Rating
1500
Battle Count: 0
Relevance
8/10
Highly relevant for quantitative researchers and practitioners focusing on factor investing and derivatives. It identifies specific option-implied factors (tail risk, kurtosis, IV convexity) that add pricing information beyond standard equity factors, which can be used to enhance alpha generation or risk models.
Implementation Complexity
7/10
Requires handling high-dimensional data, implementing or adapting the DS-LASSO algorithm, performing gradient boosting with SHAP values for feature selection, and managing complex option pricing data to derive risk-neutral moments.
Reproducibility
4/5
The paper provides detailed descriptions of data sources (CRSP, Compustat, CBOE, French Data Library), specific algorithms (DS-LASSO code from Feng et al. 2020, GBR parameters), and lists all 137 characteristics in the appendix. However, the specific code for the SHAP screening step and the exact construction of some derived option metrics might require careful implementation to match results exactly.
About this paper
Methodology: Double-Selection LASSO (DS-LASSO) with Gradient Boosting Screening. Problem types: Regression, Dimensionality Reduction, Risk Management, Portfolio Optimization.
The interactive Everscope explorer (charts, battles, favorites) loads below.