Top

Published in:

2021 | OriginalPaper | Chapter

4. Quantitative Study: The Role of OSS for Likelihood and Timing of Acquisitions

Author : Michael Vetter

Published in: Acquisitions and Open Source Software Development

Publisher: Springer Fachmedien Wiesbaden

Activate our intelligent search to find suitable subject content or patents.

search-config

AI-assisted search

Off

Abstract

The previous chapter’s qualitative study suggested that firms’ OSS activities can be relevant for acquisition decisions. The study presented in this chapter attempts to generate first quantitative evidence for whether participating in open innovation ecosystems, such as OSS development, plays a role in acquisitions.

Dont have a licence yet? Then find out more about our products and how to get one now:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

über 102.000 Bücher
über 537 Zeitschriften

aus folgenden Fachgebieten:

Automobil + Motoren
Bauwesen + Immobilien
Business IT + Informatik
Elektrotechnik + Elektronik
Energie + Nachhaltigkeit
Finance + Banking
Management + Führung
Marketing + Vertrieb
Maschinenbau + Werkstoffe
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

inform now

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

über 67.000 Bücher
über 340 Zeitschriften

aus folgenden Fachgebieten:

Bauwesen + Immobilien
Business IT + Informatik
Finance + Banking
Management + Führung
Marketing + Vertrieb
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

inform now

previous chapter Qualitative Study: How Does the Involvement of Firms in OSS Matter in the Pre-acquisition Phase?

next chapter Mixed-methods Study: The Effect of Acquisitions on Open Source Software Development

Available only for authorised users

See Stein (2017) and Hlavka (2019) for an explanation how reducing uncertainty can lead to earlier acquisitions and Ransbotham and Mitra (2010) who show that patenting activity is a positive signal towards potential acquirers leading to a higher probability of getting acquired.

See Section 2.1.4 for a more detailed description of the general benefits and downsides of a firm’s engagement in OSS without considering the context of acquisitions.

The practical relevance of this topic is highlighted, e.g., by a checklist for OSS license due diligence in acquisitions released by the Linux Foundation (https://www.linuxfoundation.org/open-source-management/2019/03/assessment-open-source-practices/; last accessed on 02.12.2020) or commercial offerings for OSS license due diligence, e.g., by Synopsis (https://www.synopsys.com/software-integrity/open-source-software-audit.html; last accessed on 02.12.2020).

https://www.freecodecamp.org/news/the-top-contributors-to-github-2017-be98ab854e87/ last accessed 07.12.2019.

https://venturebeat.com/2018/11/08/github-passes-100-million-repositories/ last accessed 12.06.2020.

I am grateful to Georgios Gousios, founder of the GHTorrent project, for giving me the opportunity to discuss my approach with him and granting me access to his GHTorrent MongoDB server, giving me access to over 12TB of raw GitHub data without being required to replicate the full database ourselves (available for download here: http://ghtorrent-downloads.ewi.tudelft.nl/mongo-daily/ as of 12.06.2020).

https://scholar.google.gr/scholar?cites=11132126230347149781; last accessed 12.06.2020.

The Crunchbase industry group “software” contains 90 software related software-related sub-industries such as “Enterprise Software”, “Online Games”, “SaaS”, “Developer Platform”, etc. See https://support.crunchbase.com/hc/en-us/articles/360043146954-What-Industries-are-included-in-Crunchbase- for the full list (last accessed 12.06.2020).

I exclude forks since they inherit all contributions from their respective parent project and hence create a large number of duplicates.

GitHub users are only associated with a firm for the time they use the firm’s domain in their commits. Before they start using the domain and once they stop using the domain, I do not consider them an employee anymore. This way, I can ensure that I only capture activities that have happened when the user was employed by the focal firm.

A firm can be acquired every month until it is actually acquired, going public or ceases operations. Hence, the one year period is always calculated relative to the focal month.

In GitHub a commit is always linked to the person who wrote the code, be a user with direct write access to the project or external contributors who have to submit their proposed commits to a project via a pull request.

https://en.wikipedia.org/wiki/Comparison_of_free_and_open-source_software_licences (last accessed 24.06.2020).

https://help.github.com/en/github/creating-cloning-and-archiving-repositories/licensing-a-repository (last accessed 24.06.2020).

I thank Georgios Gousios, founder of the GHTorrent project, for giving me the opportunity to discuss my approach to utilizing GitHub data with him and granting us access to his GHTorrent MongoDB server.

https://opensource.org/licenses/alphabetical (last accessed 24.06.2020).

The high share of permissive licenses compared to copyleft licenses seems common for OSS development in the last decade and was confirmed by my interviewees. Non-scientific research on GitHub licenses shows similar results (https://www.kaggle.com/mrisdal/safely-analyzing-github-projects-popular-licenses; last accessed 15.12.2020).

I cannot capture changes of licenses of the same project over time as there is no data available for such license changes. Yet, such changes are rare and I do not expect them to influence the results.

Publicly available in Google’s cloud platform: https://cloud.google.com/blog/products/gcp/google-patents-public-datasets-connecting-public-paid-and-private-patent-data (last accessed 05.12.2020).

The assignee and company name cleaning logic is available for download here: https://zenodo.org/record/3594743 (last accessed 05.12.2020). I thank Ali Samei for pointing me to this method.

E.g., the Crunchbase company name “Element” is contained in more than 1700 assignee names in Google Patents data, among others “Element Science Inc”, “Element Ltd”, “Element Co Ltd”, “Element Ltd Corp”, “Element KK”, “Element Labs Inc”, “Element”. As a result, this record was dropped.

Categories of the number of employees recorded in Crunchbase are as follows (my mapping to a numeric scale in brackets): 1 to 10 (1) 11 to 50 (2) 51 to 100 (3) 101 to 250 (4) 251 to 500 (5) 501 to 1,000 (6) 1,001 to 5,000 (7) 5,001 to 10,000 (8) 10,001 + (9) employees. As the classes are not evenly sized, this mapping leads to a pseudo-logarithmic scale.

I applied the weighting logic known from Coarsened Exact Matching (Blackwell, Iacus, and Porro, 2009), where I fixed the weight of all firms active in OSS development to 1 and adjusted the weights of the non-active firms accordingly. E.g., if one “OSS-active” firm matches with two “non-active firms”, each non-active firm gets the weight 0.5. If one “non-active firm” matches with two “OSS-active” firms, this non-active firm gets a weight of 2.

Hlavka (2019) found that if a target received investments, it is on average acquired earlier than if it did not, which supports the argument that investors may seek an earlier exit which in turn increases acquisition likelihood for young firms.

Adjusting for controls is a function part of Stata’s ststs-package. When adjusting for controls, Stata fits a separate CPHM on the independent and control variables and retrieves the separately estimated baseline survivor function per variable (see https://www.stata.com/manuals13/ststs.pdf; last accessed 12.01.2021). Without adjusting for controls, the graphs show the share of firms from either group that at an age of t month had not yet been acquired as estimated by the full CPHM.

I choose this OLS model in line with previous research on acquisition timing (e.g., Fischer et al., 2019; Hlavka, 2019; Stein, 2017), including research on timing and the role of characteristics of the acquirer (Hlavka, 2019). I do not use a CPHM in this section, as CPHM does not allow to examine acquirer characteristics (surviving firms don’t have an acquirer). On the other hand, I do not use the OLS model in the two previous sections to examine target characteristics, as my data is limited to the early years of a firm’s life and a large share of firms does not get acquired in the observation period, but might get acquired later (i.e., given the higher acquisition hazard for firms active in OSS development compared to non-active firms, there are more non-active firms “surviving” my observation period which could get acquired in the future). These potentially missing acquisitions are no problem when examining characteristics of the acquirer within the group OSS-active targets, which are in the focus of this section. Obviously, one needs to keep in mind, that the results obtained are only valid for young firms, which are in the focus of this chapter.

Title: Quantitative Study: The Role of OSS for Likelihood and Timing of Acquisitions
Author: Michael Vetter
Publisher: Springer Fachmedien Wiesbaden
Book: Acquisitions and Open Source Software Development
Print ISBN: 978-3-658-35083-3

Electronic ISBN: 978-3-658-35084-0

Copyright Year: 2021
DOI: https://doi.org/10.1007/978-3-658-35084-0_4

Springer Professional

Abstract

Please log in to get access to your license.

Dont have a licence yet? Then find out more about our products and how to get one now:

Springer Professional "Wirtschaft+Technik"

Springer Professional "Wirtschaft"

Premium Partner