Replication data for: Dispelling ESG Investing Risk Misconceptions (ECMODE-D-25-02812 )

## 1) Purpose
This repository creates synthetic return datasets (and optionally daily returns) and saves final outputs in a submission-ready format.

Replication materials are organized into
(i) input data folders (RF_data/, Y_data/),
(ii) code (code_submit/ and root scripts), and
(iii) generated outputs (data_submit/).
The pipeline is run from the project root by executing master.R (or sequentially make_synthetic_all_mean_var.R and make_synth_daily_returns_mean_var.R), which reads inputs from RF_data/ and Y_data/ and writes synthetic datasets and summary tables to data_submit/. Random-forest analyses with nested temporal cross-validation are implemented in code_submit/rf_nested_cv_temporal.R, with extensions for interaction analysis and SHAP-style interpretation in code_submit/rf_nested_cv_temporal_interactions_shap.R. All output files referenced in tables and figures are contained in data_submit/ and can be regenerated from the scripts using relative paths.