The Liquidity of the German Equity Market: Data, Methodology, Results

• Applicant: Professor Dr. Erik Theissen
• Subject Area: Accounting and Finance
• Term: from 2014 to 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 260413935

Our main objective is to create an extensive database containing daily liquidity measures (ca. 30 variables) for the German equity market. The coverage will be 10 years and all stocks contained in the CDAX. This database will provide important input for future research in microstructure and related areas. The raw data (approximately three terabyte of intraday data) will be provided by Deutsche Börse AG. We will carefully screen the data and then aggregate it to the daily level. The daily data wil then be fed into a database that will be freely and at no cost available to other researchers. The only requirement is that researchers have to sign a form un which they guarantee that the use the data for research purposes only (and in particular not for commercial purposes), and that they will not make the data available to third parties. Deutsche Börse has explicitly granted us the right to make the daily data available in this particular way. A second objective of our project (closely linked to the first one) is the creation of a technical document. It will describe the contents and structure of the database in a way such that other researchers can work with the data. Our third objective is to write a paper that describes and analyzes the trading activity and liquidity on the German equity market. Obviously this paper will be based on the data set we intend to create. The fourth and final objective is to write a paper that analyzes the accuracy of low-frequency liquidity measures. Such measures have been proposed in the literature (starting with Roll 1984) in order to provide measures of liquidity that can be calculated when intraday data is unavailable (or too expensive or too cumbersome to use). Previous evaluations of such measures use different methodology and yield ambiguous results. Further, almost all previous papers conclude that the low-frequency measure which has been proposed in the respective paper fares better than all other measures. We plan to perform an analysis that is based on a broader sample (Germany and the US; this is warranted because the market structure in the two countries is distinctly different). We further plan to use a broader set of methods to assure the robustness of our results.

• DFG Programme: Research Grants