Paris School of Economics - École d'Économie de Paris

La science économique au service de la société

Le modèle utilisé pour la base de données

The logical model of the database, run in an Oracle environment, has been conceived to take into account several constraints. First, it has been set up to build an open, scalable and interdisciplinary database. It has not been designed to only store financial data on securities listed on the Paris Stock Exchange and their issuers. Beyond the other French markets and potential extensions to other European stock markets, the logical model of the Equipment can cover the data on issuing companies not incorporated into the initial version and produced by other projects such as prosopographic data on directors, workers, employees, and shareholders of a listed corporation, and on unlisted companies, whatever their juridical status, such as the location, the balance sheets, the partenaires, the capital, its changes and its distribution among the partenaires.

Moreover, the logical model has been conceived to satisfy specific constraints such as the storage of the data as they appear on historical sources. It is then possible from the database to reconstruct the official lists and the yearbooks as they were printed. While linking the images of the sources directly to the database would have produced a too heavy structure, the researcher interested in the historical evolution of data will be able to observe it directly into the database. For example the text of the yearbooks’ notices on issuers is split among the various parts composing the original document, indexed and stored in a specific table in such a way the researcher can easily rebuild the text. On the same ground it is possible to rebuild the list of the securities traded at the Paris Bourse as it is on the official list (see sect. 3.3.1).
Nevertheless, storing the data as they are on the sources would limit the workability of the database for a researcher interested in its financial dimension, i.e. harmonized prices’ time series. This is why the data are also stored within a system allowing an easy extraction of them. The technologies adopted for the data entry have needed specific adaptations of the database logical model to manage and track the data production. Overall, the DFI-H database is built around different but interconnect systems.

To reach this goal, the logical model of the Antwerpen University’s SCOB database (http://www.scob.be/what.html) has been extensively developed to store data that were not taken into account such as the derivatives prices and exchange rates, to adapt it to the French sources and the D-FIH project technologies of production, and to set the database for future extensions (for example, specific systems to easierly manage cross-listings and to store quotations in different currencies are under development).