Compute IC/IR stability and quantile backtests for cross-sectional factor signals.
npx skills add https://github.com/loanntc/Paave --skill factor-research-loanntc
Please help me install this Agent Skill. Skill: factor-research Source: https://github.com/loanntc/Paave/tree/main/skills/factor-research Command: npx skills add https://github.com/loanntc/Paave --skill factor-research-loanntc
It helps you determine whether a cross-sectional trading factor truly has predictive power, and how to select and combine factors without introducing look-ahead bias.
Use the factor-research Skill to run IC/IR and quantile backtests by providing a factor CSV and a forward N-day return CSV whose dates and instrument codes align exactly.
High-intent search queries and answers about installing and using this skill.
You can validate cross-sectional factor predictive power by computing the Information Coefficient (IC) series and running quantile backtests. This measures IC stability and generates group equity curves to assess stock-selection monotonicity.
To avoid look-ahead bias during factor evaluation, you must use strictly aligned factor and forward-return CSVs. Forward returns are computed specifically after the factor observation date to ensure no future data leaks into the signal.
Quantile backtesting sorts instruments into a configurable number of groups based on factor values, producing equity curves for each group. This reveals monotonicity, long-short spreads, and nonlinearity in the factor's stock-selection performance.
Yes, you can combine multiple effective factors into a single portfolio signal using equal-weight, IC-weighted, or orthogonalized methods. This multi-factor screening process merges individual validated signals into a composite score.
You need aligned factor and forward-return CSV files with exactly matching date and instrument code columns. Pandas, numpy, and scipy are required dependencies to compute the IC series, IC summary, and quantile group equity curves.
ebrahim-saniEvaluate cross-sectional financial factors with IC/IR statistics and quantile backtests.
wudyeCompute IC/IR statistics and quantile backtests for cross-sectional factor research.
brainbytes-devDesign and optimize momentum-based trading strategies across time-series, cross-sectional, and dual momentum paradigms.
ajithkumar31082004-bitEvaluate financial factor predictive power with IC/IR and quantile backtests.
