Skip to content

Data used in the paper "A SSD approach to enhanced indexation with sector constraints"

Notifications You must be signed in to change notification settings

cristianoarbex/subsetSSDData

Repository files navigation

Data for the paper

This repository contains all data used to produce the results in the following paper:

Subset SSD for enhanced indexation with sector constraints

The following files are available:

US stocks data

  • marketDataUS.csv: A csv file containing prices for every asset in the S&P500 index, from the end of 2018 to the end of 2023. Each row represents a date and each columns an asset; assets are identifier by their ticker. A missing price means that either the price did not exist at the time or was not found. A negative price means that the asset was not part of the S&P500 at the time.
  • marketDataUS_benchmarks.csv: A csv file containing prices for the S&P500 index and all 11 sector subindices.
  • marketDataUS_tags.csv: A csv file containing the sector to which each asset belongs.

Fama-French 49 industry portfolios

  • marketDataFF.csv: A csv file containing prices for every one of the 49 Fama-French industry portfolios, from the end of 2018 to the end of 2023. Each row represents a date and each columns an asset; assets are identifier by their ticker. A missing price means that the corresponding portfolio was not computed at the time. Note that the original data is given as daily returns, we computed prices assuming an initial price of $0.1 in 1926 (when the data begins) and compounded the returns accordingly.
  • marketDataFF_EW_benchmarks.csv: A csv file containing prices for the equally-weighted benchmarks as calculated in our paper, including the equally-weighted benchmark of all 49 portfolios (code: EW) and the equally-weighted sector indices.
  • marketDataFF_tags.csv: A csv file containing the sector to which each portfolio belongs, as classified by us. Note that since real estate is a single portfolio, it was put into the financials sector. So effectivelly there are 10 Fama-French sectors instead of 11 as above in the US data.

Data for CRSP indices has not been included as it is not publicly available.

In the paper, the experiments were run from 2018-12-31 until 2023-12-31. Every rebalance uses 60-days of historical returns ending at the day of the rebalance, hence the data also includes 60 prices prior to 2018-12-31.

With respect to a minor technical point here while there are eleven sector indices according to GICS, only one of the 49 Fama-French portfolios fits into the real estate sector. To prevent a sector portfolio containing a single asset we then merged, for this experiment, the financials and real estate sectors into one. We did this for both the equally-weighted and the CRSP sector indices.

About

Data used in the paper "A SSD approach to enhanced indexation with sector constraints"

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published