This repository covers codes and data implemented in the research paper titled ''Do We Need Higher-Order Comoments to Enhance Mean-Variance Portfolios? Evidence from a Simplified Jump Process'' The paper is available on SSRN.
The data was collected from two sources. The first one is CRSP, which is based on scraping the S&P 500 constituents. The CRSP data is available here. The second is more common and relies on the book to market portfolios by Fama-French. The second data is available here.
The repository contains two main codes. The first code illustrates how to construct the final time series for estimation and portfolio formation. This file is named "create_data.R". The second file contains the main code of the research. It includes estimation functions, portfolio optimization, backtesting, and bootstrapping. The main file reproduces all results of the manuscript.