EUIPO data

European Union trade mark (EUTM) data from the EUIPO and patent data from PATSTAT were used.

Only applicants from the European Union were selected since the matching algorithms used is optimised for them. Due to lack of detailed data on applicants, IPRs applied for in national IP offices are not part of the final dataset.

Thus, the data from Invest Europe, the EUIPO and the EPO were matched by an algorithm developed for years by the different economic studies of the EUIPO, which uses the name of the firm and the country of its address, with standardised names and legal forms. This algorithm is able to find a maximum of matches with a very low number of false negatives. Later, part of the data was manually reviewed.

The information provided is the application for each of the EUTM and European patent applications.

In total, 18,786 companies were identified, with 57,918 EUTMs, and 80,465 European patents, which represents an important sample of data.

Invest Europe (EDC) data

Invest Europe, in collaboration with its national association partners, collects data pertaining to the private equity and venture capital industries across Europe. This data is collected through the European Data Cooperative (EDC), which is the most comprehensive database of European private equity and venture capital statistics. The EDC serves as a single data entry point for members of private equity and venture capital associations and other contributors across the continent, including entities associated with Invest Europe and the national associations.

The data encompasses a range of information, including data on fundraising, investments, and divestments, as well as on economic impact and Environmental, Social and Governance (ESG). Audit efforts are conducted in close coordination with data contributors and partnering national associations with the objective of ensuring the best coverage and consistent application of methodology. Invest Europe processes all available information at the time of the data collection cut-off to produce its annual statistics. Data collection is conducted throughout the year, contingent upon the schedule of Invest Europe’s publications. The primary activity statistics are gathered from November until mid-March, with auditing occurring prior to the publication in May. For the purposes of this study, audited data from 2007 to the first half of 2023 was used.

Methodology

Matching

To develop a comprehensive database, a matching exercise was conducted between the dataset of the EUIPO, PATSTAT and Invest Europe in two stages. First, all reporting profiles included in the EDC platform were matched with the IPR dataset from EUIPO and PATSTAT (see Appendix). Second, this matched dataset was integrated with Invest Europe’s investment data. The main variables used from the investment dataset for this study include the year of investment, the sector in which the investment was made, the fund stage focus, and the investment stage.

The resulting database exhibits the full Invest Europe investment database between 2007 and H1 2023, and comprises observations of IPR activity of 56,042 portfolio companies before and after each round of investment by the private equity and venture capital funds, and 103,131 filings between patents and trade marks. The key variables of interest in this study are the portfolio company’s records of patent and trade mark applications before and after the investment round, identified by the quarter of the year in which the investment took place, and the amounts invested by the funds during the quarter.

To understand how the relationship between private equity and venture capital investing and IPR activity is shaped, the database was also broken down by the investment stages of the company, resulting in chunks made of venture capital⁴, growth, and buyout companies.

Models

The resulting database contains information not only about the number of IPR applications filed by the portfolio company before and after each round of investment but also about the amounts invested by the respective private equity and venture capital funds in each period. Linear regression models and logit regression models are run to investigate the relationship between investment amounts deployed by private equity funds and the IPR activity of the backed portfolio companies. Logarithmic transformations⁵ were performed to normalise some data and improve the model fit. Consequently, the coefficients of the linear regression models are interpreted as elasticities and semi-elasticities⁶, reflecting the effects of percentage changes rather than unit changes.

Control variables

Several control variables were considered, including the sector in which the company operates, the country where it is based, and its current stage of investment.