This repository contains a replication of Lewellen (2015), "The Cross-Section of Expected Stock Returns", published in Critical Finance Review. The study explores how investors can use firm characteristics to generate real-time forecasts of a stock’s expected returns using Fama-MacBeth regressions.
The project aims to replicate Table 1, Table 2, and Figure 1 from the original paper using CRSP and Compustat data. It examines the predictive power of expected returns computed using multiple firm characteristics.
- CRSP (Center for Research in Security Prices)
- Compustat (Standard & Poor’s financial database)
To quickest way to run code in this repo is to use the following steps. First, you must have the conda package manager installed (e.g., via Anaconda).
Thenm open a terminal and navigate to the root directory of the project and create a conda environment using the following command:
conda create -n env_name_here python=3.12
conda activate env_name_here
and then install the dependencies with pip
pip install -r requirements.txt
Finally, you can then run
doit
And that's it!