Arnold Polanski, Evarist Stoja, and Frank Windmeijer, "Telling Tales from the Tails: High-Dimensional Tail Interdependence", Journal of Applied Econometrics, Vol. 34, No. 5, 2019, pp. 779-794. The stock price data used in this paper are obtained from Eikon/Datastream and are the property of the Thomson-Reuters company, who do not permit open access. They offer a paid-for subscription, and hence the data cannot be lodged here. The data are held at the Thomson-Reuters' platform Eikon https://eikon.thomsonreuters.com/index.html but can be easily obtained if one has a subscription to this paid-for service. The analysis in the paper is based on the first 250 constituents (please see information below regarding the mnemonics of companies) of the Standard & Poors 500 index which have complete data between 02 January 1990 and 29 December 2017 (7304 observations). In addition, we also use the Fama-French-Carhart factors for the same period as above (02 January 1990 and 29 December 2017) obtained for free from Kenneth French's website http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html. This dataset yields 7057 observations for the Fama-French-Carhart factors. The dates on the two datasets do not perfectly match. In particular, Kenneth French's dataset does not have observations for every single day in the Thomson-Reuters data. In this case, we synchronise the data using Kenneth French's data calendar as the benchmark. This results in 248 (7304-7056) observations from the Thomson-Reuters dataset being omitted from the analysis. These observations are the price data from Thomson-Reuters which do not have a matching observation in that particular day in Kenneth French's dataset. In the CSV file that accompanies this readme file, we provide the Eikon mnemonics as well as the names of the S&P companies (in the top row) and the dates of Kenneth French's data in the first column. The zip file psw-files.zip contains two files. One contains Mathematica code, and one contains mnemonics for the S & P series. These are ASCII files in DOS format. Unix/Linux users should use "unzip -a". Evarist Stoja E.Stoja [AT] bristol.ac.uk