance visualization, but it does not show numerical results or the referenced portfolio curves. It computes annualized return, volatility, and drawdown, although the selection criterion is Sharpe ratio assuming a zero risk-free rate. The account gives no details on rebalance timing, transaction costs, data availability, or out-of-sample validation. Searching many weight combinations on the same history can overfit, so the winning weights should not be treated as validated without further testing.
Key ideas
- The stock score blends ranks for market capitalization, turnover, and recent price change.
- Eligibility filters exclude risk-warning stocks, newer listings, selected sectors, and firms with nonpositive trailing earnings.
- A grid search evaluates combinations of factor weights and chooses the highest observed Sharpe ratio.
- Portfolio value curves are collected and plotted to compare candidate weight combinations.
- The document shows no numerical outcomes and does not establish out-of-sample performance.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.