Entrepreneurship has long been a key factor contributing to a country’s economic development. During the 1980s, while Fortune 500 firms lost 4 million jobs, firms with less than 100 employees added 16 million jobs (Birch, 1990). Advancements in technology and an expansion of the financing industry meant increasing startup activity through the 90s and up to the 2008 financial crisis. This trend was largely reversed as a consequence of the financial crisis, however in 2015 a large reversal in startup activity took place (Fairlie, 2015). Entrepreneurship remains a very relevant area economically and it has gained traction in theoretical research in the 21st century.
Startup activity is strongly related to financing availability; entrepreneurs require funding to pursue their business ventures and most of the time they will not have enough capital to avoid external sources of funding. Startups are typically not yet profitable, lack tangible assets and do not have a proven track record (Denis, 2004). These problems make it very challenging for entrepreneurs to gain access to funding. Furthermore, agency problems and information asymmetries worsen this challenge.
A vast amount of research on startups funding is focused on the reasons for entrepreneurs to choose between alternative capital providers. Nevertheless, Robb and Robinson (2012) argue that the pronounced information asymmetries that entrepreneurs face may imply that they seek capital where it is most plentiful, implying that they do not freely choose between available financing channels. This insight stimulated me to focus on why financiers decide to fund certain startups instead. The choice of funding certain startups by financiers will be primarily driven by the expected capital gains to be realized from investing in the startup. However, considering that most startups are difficult to value, and that a significant number of new ventures fail, financiers will take into account a multitude of factors to screen for attractive new business ventures ex-ante (prior to investment choice).
Cochrane (2005) analyzes the risk and return of venture capital investments using the VentureOne database. The findings from his research are that returns on venture capital investments are extremely volatile, with early stage deals being much riskier than later stage deals. Considering the volatility of VC investments, the financier’s ex-ante process by which startup-projects are chosen must be an important factor affecting the financier’s investment success. Due to the presence of agency problems and information asymmetries in entrepreneurial finance, startups must attempt to convey their capabilities and technological accomplishments to financiers in an attempt to demonstrate the actual value of their business venture. Conti et al. (2013) examine the role of patents as signals used by investors to reduce information asymmetries. A patent publicly discloses the scope and specification of an invention (Conti et al. 2013). Moreover, research from Graham and Sichelman 2008 and Graham et al. (2009) finds that a primary reason for entrepreneurs to patent is to secure financing. Patents are assets that startups can use as collateral to access formal debt financing. Furthermore, since patents with misinformation can be overturned, they provide a way to reduce information asymmetry and convey startups valuable information to potential investors without risking information spillover or competitors exploiting their inventions (Long, 2002; Kortum & Lerner, 2000). Therefore, patents should be of particular importance to financiers when conducting their process of selecting startups that they are willing to provide funding to. The importance of patents in securing financing for startups led to my main research question:
To what extent does patenting increase a startup’s access to formal external funding?
Moreover, I will explore how patent quality, proxied with a variety of variables, affects a startup’s choice of funding and the amount of funding realized. Finally, I will examine whether the startups alternative sources of capital and the number and quality of the startups patents affect the eventual success of the business venture. Therefore, I propose:
- Hypothesis 1: The amount of successful patent applications by a startup will be negatively related with the time taken to acquire external funding.
- Hypothesis 2: The amount of successful patent applications by a startup will be positively related with the amount of financing it raises in early-staged financing rounds.
- Hypothesis 3: The amount of successful patent applications by a startup will lead to higher valuation at time of initial public offering or acquisition
From an entrepreneur’s perspective, understanding the role of patents as a signal of venture quality is of extreme importance. Understanding the role that patents have in helping entrepreneurs access alternative sources of funding is critical. Theoretically patents are a good example of a signal as theorized by Spence, they are costly to obtain and can be used to assess the quality of a startup’s innovation capabilities (Spence, 1973). In addition, patenting is a time-consuming and expensive endeavor which should only be pursued if the costs involved are equal or lower than the expected benefits it will secure. The role of quality signaling that patents provide may be overlooked by entrepreneurs, this could mean that new business ventures systematically undermine the value of patents and forego a value-adding process that would enhance the probability of startup success. Considering that the costs associated with the patenting process are approximately $25,000 (Lemley, 2000), and that the average costs for an EPO (European Patenting Office) patent are around €32,000 (Haeussler et al. 2009) it is likely that a large number of young startups will avoid patenting until they gain access to external financing.
Most of the existing research on patenting by new business ventures has focused on its effect on availability of venture capital funding. However, as Moskowitz and Vissing-Jorgensen 2002 point out, venture capital accounts for less than 1% of the private equity market. Therefore, my research on the effects of patenting for entrepreneurs will contribute to the literature by extending patenting research to other alternative sources of capital, such as bank loans and angel investors.
Furthermore, empirical evidence on this topic is largely based on later-staged financing activities. It focuses on already well-recognized, successful startups seeking for additional funding from public equity markets. The role of signaling that patents serve is more difficult to explore by examining later-staged financing rounds. After the startup has received funding from an investor or several investors, it is very likely that information asymmetries have been significantly reduced. Therefore, the impact of existing patents and filing new patents, may not have an effect on attracting additional investor funding, or at least its effect will be greatly diminished. Another problem of exploring later-staged financing is that new patents being filed may be a result of initial investor financing, where the backer influence the startup to pursue patenting as a means of protecting their technological advancements/inventions from potential competitors (Stuart et al. 1999, Hsu and Ziedonis 2008). The focus of my research will be on early-staged financing to try and isolate the role of patents as a signaling mechanism for the startup to obtain funding, from the role of protection. I therefore propose:
Hypothesis 4: The value of patents in reducing information asymmetries will be more pronounced in early funding rounds versus later rounds.
Furthermore, a large number of studies focus on new ventures in industries where patenting is seen as an important requirement. For example, Haeussler et al. (2009) focus on biotechnology firms, Hsu & Ziedonis 2006 focus on semiconductor startups. Moreover, Mansfield 1986, studies a sample of new firms divided into 5 industries (pharmaceuticals, chemicals, petroleum, machinery, and fabricated metal products) where patents are deemed more important and 7 industries (primary metals, electrical equipment, instruments, office equipment, motor vehicles, rubber, and textiles) where patents seem less important. Mansfield’s results indicate that over 80% of patentable inventions in former group of industries were patented, and over 60% of those in latter group of firms were patented (Mansfield, 1986). These findings provide evidence that patenting still has a strong presence in industries that rarely regard patent protection as necessary. This suggests that patents could be widespread in these industries due to their value as a signal to outside investors. Therefore, I will examine the role of patents across different industries to determine whether patents are pursued for different reasons and if their effects of startup funding differ.
Haeussler et al. (2009) find that having filed at least one patent application reduces the time to the first VC investment by 76% (Haeussler et al. 2009). Therefore, there is existing evidence supporting my first hypothesis. Haeussler et al. (2009) focus on British and German biotechnology firms, it is not clear whether their results would still hold in other industries and countries.
Graham and Sichelman report that scholars have found that increased patenting by venture-backed companies is significantly correlated with total investment and total number of financing rounds (Graham & Sichelman, 2008; Baum and Silverman, 2004; Mann and Sager, 2007; Hsu and Ziedonis, 2008). This evidence supports my second hypothesis. Considering that the positive relationship between patenting and investment collected in financing rounds has been reported across many studies and different industries, it would be very surprising if I wouldn’t find evidence that corroborates hypothesis 2.
Lerner shows a positive impact of the patent stock of high-technology companies on VC valuation (Lerner, 1994). Therefore, empirical evidence exists supporting my hypothesis 3.
There is wide range of literature on the topic of patents and new business ventures. However, there are still a lot of areas which require further examination in order to fully comprehend why startups choose between alternative sources of capital and how these financiers value startups patents applications and grants.
Kortum and Lerner examine the influence of venture capital on patented inventions in the US across 20 industries over 3 decades (Kortum & Lerner, 2000). The authors analyze yearly data from 1965 to 1992 for 20 manufacturing industries. Their dependent variable is U.S. patents issued to U.S. investors by industry and date of application, from the USPTO. Their explanatory variables are measures of venture funding collected by Venture Economics and industrial R&D expenditures collected by the U.S. National Science Foundation. Their empirical results suggest that there is a strong association between venture capital and patenting, and that corporate R&D and venture funding have almost identical effects in generating innovations. The authors address the issue of endogeneity, where both venture funding and patenting could be positively related to an unobservable factor, due to the effect of the arrival of technological opportunities. Firstly, they exploit the 1979 Employee Retirement Income Security Act, which allowed pensions to invest in venture capital. R&D expenditures are used as a control variable for the arrival of technological opportunities that are anticipated by economic actors. After addressing the endogeneity concern, their results suggest that VC funding has a strong positive impact on innovation. A dollar of venture capital appears to be about three times more potent in stimulating patenting than a dollar of traditional corporate R&D (Kortum & Lerner, 2000). Furthermore, the authors address concerns that VC may stimulate patenting activity while having no impact on innovation. They compare indicators of patent-quality between 122 VC backed and 408 non-VC backed firms based in Massachusetts. Results indicate that VC-backed firm’s patents are more frequently cited by other patens and engage in frequent and protracted litigation of both patents and trade secrets (Kortum & Lerner, 2000).
Kortum & Lerner provide 2 reasons for entrepreneurs to patent inventions. Firstly, they may fear that once they receive VC funding the investors will exploit their ideas (Patent after VC funding). Secondly, financiers may have difficulties in assessing the true quality of a new venture’s holdings. Therefore, startups apply for patents on valuable technologies to increase the probability of obtaining external financing (Patent before external funding). Kortum & Lerner also examine 3 measures of innovative activity. They use citations per patent as a measure of the economic importance of a patent, Trajtenberg demonstrated a strong relationship between these (Trajenberg, 1990). Secondly and thirdly, they examine the frequency of patent and trade-secret litigation in which the firm has engaged and the extent of the litigation. Models in the law-and-economics literature predict that parties are more likely to litigate and pursue these cases to trial when (i) the potential benefits from the dispute are high relative to the litigation costs or (ii) the outcome of the case is unclear (Gompers and Lerner, 2004). Therefore, litigation is theoretically a good proxy for the economic importance of a patent, this was corroborated empirically by Lanjouw and Schankerman (1997).
Haeussler, Harhoff and Mueller investigate how patent applications and grants held by new business ventures improve their ability to attract VC financing. They argue that investors rely on patents as signals to assess the value of startups. The authors use a survey sample of VC-seeking German British biotechnology companies conducted in 2006 and identify all patents filed at the European Patent Office (EPO). Besides, 162 German and 118 British companies were interviewed face-to-face. They examine the quality of patents, through the number of citations. Their dependent variable is time of first VC financing (dummy variable measured on quarterly basis). Their independent variables are a number of patent related variables, controls for firm characteristics and the number of early stage VC financings as a proxy for supply conditions in the VC financing market. Among the patent related variables, there is the average number of citations, defined as the total number of citations received divided by application stock (cumulative number of patent applications filed at EPO). Furthermore, they compute the number of X-type references (reference that a claimed aspect of invention cannot be considered novel, claim does not deserve patent protection). Also, the share of patent applications that received an opposition.
Haeussler et al. (2009) use a proportional hazard model with time-varying covariates, to estimate the effect of a firm’s patenting activities on the hazard of acquiring VC financing in a specific quarter. The authors find that having filed at least one patent application reduces the time to the first VC investment by 76% (Haeussler et al. 2009). Furthermore, they find that ventures with higher patent quality receive VC funding faster. This finding is important as patent citations largely occur after the VC investment, indicating that investors are capable of distinguishing patent quality before their investment decision. Moreover, they find that opposition by competitors is taken as a positive signal by VCs. The signal “opposition” may be interpreted by the investor as evidence that the company is developing a technology of high commercial value (Gambardella et al. 2008).
The authors provide evidence that signals generated in the patenting process help to reduce informational asymmetries that exist in the process of new ventures seeking external financing. The classical view of patents states that patents foster incentives for innovation, but they do so at the potential blocking of technological developments (Heller & Eisenberg 1998). However, Haeussler et al. (2009), find that patenting stimulates the entry of new business ventures. Furthermore, the authors find that the average firm received VC financing long before the first patent is granted, suggesting that maybe patent applications have a larger influence on VC financing than patent grants. In their interviews, they find that the protection and quality signaling functions of patents are very important for VCs.
Their research has multiple theoretical implications. Kortum & Lerner 2000 observed that VC financed ventures are more active in patenting than non-VC financed companies, however they did not explain whether these results were due to selection or nurturing by VCs. Haeussler et al. (2009) provide evidence in favor of selection of valuable startups with technological capabilities. Furthermore, their results suggest that the patenting process generates valuable quality signals which help start-ups acquire funding.
Hsu and Ziedonis (2006) investigate the extent to which VC investors use information from patents to assess the quality of new ventures. They use a sample of 370 U.S. VC-backed semiconductor startups founded between 1975 and 1999 (Hsu and Ziedonis, 2006). Their dependent variable is startups valuation measured as its pre-money valuation in a financing round. Their independent variables include a cumulative count of the startup’s pending U.S. patent applications, VC experience and founder experience. The authors base VC experience on a Bonacich centrality score of the lead venture capitalist in a funding round, allowing them to classify the importance of investors within the VC community (Hsu and Ziedonis, 2006). Founder experience is a dummy variable representing whether the founding team lacks prior entrepreneurial or IPO experience. As control variables, they employ startup age, round type, prominent corporate affiliations and year of initial financing.
Their main regressions estimate the effect of patent filings on start-ups valuation across funding rounds, holding unobservable time invariant effects constant (Hsu and Ziedonis, 2006). The authors results show a statistically significant and economically sizeable effect of patent filings on investor estimates of startup value. A doubling in the patent application stock of a startup is associated with a 28% increase in valuation, approximately $16.8 million for the average start-up in their sample (Hsu and Ziedonis, 2006). Furthermore, consistent with the signaling value of patents, they find that the effect of patenting on startup valuation is more pronounced in earlier financing round when uncertainty is greatest. However, surprisingly they find no evidence that the signaling value of patents is more pronounced for inexperienced entrepreneurs than for more experienced ones. Finally, they find that the signaling value of patents is higher when funds are secured from more prominent investors (Hsu and Ziedonis, 2006). This finding suggests that startups backed by prominent VCs may have access to superior legal and organizational resources, improving the probability that their patent filings will be of greater economic value. Hallen (2008) noted that landing a prominent VC in an early stage makes it more likely that a prominent VC will invest in future rounds (Hallen, 2008).
Mansfield (1986) explores the extent to which the rate of development and introduction of inventions would decline in the absence of patent protection. Also, the extent to which firms make use of the patent system, and the differences that exist among firms and industries and over time in the propensity to patent (Mansfield, 1986). Mansfield (1986) chooses a random sample of 100 firms from 12 industries in the U.S. From each firm, they obtain an estimate of the proportion of its inventions developed between 1981 and 1983 that would not have been developed if it could not have obtained patent protection. Furthermore, they estimate the proportion of firm’s inventions commercially introduced between 1981 and 1983 that would not have been commercially introduced if it could not have obtained patent protection (Mansfield, 1986). When individual firm estimates are combined into industry-wide estimates, the results indicate that patent protection was deemed essential for the development or introduction of 30% or more inventions in the Pharmaceuticals and Chemicals industries and 10 to 20% for Petroleum, Machinery and Fabricated Metal Products industries (Mansfield, 1986).
Economists seem to believe that patent protection tends to be more important for smaller ventures than larger ones. Mansfield (1986) tests this assertion by correlating a firm’s size (measured by 1982 sales value) with the proportion of its inventions (developed between 1981 and 1983) that would not have been developed if it could not have obtained patent protection. His findings are mixed. In the fabricated metal products industry, the correlation coefficient is negative and statistically significant. In the instruments and machinery industries, the correlation coefficient is positive and statistically significant (Mansfield, 1986).
Furthermore, in most of the industries included in his research, there is a positive correlation between firm’s size (measured by 1982 sales value) and the percentage of its patentable inventions that were patented. In industries like pharmaceuticals and chemicals, where patents seem to be important, over 80 percent of the patentable inventions are patented. Suggesting that firms do not prefer relying on trade secret protection when patent protection is possible (Mansfield, 1986).
Graham et al. (2008) study a comprehensive survey of patenting and entrepreneurship in the U.S., summarizing responses of 1332 early-stage technology companies founded since 1998 (Graham et al. 2009). The authors focus primarily on companies in the biotechnology and software industries. They use Dun & Bradstreet’s data on technology and founding date on listed companies, they combine this data with Thompon’s VentureXpert data covering a substantial share of VC-backed firms in U.S. Combining both data sources, their final list of target firms contained 15,000 unique entities (Graham et al., 2008). They find that VC-backed companies are much more likely to hold and file for patents. Furthermore, from the survey results they find that patents offer mixed to weak incentives to engage in innovation. However, they do suggest that patents support other activities crucial to technology startups, gaining access to necessary investments to develop and grow, increasing the odds and quality of a liquidity event (acquisition or IPO) and serving strategic goals in negotiating and defending against patent infringement suits (Graham et al., 2008). The authors find that entrepreneurs have varied reasons for using the patent system, and that many of the reasons diverge from the traditional theory that patents provide incentives for innovation.
There are 3 major findings from this study. Firstly, patents help startups in technological competition. The author’s results demonstrated that patenting plays an important role for high-technology startups in safeguarding their competitive advantage from their technological innovations (Graham et al., 2008). Secondly, startups have differing motives for patenting. Across all survey respondents who report filing for patents, “the most important reason for patenting is to prevent others from copying the startup’s products and services” (Graham et al., 2008). Next in importance is improving chances of securing investment. Their findings imply that patents play an important role in the financing of many startups, providing evidence to the signaling role of patents. Thirdly, they find that technology entrepreneurs vary in their reasons for not seeking patents. Authors find that, among technology startups, the costs associated with the patenting process are the main reason for not patenting a major technology (Graham et al., 2008).
Nanda and Kerr (2014) review the recent literature on the financing of innovation and provide a comprehensively structured overview. The authors provide reasons for the important frictions that the R&D process introduces, which can lead to financing constraints for firms aiming to undertake R&D. Firstly, the innovation process is inherently uncertain. From the financier’s perspective, this makes it very difficult to distinguish the potential for innovation of different projects requiring funding. Furthermore, this challenge is exacerbated by the extremely skewed returns from the innovation process (Scherer and Harhoff, 2000). In addition, financiers have a first-order effect on innovation since their backgrounds and incentives may lead them to act on information differently (Kerr and Nanda, 2014). Another issue is that firms engaged in innovation have a high percentage of intangible assets, where knowledge is rooted in the firm’s employees (Kerr and Nanda, 2014). This core resource disappears when workers leave or are fired (Hall and Lerner, 2010). This leads firms to smooth their R&D spending over time, to abstain from having to fire their research and knowledge specialists, driving R&D spending at the firm-level to act as though it had high adjustment costs (Kerr and Nanda, 2014).
Furthermore, Nanda and Kerr (2014) highlight the importance of bank finance in innovation. Changes in the availability or costs of bank financing have shown to be related to the rate and nature of innovation by firms. Mann (2014) shows that debt financing is common for innovating firms and that patents are often used as collateral in such instances. He notes that 16% of the aggregate stock of patents at the USPTO has been pledged as collateral at some point, and that companies with patent-backed debt have performed over 40% of USPTO patenting since 2003 (Mann, 2014). Similarly, Hochberg, Serrano and Ziedonis (2014) find patents are used as collateral for venture debt. Chava, Chong and Nanda (2012) also find that firms with significant patent activity and higher-quality patents receive cheaper bank loans.
Nanda and Kerr (2014) also explore the evidence of the effects of institutional ownership on innovation. Aghion, Van Reenen and Zingales (2013) find that greater institutional ownership leads to greater innovation. They find that increased institutional ownership raises the incentives to monitor CEOs and protects them from being fired for bad outcomes unrelated to their management decisions, which promotes innovation.
Finally, Nanda and Kerr (2014) focus on finance and innovation in startups. Ewens and Fons-Rosen (2013) find that individuals who transition from large companies to founding a startup change their patenting behavior to focus on riskier, more innovative patents. Bernstein, Giroud and Townsend (2014) show that one mechanism through which VC investors impact startup innovation is through their role in monitoring, suggesting the important role of governance in innovation. Lerner (1995), Gompers (1995), Hellman and Puri (2000; 2002) and Chemmanur, Krishnan and Nandy (2011) document the important role that VC investment contracts play in overcoming agency issues through staged financing, monitoring, board representation, and replacing founders with professional CEOs in the case of underperforming ventures (Wasserman, 2003). Nanda and Rhodes-Kropf argue that during hot markets— time periods when financing risk is low—projects with the highest real option values are funded because the continuation risk is lower (Nanda, Kerr and Rhodes-Kropf, 2014). Empirically, during hot markets, VC-backed firms have the highest failure rates, but conditional on being successful, these startups have higher-valued exits compared to startups funded in less-active periods (Nanda, Kerr and Rhodes-Kropf, 2014).
In conclusion, there is clear evidence that financing constraints can create considerable consequences for firms engaged in R&D and innovation, shaping both the rate and the trajectory of innovation (Nanda and Kerr, 2014). Bank finance is an important source of finance, particularly for firms with intangible assets (patents) to pledge as collateral. The standard approach to measuring innovations has been looking at the number of patents and patent citations, these metrics have been linked with economic value (Trajtenberg 1990, Kogan et al. 2012). Finally, the clustering of entrepreneurial finance and innovative firms reduces some innovation adjustment costs (Carlino and Kerr, 2014), this finding suggests that startup concentration has an effect on patenting costs. Therefore, the level of entrepreneurship concentration in a startup’s location will be employed to randomly assign startups to different patent-cost groups for my natural experiment.
Lerner 1994 examines the impact of patent scope on firm value. Lerner develops a proxy for patent scope based on the International Patent Classification (IPC) scheme. A sample of 535 financing rounds at 173 privately held venture-backed biotechnology firms is used. Lerner uses a proxy to measure patent scope, the number of subclasses into which the USPTO assigns the patent. To validate the economic importance of the proxy, Lerner analyzes the impact of patent scope on citations through regression analysis. The dependent variable is the number of citations in US patent documents, the independent variable is the number of IPC classes to which the patent was assigned. As another independent variable, the author uses the time from patent award until Dec 22 1992, to control for the differing exposure periods of time that these patents have had to garner citations. The coefficient of the patent scope variables is highly significant; a one standard deviation increase in patent scope increases the expected number of citations per year by 11%. This finding is valuable since Trajtenberg (1990) showed that the number of citations reflects the economic importance of a patent.
In addition, Lerner examines whether a patent has been subject to litigation. Since litigation is costly and time-consuming, Lerner assumes that most firms will only litigate economically important patents. Patent litigation is identified through two sources. Firstly, the American Intellectual Property Law Association periodically compiles a listing of all pending and terminated litigation involving biotechnology patents. Secondly, this list is manually updated and supplemented with news stories published in BioWorld Today, a biotechnology news service. Lerner proceeds with a probit regression, where the dependent variable is a dummy variable which is equal to 1 if a patent has been subject to litigation through May 31 1993. As the independent variable, Lerner uses his proxy for patent scope. As another independent variable, the author uses the time from patent award until May 31 1993, again to control for the differing exposure periods. The regression results indicate that broader patents are significantly more likely to have been subject to litigation. A one standard deviation increase in patent scope increases the probability of litigation by 41% (Lerner, 1994). Furthermore, Lerner finds that increasing scope of patent protection is associated with higher firm valuations.
Data & Methodology: Hypothesis and statistical tests
Hypothesis 1: The amount of successful patent applications by a startup will be negatively related with the time taken to acquire external funding.
The null hypothesis – that patent applications is not related to time taken to acquire external funding will be tested against the alternative hypothesis – that it is negatively related to time taken. The following equation will be estimated across our sample of firms:
TIME TO FUNDINGi=αi+β1*#PATENTSi+β2*#CITATIONSi+β3*STARTUP SIZEi+β4*PRIOR EXPERIENCEi+β5*PERSONAL CONSUMPTIONi+εi
The key dependent variable is TIME TO FUNDING, the time to obtain external funding collected from Crunchbase, calculated as the difference between the date the firm was founded and date at which first funding was received. The independent variables are, #PATENTS, which is the number of patent applications at time of funding, and #CITATIONS the number of citations at time of funding obtained from USPTO and the European Patent Office (EPO).
Furthermore, a number of control variables will be employed to isolate other effects which may lead to different durations in accessing external funding. STARTUP SIZE is the number of employees that the startup has at the time it receives external financing, which aims to control for venture size and startup’s life stage (Hsu & Ziedonis, 2008). PRIOR EXPERIENCE is a dummy variable which will take the value =1 when a member of the startup had previously founded another startup (=0, otherwise). PERSONAL CONSUMPTION is a variable measuring the personal consumption expenditures collected by the U.S. Bureau of Economic Analysis. This will be used as a control variable for the overall economic market conditions.
The test of H1 is whether
β2are < 0, that patent applications and patent citations lead to a decrease in time taken to access funding.
Hypothesis 2: The amount of successful patent applications by a startup will be positively related with the amount of financing it raises in the first financing round.
The null hypothesis – that patent applications is not related to the amount of funding raised in the first financing round – will be tested against the alternative hypothesis – that it is positively related to amount of funding raised. The following equation will be estimated across our sample of firms:
FUNDING AMOUNTi=αi+β1*#PATENTSi+β2*#CITATIONSi+β3*STARTUP AGEi+β4*PRIOR EXPERIENCEi+β5*PERSONAL CONSUMPTIONi+β6*INVESTOR EXPERIENCEi+β7*YEARi+εi
The dependent variable, FUNDING AMOUNT is the total amount of financing raised in a round, collected from Crunchbase. The independent variables are, #PATENTS, which is the number of patent applications at time of funding, and #CITATIONS the number of citations at time of funding obtained from USPTO and the European Patent Office (EPO).
As in H1 a number of control variables will be employed. The variables PRIOR EXPERIENCE and PERSONAL CONSUMPTION will be included again. Additionally, STARTUP AGE, measured at the time when the startup receives financing is added as a control variable. Moreover, INVESTOR EXPERIENCE, the financier’s experience measured as the difference between the year financier made first investment and the year it invested in the startup is added as a control variable. Finally, YEAR, the year when the investment was made will be included to control for periods of expansion vs. recession which will likely affect the supply of financing.
The test of H2 is whether
β2are >0, that patent applications and patent citations lead to an increase in the amount of funds raised in financing rounds.
Hypothesis 3: The amount of successful patent applications by a startup will lead to higher valuation at financing rounds, time of initial public offering and acquisition.
The null hypothesis – that patent applications is not related to the valuation of a startup (in funding round or at IPO/acquisition)– will be tested against the alternative hypothesis – that it is positively related to the startup’s valuation. The following equation will be estimated across our sample of firms:
VALUATIONi =αi+β1*#PATENTSi+β2*#CITATIONSi+β3*STARTUP AGEi+β4*PRIOR EXPERIENCEi+β5*PERSONAL CONSUMPTIONi+β6*INVESTOR EXPERIENCEi+β7*YEARi + εi
The dependent variable, VALUATION, is the total value assigned to a startup at the financing round obtained from Crunchbase. Furthermore, IPO value will be retrieved from CRSP and acquisition value acquired from SEC filings.
The test of H3 is whether
β2are >0, that patent applications and patent citations lead to higher valuation at financing rounds and exits (IPO or acquisition).
Data sources discussion
The main data source that I will use for data on startups and financing rounds will be Crunchbase. They provide international data on startups. Extensive information for each startup regarding its country of origin, industry it operates in, number of employees, date it was founded, is accessible. Furthermore, they have detailed information on different investment rounds and on different types of investors (venture capital, debt financing, crowdfund, university program). Finally, they also have data on startups which were acquired or had a successful IPO.
The main problem with the Crunchbase data is that startups are only mentioned by name, they do not have a unique numerical identifier that can be used to search for patents in USPTO and EPO. Therefore, the most time-consuming and complicated process is going to be matching the startups in Crunchbase to their patent applications through USPTO and EPO. An issue, which will be encountered when attempting to find each entity’s patents, is that firm’s unique GVKEYs can change over time due to ownership structure changes or accounting changes. Therefore, a single startup may have multiple different entries in the USPTO data under different GVKEYs.
Furthermore, Kogan and Stoffman (2012) made their data (1926-2010) available online. This data contains patent level data: patent number, filing date, grant date, publication date, CRSP PERNMO and number of citations. Also, it has firm-level innovation data: CRSP PERNMO, year, number of patents. However, their research doesn’t solely focus on startups, thus their data will be analyzed to assess the effect that patenting rates and patent quality have on firm’s accessing external financing. While the data is not solely focused on startups, it would be interesting to compare the effects of patenting for mature firms and startups.
Finally, the National Bureau of Economic Research (NBER) made their data available to facilitate matching patent data and Compustat data. They introduce a unique identifier PDPCO which includes multiple GVKEYs if the entity was reorganized (Bessen, 2009). Furthermore, since patent owners change over time, the NBER records dynamic changes in ownership. This data also does not solely focus on startups, however due to the limited data available for startups it will be a useful source to explore the effects of patenting. They do filter their patent data creating a variable “COD” which identifies assignee type. The different COD classifications identify patent assignees like US corporation, Foreign corporation, US individual and foreign individual (Bessen, 2009). Focusing on individuals holding patents may yield interesting insights, since they are more likely to be part of a startup or of a firm in earlier stages of its life cycle.
Throughout this paper, patent data will be used to explore the effects that patenting has on startup’s access to external financing. Patents are legal rights which exclude other from making, using, or selling the patented invention or process for a certain time-period (Carlino & Kerr, 2014). Therefore, patenting should lead to new revenue streams for the startup, through selling the aforementioned product or process without (or with limited) competition, by selling the right to production to a third-party or entering into royalty agreements. Securing these revenue streams will lead to higher economic valuations of startups that get a patent approved. External financiers (e.g., venture capitalists, angel investors) are constantly looking for new investment opportunities, screening the population of new ventures through a set of factors which they deem important for future value creation. Kaplan and Strömberg (2000), look at how VCs select and screen-for investments. They find that VCs look, among others, at the management team, the technology, and the competition (Kaplan & Strömberg, 2000). VCs are likely to continually monitor patent applications and grants to assess technological capabilities and the degree of competition.
Since patent grants, on average (as of February 2017) take a minimum of 25 months from application date, we use patent applications instead. VCs and other investors are knowledgeable so we assume that they can assess the quality of a patent shortly (in far less than 25 months) after a firm submits the patent application. Therefore, we will make use of patent application data which, most likely, will provide better insights into the effect of patenting on access to external financing.
A concern with using patent data is that the value of patents is highly skewed. Most patents are worth very little, while only a small number are very valuable (Scherer and Harhoff, 2000; Harhoff et al., 1999). This raises the concern that maybe VCs and other external financiers don’t focus on patent data for investment-selection since most are of little value. However, since there is a vast difference between the economic impact of different patents external financiers are likely to thoroughly assess patenting activity to find the limited number of attractive startups that have truly valuable patents.
Furthermore, the number of citations that a patent receives will be used alongside the number of patents that an entity applied for, to distinguish between less and more valuable patents. Researchers (e.g., Haeussler et al. 2009; Trajtenberg, 1990) make use of citation data to assess patent quality.
Another issue with patenting data is the large difference in tendency to patent across different industries (Carlino & Kerr, 2014). Most research is focused on industries with high patenting rates, like pharmaceuticals, software and chemicals. It is unclear whether patenting data on industries with low patenting rates will provide valuable insights on the funding process of startups.
In an attempt to infer causality, on the effect of patents and financing, I will conduct a natural experiment. In this experiment, I will use distance to a patent office as a control condition, leading to a random assignment of firms that can more easily/economically patent (close to patent office) against firms that find it more expensive (further to patent office).
I will then conduct my hypothesis again but separately for each control group. If the group of firms that is closer to patent offices receives funding earlier, and in greater amount, than the other group that is further away; the experiment would provide insights about the role that patenting plays in acquiring finance for new ventures.
Furthermore, considering the findings of Carlino and Kerr (2014), that startup concentration has an effect on patenting costs, I will also employ degree of startup concentration as a control condition to randomly assign firms to more or less expensive patenting groups.
Plan for remainder:
Summary statistics of the data will be displayed. Statistics such as:
- Average financing raised in a round
- Average number of financing rounds
- Average number of patents
- Average number of citations
- Regression results will be reported
- Assess the results and whether they are significant or not
Natural experiment and results
- Were results as expected?
- Is there significant difference in results across different industries?
- Any econometrics issue with data/tests
- Further developments to expand research in the topic
Aghion, P., Van Reenen, J., & Zingales, L. (2013). Innovation and institutional ownership. American Economic Review, 103(1), 277-304.
Baum, J. and Silverman, B. (2004). Picking winners or building them? Alliance, intellectual, and human capital as selection criteria in venture financing and performance of biotechnology startups. Journal of Business Venturing, 19(3), pp.411-436.
Bernstein, S., Giroud, X., & Townsend, R. (2014). The impact of venture capital monitoring: Evidence from a natural experiment. Working Paper.
Bessen, J. (2009). Matching Patent Data to Compustat Firms. NBER PDP Project. https://sites.google.com/site/patentdataproject/Home/downloads.
Birch D. 1990. Sources of job growth and some implications. In: Kasarda, J. (Ed.), Jobs Earnings, and Employment Growth Policies in the United States. Kluwer Academic Publishers, Norwell, MA, pp. 71 – 76.
Carlino, G., & Kerr, W.R. (2014). Agglomeration and innovation. Harvard Business School Working Paper No. 15-007.
Chava, S., Chong, X., & Nanda, V. (2012). Funding innovation: The role of lender expertise and control rights. Georgia Institute of Technology Working Paper
Cochrane, J. 2005. The risk and return of venture capital. Journal of Financial Economics, 75(1), pp.3 52.
Conti, A., Thursby, J. and Thursby, M. 2013. Patents as Signals for Startup Financing. The Journal of Industrial Economics, 61(3), pp.592-622.
Cooter, R.D. and Rubinfeld, D.L. 1989. Economic Analysis of Legal Disputes and Their Resolution. Journalof Economic Literature, Vol. 27, pp. 1067-1097
Denis, D. 2004. Entrepreneurial finance: an overview of the issues and evidence. Journal of Corporate Finance, 10(2), pp.301-326.
Ewens, M., & Fons-Rosen, C. (2013). The consequences of entrepreneurial firm founding on innovation. Working paper.
Fairlie, R. 2015. Kauffman Index of Entrepreneurial Activity: 2015. SSRN Electronic Journal.
Gambardella, A., D. Harhoff, B. Verspagen 2008. ”The Value of European Patents,” European Management Review, 5, 69–84.
Gompers, P. and Lerner, J. (2004). The venture capital cycle. 2nd ed. Cambridge, Mass.: MIT Press.
Graham, S. and Sichelman, T. 2008. Why do Start-ups Patent?. Berkeley Technology Law Journal, [online] 23(3). Available at: http://scholarship.law.berkeley.edu/cgi/viewcontent.cgi?article=1757&context=btlj [Accessed 22 Feb. 2017].
Graham, S., Merges, R., Samuelson, P. and Sichelman, T. 2008. High Technology Entrepreneurs and the Patent System: Results of the 2008 Berkeley Patent Survey. SSRN Electronic Journal.
Hall, B.H., Griliches, Z., &. Hausman, J.A. 1986. Patents and R&D: Is there a lag?. International Economic Review, 27(2), 265-283.
Hall, B., & Lerner, J. (2010). The financing of R&D and innovation. In Handbook of the Economics of Innovation, eds. Hall, B., & Rosenberg. N., Netherlands: Elsevier, 609-639.
Hallen, B. and Eisenhardt, K. 2008. How Entrepreneurs Form Relationship with other Organizations: The Securing of Investments From Venture Capital and Corporate Investors. Academy of Management Proceedings, 2008(1), pp.1-6.
Harhoff, D., Narin, F., Scherer, F., Vopel, K. (1999). “Citation frequency and the value of patented inventions.” Review of Economics and Statistics, 81, 511-15.
Häussler, C., D. Harhoff, and E. Muüller. 2012. “To Be Financed or Not . . . The Role of Patents for Venture Capital Financing.” SSRN Working paper.http://papers.ssrn.com/sol3/papers.cfm?abstract id=1393725.
Heller, M.A., R.S. Eisenberg. 1998. Can patents deter innovation? The anticommons in biomedical research. Science 280 698-701.
Hochberg, Y., Ljungqvist, A. and Lu, Y. 2007. Whom You Know Matters: Venture Capital Networks and Investment Performance. The Journal of Finance, 62(1), pp.251-301.
Hochberg, Y., Serrano, C., & Ziedonis, R. 2014. Patent collateral, investor commitment, and the market for venture lending. NBER Working Paper No. 20587
Hsu, D. and Ziedonis, R. 2008. Patents as Quality Signals for Entrepreneurial Ventures. Academy of Management Proceedings, 2008(1), pp.1-6.
Kaplan, S. and Strömberg, P. 2000. How Do Venture Capitalists Choose Investments? Graduate School ofBusiness, University of Chicago.
Kogan, L., Papanikolaou, D., Seru, A., & Stoffman, N. (2012). Technological innovation, resource allocation, and growth. NBER Working Paper No. 17769
Kortum, S. and Lerner, J. (2000). Assessing the Contribution of Venture Capital to Innovation. The RAND Journal of Economics, 31(4), p.674.
Lanjouw, J. and Schankerman, M. 1997 Stylized Facts of Patent Litigation: Value, Scope and Ownership.” Working Paper no. 6297. National Bureau of Economic Research
Lemley, M.A. 2000. Reconceiving patents in the age of venture capital. Journal of Small and Emerging Business Law 4 137-148
Lerner, J. (1994). The Importance of Patent Scope: An Empirical Analysis. The RAND Journal of Economics, 25(2), p.319.
Lerner, J., & Seru, A. (2014). The use and abuse of patent data. Working Paper.
Long, C. 2002. Patent signals. University of Chicago Law Review 69 625-679.
Mann, W. (2014). Creditor rights and innovation: Evidence from patent collateral. Working paper.
Mann, R. and Sager, T. (2007). Patents, venture capital, and software start-ups. Research Policy, 36(2), pp.193-208.
Mansfield, E. 1986. Patents and Innovation: An Empirical Study. Management Science, 32(2), pp.173-181.
Moskowitz, T., Vissing-Jorgensen, A., 2002. The returns to entrepreneurial investment: a private equity premium puzzle? American Economic Review 92, 745 – 778.
Nanda, R., & Kerr, W. (2014). Financing Innovation. Harvard Business School Working Paper No. 15-034.
Nanda, R., Kerr, W., & Rhodes-Kropf, M. (2014). Entrepreneurship as experimentation. Harvard Business School Working Paper No. 15-005
Nanda, R., & Rhodes-Kropf, M. (2010). Financing risk and innovation. Harvard Business School Working Paper No. 11-013.
Nanda, R., & Rhodes-Kropf, M. (2013). Investment cycles and startup innovation. Journal of Financial Economics, 110(2), 403-418.
Nanda, R., & Rhodes-Kropf, M. (2012). Innovation and the financial guillotine. Harvard Business School Working Paper No. 13-038.
Robb, A. and Robinson, D. 2012. The Capital Structure Decisions of New Firms. SSRN Electronic Journal.
Scherer, F.M., & Harhoff, D. (2000). Technology policy for a world of skew-distributed outcomes. Research Policy, 29(4–5), 559-566.
Spence, M. 1973. Job market signaling. Quarterly Journal of Economics 87 355-374
Stuart, T. E., H. Hoang, R.C. Hybels, R. C. 1999. Interorganizational Endorsements and the Performance of Entrepreneurial Ventures. Administrative Science Quarterly 44 315-349.
Trajtenberg, M. (1990). A Penny for Your Quotes: Patent Citations and the Value of Innovations. The RAND Journal of Economics, 21(1), p.172.
Wasserman, N. (2003). Founder-CEO succession and the paradox of entrepreneurial success. Organization Science, 14, 2149-2172
Cite This Work
To export a reference to this article please select a referencing stye below: