The creation of value through innovation is among the defining traits of new technology-driven ventures. In this context, patents are an important signalling device to attract external financing. In this paper we contribute to the literature by investigating the value of innovations for start-ups supported by the European Investment Fund (EIF), through its venture capital (VC) instruments, in the years 1996–2014. The value of innovations is measured through patent applications and renewals. We employ an established econometric model to estimate the euro value of innovations based on patent renewal decisions. We find that start-ups in the life sciences hold, on average, the most valuable innovations. At the same time, we find compelling evidence that selection bias, causing less promising inventions to be excluded a priori from patenting, is pervasive across industries and/or regions of Europe. Implications for policy and research are discussed.
Bitte loggen Sie sich ein, um Zugang zu diesem Inhalt zu erhalten
We considered implementing a full-fledged real option pricing model (e.g. Schwartz 2004). However, the data at our disposal could not satisfy the level of granularity and specificity required by such approach.
For the remaining 1957 innovations, current ownership did not coincide with the original applicants. It ensures that these innovations were acquired by EIF-backed start-ups. Interestingly, the ownership of about 38% of acquired innovations further transitioned to other entities, following either the acquisition of the start-up or its bankruptcy.
The figure does not account for utility models and designs, excluded from the analysis. In addition, note that the initial year of the innovation typically equates to the priority year of its underlying patents.
We find over 7% of patenting ICT start-ups consistently following this route, while in other sectors, the incidence is lower than 1%. Nevertheless, 37% of patentors adopted such practice for at least one innovation.
The upper bound restriction is due to renewal data being collected up until 31 December 2016. As such, most applications submitted after 2012 will not have witnessed enough time for the accrual of renewal fees.
The model’s sensitivity to this assumption is tested by varying s in the range of 5–15%. Because of the model’s parametric form, all original MLE estimates are maintained, save for δ which shifts accordingly to counteract the increase or decrease in s. For additional robustness, we tested a firm-specific discount rate s, leveraging on firms’ weighted average cost of capital (based on the methodology of Lünnemann and Mathä 2002). Results are very similar to the ones reported in the remainder of the paper.
This phenomenon may not only be limited to innovations lacking the potential to produce outstanding economic returns but also covers IPs whose revenues may be harder to protect, easier to imitate, etc.
Unfortunately, our dataset does not track financing rounds. Thus, in the remainder, we rely on the assumption that more mature start-ups face a higher likelihood of follow-on investment than younger ventures.