3 files

The Epps effect under alternative sampling schemes: Dataset

posted on 19.11.2020, 09:42 by Patrick Chang, Etienne Pienaar, Tim Gebbie
The data consists of transaction data for 10 equities from the Johannesburg Stock Exchange. The data consists of five trading days ranging from 2019-06-24 to 2019-06-28. The data has been processed to only contain transactions. Furthermore, transactions with the same time stamp have been aggregated using a volume weighted average so that there is only one trade per time stamp. Missing data is indicated with NaN's.

The 10 equities included are: FirstRand Limited (FSR), Shoprite Holdings Ltd (SHP), Absa Group Ltd (ABG), Nedbank Group Ltd (NED), Standard Bank Group Ltd (SBK), Sasol Ltd (SOL), Mondi Plc (MNP), Anglo American Plc (AGL), Naspers Ltd (NPN) and British American Tobacco Plc (BTI).

The data structure in each csv file is 10 columns which contain the trading information for the assets traded. The transaction data are in chronological order. The three files have the exact same structure with each file containing information for the transaction tuple: price, time and volume.

The data should only be used to aid the reproducibility for the paper:

The Epps effect under alternative sampling schemes. The steps to reproduce the results can be found in our GitHub site: https://github.com/CHNPAT005/PCRBTG-VT.

The research focuses on investigating the Epps effect under different definitions of time.

The work is funded by the South African Statistical Association. The original data was sourced from Bloomberg Pro. The code for the research is done using Julia Pro.


South African Statistical Association



Statistical Sciences



Logo branding