1.
STANFORD UNIVERSITY AND HIGH-TECH ENTREPRENEURSHIP: AN EMPIRICAL STUDY Hervé Lebret, Ecole Polytechnique Fédérale de Lausanne, Switzerland ABSTRACTThis study examines more than 2’700 companies founded by alumni of Stanford University orhaving licensed a technology from this university. Stanford University is with MIT one of themost entrepreneurial university in the world, and surprisingly not much data is available on itsspin-offs and start-ups. Some important features are described such as the use of venture capital,the dynamics of growth and exits through acquisition or initial public offering. Somecharacteristics of the founders are also considered such as the time lag between their academicactivity and the start-up creation as well as the characteristics of serial entrepreneurs. INTRODUCTION Academic entrepreneurship as well as the role of universities in high-tech entrepreneurshipthrough their alumni has been a much-studied topic in the recent past. Two extensive studies(Shane, 2004 and Djokovic & Souitaris, 2008) illustrate the amount of work done recently. Manyof these analyses (Shane, 2004; Roberts, 1991; Hsu et al. 2007; Roberts & Eesley, 2009) werefocused on the Massachusetts Institute of Technology (MIT). Other authors (Saxenian, 1994;Zhang, 2003, 2009) have compared the Boston Area and Silicon Valley in particular through theangle of venture capital funding and have shown the critical role of both MIT and StanfordUniversity in academic entrepreneurship. It would be impossible to make here a list of all paperspublished on the topic and Djokovic has done a very interesting compilation of papers studyingspinouts from academic institutions. Another synthesis summarizing lessons learnt on universitiesand start-ups (Lerner, 2005) was also published after many articles on the topics related to spin-offs and venture capital. Whereas Silicon Valley has been extensively studied (Saxenian 1994, 1999; Kenney, 2000;Lee et al., 2000), it appears that Stanford University has not been studied as much as MIT or manyother universities, which have been much less entrepreneurial than Stanford. Here can bementioned the cases of UT-Austin (Smilor, 1990), the University of Cambridge in the UK(Garnsey & Heffernan, 2005), Oxford University (Lawton Smith & Ho, 2006), ETH Zurich(Oskarsson & Schläpfer, 2008) or the broader subject of universities and venture capital (Zhang,

2.
2009). Stanford has been studied through the limited cases of departments or laboratories (Kenney& Goe, 2004; Jong, 2006, Lebret, 2007) including some unpublished work (Lenoir, 2002). Abroader view of Stanford University and its connection with the military-industrial complex is thebook by Lowen (1997). Stanford remains however some lesser-known territory that deserves moreattention.Start-up and Spin-off The definition of a spin-off has also been the subject of many studies and the debate may notbe totally closed. Djokovic and Zhang among others have shown the variety of definitions used.They usually include the transfer of technology and/or people but the definition of transfer oftechnology may be formal (through a contract or license process) or informal. A good example ofthe difficulty is the famous example of Google vs. Yahoo at Stanford University (Ku, 2002).Yahoo was not considered as a Stanford spin-off because the two founders built the web site as ahobby on their spare time so that no license was needed from Stanford when Yahoo was movedout of the laboratory to a stand-alone company. Stanford had filed a patent on the PageRanktechnology, which was licensed to Google when the company was incorporated. Fundamentally,however there was the same transfer of people and there was a similar transfer of technology evenif it was not formalized through a patent application in the Yahoo case. If the definition of a spin-off looks quite clear and simple, it does not mean that Yahoo was not created thanks to theuniversity facilities and (cultural) ecosystem. This is another motivation for studying not onlyStanford spin-offs but also the related start-ups founded by alumni without a license fromStanford.Venture Capital and Founders of Start-Ups Why are start-ups scrutinized so much? One important reason is certainly the value creation ofthese companies in the last fifty years. From Intel to Google, and in between companies such asGenentech, Apple, Microsoft, Oracle, Cisco, the American economy has positively benefited fromthese fast growing companies. Billions of dollars of sales and hundreds of thousands of jobs havebeen created by a relatively modest number of companies in a very short period of time. SiliconValley and its informal ecosystem have both been at the origin and the beneficiaries of this valuecreation that very few other regions on the planet have experienced. Some interesting and rather unique characteristics of these companies have also explained thisattention. Venture capital has been a critical tool for the growth of these companies and it hasbecome a structured activity in parallel to the development of the Silicon Valley and Bostontechnology clusters and in particular their start-ups. However, even if venture capital has beenextensively studied, the author is not aware of studies that link start-ups and venture capital in asystematic manner. What about start-ups which do not use venture capital? Do venture-backedcompanies succeed better than others? The founders of start-ups are also critical. The names of the founders of the companiesmentioned above are all famous. Why is this? There are certainly elements of leadership andcharisma with start-up founders which may be much more important when companies are smalland fast-growing. More importantly, these founders have become role models for the newentrepreneurs. Steve Jobs had Robert Noyce as a mentor, Brin and Page met Andy Grove.Founders are important outside their own companies.

3.
This article does not have the ambition to consider hypotheses that could be validated or notbut has the objective of illustrating some features of start-ups (that the author believes areimportant and even if somehow quite well-known, not always described with facts and figures). Italso has the ambition of showing that some of these features might be used as success or growthmeasures of start-ups: these are venture capital resources, value creation, time-to-exit for example. DATA AND RESULTS This paper studies three different groups of start-ups linked to Stanford University. The firstone is the group of start-ups which obtained a license from the Office of Technology Licensing(OTL) of Stanford University (called the “spin-offs”). The second one is based on a studycommissioned by OTL (Leone et al. 1992). The third group known as the Wellspring ofInnovation (http://www.stanford.edu/group/wellspring) is a list of companies founded by StanfordAlumni and was retrieved on February 6, 2009 (this web site is an ongoing project). This makes atotal of more than 2’700 start-ups. Essentially the names of the companies and Stanford founderswere available in these lists. We have empirically built consistent data over the three groups: thefields of activities, the resources provided by venture capital and other investors, the year offoundation, the year of a liquidity event if any (Initial Public Offering – IPO, Trade Sale – M&Aor Cessation of Activity). The value creation is also studied in three ways: the sales, theemployment and the value creation (market capitalization) when the company is public or thevalue of the M&A if the company was acquired. The time span between activity at Stanford andcreation of the start-up, and between creation and liquidity of the start-up has been studied. Thelast but related features are linked to the founders: are these serial entrepreneurs? What about theirpast experience before becoming founders. What about the role of professors as founders?Value creation For the sake of efficiency and space available in this article, we will compare in this first partof our results the spin-offs and a subset of the Wellspring of Innovation (“WI”). StanfordUniversity generated 204 spin-offs (see Table 1). The number of start-ups in WI which did notbelong to the two other groups is 2’140. Out of these, 1’467 can be considered as high-techcompanies as the creators of the WI list also included entrepreneurs in non technical activities suchas consulting, finance. Table 1 indicates the fields of activities of the companies and the number ofVC-backed companies. Biotechnologies and medical devices (“life sciences or LS”) representabout 50% of the spin-offs and information technologies (“IT”) about 45%; however for the WIgroup, once the non high-tech companies are excluded, LS account for 15% and IT for 85%. Asecond important comment is that in high-tech, and in both groups, about 50% of the companiesare venture-backed. More precisely, since 1985, 4 out of the spin-offs founded each year raised onaverage $30 million during their lifetime. In the WI group, 26 start-ups founded per year raised$41 million each. Table 2 summarizes the value creation of these start-ups. It also includes thethird group not mentioned until now. The group of spin-offs raised a total of $2.9 billion ofventure-capital. The acquisitions represent a cumulative value of $8.2 billion and the value ofpublic companies as of October, 3, 2009 was $22.4 billion, excluding Cisco ($131 billion) andGoogle ($153 billion). The WI group in its high-tech part had respectively $27 billion of venture-capital, $173 billion of M&A and $183 billion of public value.Dynamics of growth The numbers shown in Tables 1 and 2 are not fundamentally new. The value creation of high-tech start-ups is well-known, even if it may have not been linked to Stanford in such a systematic

4.
manner. A related feature of this value creation is the speed at which the value is created. Theauthor tried to systematically look for the year of incorporation of the companies as well as thetime of exit, if any, i.e. the year of an acquisition, of an initial public offering or of a liquidation.Many studies focus on the survival rate of start-ups (e.g. Oskarsson & Schläpfer, 2008) butTable 3 shows that the dynamics of exits may be much more relevant. Only about a third of thecompanies remain privately-held and much less (16%) in the VC-backed companies of the WIgroup. Another third has been acquired (55% of the VC-backed in WI). A smaller group (5 to15%) is public and the difference is made of liquidated companies. Figures 1 and 2 show the timeto liquidity in years of the companies which are not privately held anymore. For the spin-off case(Figure 1), the number of companies is 61 for the VC-backed ones and 31 for the others. Theaverage number of years to liquidity is 5.97 for start-ups with VC money and 6.55 for the others,the overall average being 6.17 years. The WI group (Figure 2) has 630 VC-backed companies and574 start-ups in the second group. The average time to liquidity is 5.3 years for the first subgroupand 8 years for the other one, with an overall average of 6.6 years. One clear difference betweenspin-offs and start-ups is the impact of non-tech companies. First, as Figure 4 shows, 55% of theWI non-tech companies are still private (vs. 21% of the WI high-tech ones). Secondly, the averagetime to liquidity for non-tech companies is 10 years.Founders Founders are a critical component of companies. However, no formal definition exists.Founders should not be confused with entrepreneurs who usually work in the companies they startnor with managers who may not be part of the founding team, even if some were early employees.The only simple definition of the group of founders is the group of people who recognizethemselves as such. The author could identify 2’711 unique names of individuals for the 2’727companies (Table 4). However, for 62 companies, no founder could be identified as a member ofthe Stanford community (i.e. professor, staff or alumnus). 2’203 companies had one Stanfordfounder and 462 had more than one. Professors are active founders: 167 unique professors werefounders in 243 companies. In 140 of these companies, they were the only Stanford founder. Only82 of these companies had a license and therefore belong to the spin-off group. The experience of founders is one of their key features. It is usually illustrated by their priorprofessional activity or their age. The author did not have access to such information. However,thanks to data available through the Stanford Alumni association, it was possible to find when afounder graduated from Stanford University and therefore obtain the number of years between theactivity at Stanford (graduation year if an alumnus or latest date of activity if a professor or staff)and the year of incorporation of a company. In the case of the spin-offs however no informationcould be obtained for 42 out of 204 companies (some of these companies do not have Stanfordfounders) and 163 companies had missing information out of the 2’523 other start-ups. When acompany had several founders, the time difference was taken as the year of foundation minus theaverage of the activity years of all founders. Figure 4 is the histogram of these time differences forthe spin-offs. Even if the average value is 2 years, it appears clearly that a majority of spin-offs iscreated by individuals active at Stanford at the time of incorporation. Figure 5 gives theinformation for the start-ups, i.e. the WI and the 1992 study. The average is 9.2 years. Figure 5shows two interesting features: first, nearly 250 companies (a little under 10% of the group) werecreated at year 0; the numbers decrease immediately around 100 for years 1-3 but increase againfor years 4-6, then decrease smoothly thereafter. Figure 6 shows the three groups (spin-offs andstart-ups) by field of activities. High-tech companies are created in the range of 6-8 years (e.g. 5.7for biotech, around 7 for medical technologies or “medtech”, semiconductor, IT, software and

5.
Internet) where as non-technical are above 10 years (about 12 for finance and non-technicalservices). Serial entrepreneurship (founders in our study) is an interesting topic as it is regularlymentioned in the general press. Many investors claim they favor entrepreneurs with experience,even if they have failed in their prior venture. The literature is surprisingly not very rich and tworecent works (Bengtsson, 2008 and Gompers et al., 2009) focus on venture-backed companies butseem to reach slightly different conclusions. This article does not focus on serial entrepreneurshiponly and a dedicated article is under preparation. However, it is of interest to show some results.Among the 2’711 founders, 445 individuals (including 44 professors) created more than onecompany. The total number of companies launched by these serial founders is 988 (somecompanies have several serial entrepreneurs). Table 5 compares the resources and value creationof companies which did not have serial entrepreneurs, as well as the first, second, third and fourthcompanies created by the serial founders. The results tend to show that serial entrepreneurs havemore resources with their new ventures in terms of venture capital money but do not create onaverage more value with their new companies than with the prior ones or compared to one-timeentrepreneurs. The only exception is the M&A value of the second ones compared to that of one-time entrepreneurs. DISCUSSION AND CONCLUSION The results of the analysis of Stanford high-tech entrepreneurship are manifold. The valuecreation is exceptionally high thanks, in part, to a very high level of venture-capital money.Venture capital does not explain alone this success. Hewlett-Packard belongs to the 1992 study, itis the largest of the companies considered in the three groups and was not financed by venturecapital. However the levels of money raised (on average more than $30 million for the companiesaccessing VC money) are high and should be an indicator for many academic spin-offs of thelevels of resources used to succeed. The value creation is also extremely high. Table 2 shows theamounts of sales and employment generated by the existing public companies. In high-tech, thesales were close to $350 billion in 2008 and the employment was close to 1 million jobs.Nevertheless the top 5 tech companies generated about 2/3 of the value creation and the top 10,about 3/4. Many smaller companies contribute to this creation but are not small companies even ifthey are called start-ups. The resources used by these top 5 and top 10 companies in terms ofventure-capital are relatively much smaller. Finally, a total of 1’050 companies could be identifiedas VC-backed out of the 2’727. The value creation of this subset is $186 billion in M&A and $543billion in public value (against $82 billion of M&A and $286 billion for those which were notassociated with venture capital). Table 1 also shows which fields are most developed by the spin-offs and financed by venture-capital: they are the same. Indeed life sciences and information technologies represent 99% of thespin-offs as well as the companies backed by venture capital both in the spin-off and WI groups.An interesting feature of spin-offs is the very high share of life sciences, more than 50% of thespin-offs and the VC-backed companies whereas they represent less than 15% of the WI group!There is no doubt that intellectual property licensed from universities play a role in this difference.Many studies (e.g. Oskarsson & Schläpfer, 2008; Roberts and Eesley, 2009) do not show the samefield repartitions. This feature may explain why some universities do not experience the sameratios of fast growing companies. Growth is not measured only by the available resources. Time to exit is another such measure.Whereas some studies emphasize the survival rates of start-ups as a measure of success, this article

6.
shows that fast growth is a general feature of Stanford high-tech companies. More surprisinglymaybe, biotech companies do not show longer time to exits. This apparent mystery may be easilyexplained by the fact that many biotech companies go public without any sale at an early stage oftheir development. However even if some non VC-backed companies are slower to exit (medtechor electronics), others are as fast with or without VC Money (software and biotech) as Figures 1and 2 show it. As interesting to illustrate the fast growth is the number of privately heldcompanies. A note of caution is necessary: the spin-off and WI groups are different in nature;spin-offs include companies founded until 2008 whereas the WI group includes companiesfounded before 2005, therefore the dynamics of exits are obviously different for the most recentcompanies. Furthermore, the WI group includes non-tech companies which may survive moreeasily and longer with revenues generated from their customers in sectors such as finance or non-tech services (Figure 4). Therefore even if the explanations for the number of private companiesmay be diverse, the numbers are very low in both cases. High value creation and fast growththanks to adequate resources appear clearly. These results should not be surprising. The contribution of Silicon Valley to the Americaneconomy is well documented but what may have been often neglected is how much a universitymay directly (licenses) or indirectly (alumni) contribute to a dynamic ecosystem. Similarly to MITfor the Boston Area, Stanford is a major contributor to Silicon Valley, but it less clear that otheruniversities or clusters have been as successful. The correlations between lesser entrepreneurialregions and their university spin-off success might be similar. In a much smaller-scale study, itwas described how some US start-ups went public five years on average after their incorporation,whereas it took ten years for European start-ups to become public (Lebret, 2007). The “classical 5to 7 years” that venture capitalists look for as a holding horizon when they invest in start-upscould also explain the results. In terms of recommendations for metrics and benchmarks ofacademic innovation, the dynamics of growth and value creation should therefore be used and notonly quantitative measures such as company creation or patenting and licensing activities. Thelevel of resources used by start-ups may be an indication that many non-US start-ups areundercapitalized to succeed. The characteristics of the founders represent the second part of this study. One key feature isthe founders’ experience after leaving or not Stanford University. A little less than 340 companieswere created at year 0, another 360 companies were created between year 1 and year 3 of theaverage activities at Stanford (average on all founders) and 390 between years 4 and 6. Companiescreated at year 0 (12% of total) represent about 50% of the value of public companies, 25% of theM&A value and 28% of the VC money. The second group (years 1-3) and third group (years 4-6)represent respectively 13% and 14% of the number of start-ups, 6% and 20% of the public value,13% and 20% of the M&A value, and 12% and 15% of the VC money. Though these commentswould require a deeper analysis, companies created at year 0 seem to create much more value thanothers. Experience may not matter so much as possibly the quality of the technologies. This couldbe correlated to the young age of many extremely successful entrepreneurs in Silicon Valley.Experience of founders may not be a fundamental requirement. In terms of research on technologytransfer, the comparison of Figures 4 and 5 indicates that formalizing the spin-off creation throughlicensing may be a necessary process but probably insufficient in describing the value creation ofuniversities. Though difficult to quantify, it is likely that many start-ups created at year 0 inFigure 5 include companies which could have been considered as spin-off similarly to the Yahoo-Google analogy considered in our introduction. A final and interesting feature is the serial entrepreneurship factor. Whereas the generalagreement seems to claim that serial founders would be important because of the experience they

7.
bring on board of start-ups, the results of this study does not seem to confirm this general belief.What is also interesting is that serial entrepreneurs are more successful with their first companiesthan one-time entrepreneurs and on average they raise less VC money. However the situation isinverted with the following ones with two exceptions: the average M&A value of the 2nd one isstill higher than that of one-time entrepreneurs and the VC money raised by the 4th ones is smaller!It is also worth noticing as described in the previous paragraph that most of the value creationseems to be linked to the time proximity to Stanford. Some explanations have been given thatserial entrepreneurs could be over-optimistic and less motivated. One other explanation might bethat more disruptive (and therefore promising) technologies belong to companies close toStanford. One cannot avoid thinking that the luck factor may have an important role in high-techentrepreneurship. The value creation would therefore be explained by the statistical effect of thelarge number of ventures. As a conclusion, we would like to propose some areas for future research and also mentionsome limitations and difficulties linked to the data. With the exception of public companies whichdisclose an enormous and rich amount of information in their SEC documents and in particular intheir IPO prospectus, high-tech start-ups do not disclose much information. It is the author’sexperience that information is not only difficult to find for private companies but it is alsosometimes doubtful. Money raised through venture capital, value of M&A transaction should betreated with some caution. A company may announce numbers which are not always accurate(because of milestones-based financing as an example). The revenue and employment figures wereprovided on the basis on public companies only and the author considered that private companieshave lower numbers, therefore the numbers give values lower than the real ones. Similardifficulties arise with founders as we mentioned earlier in our introduction. Who knows that AppleComputer did not have two founders (Wozniak and Jobs) but three (with the addition of RonaldWayne)? Building a database of founders and start-ups was not an easy task and the author willnot claim that it is void of mistakes or inaccuracies. Stanford founders are not the only founders ofthese companies, which is another limitation of the study. There might also be a bias in favor ofsuccessful companies and founders who may be easier to identify so that this may explain a ratherhigh level of success rate. When it was possible, all data were double checked but the authorrecognizes that there are strong limitations. It is an intuition he had when observing the start-upworld and its studies. The examples of experiences of founders and serial entrepreneurship areillustrations that general beliefs may have to be reconsidered. In terms of recommendations and future research, the author believes that the present workmay help in reassessing what could be benchmarks and good metrics for (academic) start-ups interms of value creation and growth. Stanford University and MIT are obviously exceptionaluniversities, but because innovation is a global phenomenon, it is not clear why other universitiesshould not measure their results according to the performance of these two institutions. Oneinteresting work might be to compare MIT and Stanford and analyze how similar or different theyare and if so why. Another possible study that the author did not have data to analyze would be theage of the founders and not their experience. The topic of high-tech entrepreneurship is fascinatingand even if decently well-known, opened to many new directions of research. What is the impactof venture capital in the value creation? What is the real impact of professors (vs. their students) inacademic start-ups? What is the role of luck vs. experience? Are there areas of activities which aremore favorable to start-ups than others? The author believes that he has contributed in a modestbut valuable manner to a better understanding of the dynamics of high-tech entrepreneurship.CONTACT: Hervé Lebret, herve.lebret@epfl.ch; (T): +41 21 693 7054; (F): +41 21 693 14 89;EPFL, 1015, Lausanne, Switzerland

8.
ACKNOWLEDGEMENTSThe author would like to thank Katarine Ku, head of the Office of Technology Licensing atStanford University for providing data on the spin-offs and the links to the 1992 study. He wouldalso like to thank the organizers of the 2009 Society for Entrepreneurship Scholars Meetingsponsored by the Marion Ewing Kaufman Foundation, where data used in this study were firstshown and discussed. REFERENCESBengtsson, O. (2008). Relational Venture Capital Financing of Serial Founders. Under Review.Carey, P. (2001, December 1). A Start-Ups True Tale: Often-told story of Ciscos launch leaves out the drama, intrigue. San Jose Mercury News.Di Gregorio, D. & Shane, S. (2003). Why Do Some Universities Generate More Start-ups than Others. Research Policy 32, 209-227.Djokovic, D. & Souitaris, V. (2008). Spinouts from academic institutions: a literature review with suggestions for further research. Journal of Technology Transfer 33, 225–247.Garnsey, E. & Heffernan, P. (2005). High-technology Clustering through Spin-out and Attraction: The Cambridge Case. Regional Studies 39 (8), 1127–1144.Gibbons, J. F. (2000). The Role of Stanford University: A Deans View. In: Lee, C., Miller, W., Hancock, M., Rowen, H. (Eds.) The Silicon Valley Edge (pp. 200-217). Stanford, CA: Stanford University Press.Gompers, P., Kovner, A., Lerner, J., & Scharfstein, D. (2009). Performance Persistence in Entrepreneurship. Harvard Business School. Working Paper 09-028Hsu, D. H. Roberts, E. B. & Eesley, C. E. (2007). Entrepreneurs from Technology-BasedUniversities: Evidence from MIT. Research Policy 36, 768–788.Jong, S. (2006). How organizational structures in science shape spin-off firms: the biochemistry departments of Berkeley, Stanford, and UCSF and the birth of the biotech industry. Industrial and Corporate Change 15 ( 2), 251–283Kenney, M. (2000). Understanding Silicon Valley. Stanford, CA: Stanford University Press.Kenney, M. & Goe, W.R. (2004). The role of social embeddedness in professorial entrepreneurship: A comparison of electrical engineering and computer science at UC Berkeley and Stanford. Research Policy, 33, 691 - 707.Ku, K. (2002). Software Licensing in the University Environment. Computing Research News 14 (1), 3,8.Lawton Smith, H. & Ho, K. (2006). Measuring the performance of Oxford University, Oxford Brookes University and the government laboratories’ spin-off companies. Research Policy 35, 1554–1568.Lebret, H. (2007). Start-Up, What We May Still Learn From Silicon Valley. Scotts Valley, CA: CreateSpace.Lee, C. et al. (editors), (2000). The Silicon Valley Edge. Stanford CA: Stanford University Press.Lenoir, T. (2002). The “Stanford Startup Report” project - Inventing the Entrepreneurial University: Stanford and the Co-Evolution of Silicon Valley. Retrieved from <http://www.stanford.edu/dept/HPS/TimLenoir/Startup/QuarterlyRpts>Leone, A., Vamos, J., Keeley, R., & Miller, W. (1992). Technology-based Companies Founded by Members of the Stanford Community. A Study Commissioned by the Stanford University Office of Technology Licensing.

9.
Lerner, J. (2005). The University and the Start-Up: Lessons from the Past Two Decades. Journal of Technology Transfer 30 (1/2) 49–56.Lowen, R. (1997). Creating the Cold War University. Berkeley, CA: University of California Press.Oskarsson, I. & Schläpfer, A. (2008). The performance of Spin-off companies at the Swiss Federal Institute of Technology Zurich. Zurich, Switzerland: ETH Transfer.Roberts, E., B. (1991). Entrepreneurs in High Technology: Lessons from MIT and Beyond. Oxford, UK: Oxford University Press.Roberts E. B. & Eesley, C. E. (2009). Entrepreneurial Impact: The Role of MIT. Kansas City, MO: the Ewing Marion Kauffman Foundation.Saxenian, A. (1994). Regional Advantage. Cambridge, MA: Harvard University Press,Saxenian, A. (1999). Silicon Valley New Immigrants Entrepreneur.s San Francisco, CA: Public Policy Institute of California.Shane, S. (2004). Academic Entrepreneurship, University Spinoffs and Wealth Creation. Cheltenham, UK: Edward Elgar.Smilor R., Gibson D., Dietrich G., (1990) University Spin-Out Companies: Technology Start-Ups from UT-Austin. Journal of Business Venturing 5, 63-76Zhang, J. (2003). High-Tech Start-Ups and Industry Dynamics in Silicon Valley. San Francisco, CA: Public Policy Institute of California.Zhang, J. (2009). Why do some US universities generate more venture-backed academic entrepreneurs than others? Venture Capital 11 (2), 133–162.

10.
Table 1: Stanford Spin-offs and Start-upsFields of activity Spin-offs Wellspring of Innovation All % VC- % All % all % high- VC- % all % high- backed tech backed techBiotech 72 35% 38 38% 66 3% 4% 41 5% 6%Medtech 34 17% 13 13% 113 5% 8% 64 8% 9%Computers 2 1% 2 2% 30 1% 2% 18 2% 2%Semiconductor 106 5% 7% 70 9% 9%Electronics 42 21% 19 19% 114 5% 8% 44 6% 6%Telecom 184 9% 13% 129 17% 18%IT & SW 40 20% 26 26% 310 14% 21% 153 20% 21%Internet 356 17% 24% 211 28% 29%Energy – Env 3 1% 1 1% 23 1% 2% 0 0% 0%Manufacturing 20 1% 1% 1 0% 0%Eng. Services 145 7% 10% 6 1% 1%Others 628 29% 24 3%Unknown 11 5% 1 1% 45 2% 0 0%Subtotal (high-tech) 1467 69% 100% 737 97% 100%Total 204 100% 100 100% 2140 100% 761 100%Table 2: Value Creation of Stanford Start-ups and Spin-offs Group Number VC ($M) M&A ($M) Public ($M) Sales ($M) Jobs Stanford Spin-offs 204 2969 8214 307136 65410 105281 1992 Study (not 383 1842 75406 185175 171579 454082 including licenses) Wellspring of Innovation 1’467 27125 173375 183615 111696 401453 (tech. only) Total-Tech 2’077 31936 256995 675926 348685 960816 WI (non tech) 673 272 11892 154413 46348 204895 Total 2’727 32208 268887 830339 395033 1165711 Top 5 high-tech 1719 75800 445000 223929 603528 5% 29% 66% 64% 63% Top 10 high-tech 2794 110200 497010 259828 680087 9% 43% 74% 75% 71%

11.
Table 3: Status of spin-offs and start-ups Status Spin-offs Wellspring of Innovation All VC-backed All VC-backed Public 8% 14% 5% 10% Private 39% 37% 33% 16% M&A 29% 36% 34% 55% Ceased 16% 13% 21% 19% Unknown 8% 7% Total 204 100 2140 754Figure 1: Number of years from foundation to liquidity event of spin-offs Unknown Energy ‐ Env IT & SW No VC or Electronics unknown Computers All Medtech VC backed Biotechnology 0 2 4 6 8 10 12 Nb. of companies VC backed No VC or unknown Unknown 2 Energy – Env. 3 IT & SW 16 5 Electronics 16 6 Computers 2 0 Medtech 2 4 Biotech 25 11 Total 61 31

12.
Figure 2: Number of years from foundation to liquidity event of WI companies Nb. of companies VC backed No VC or unknown Consumer goods Consumer Goods 8 45 Finance Finance 0 52 Non tech services Non Tech Services 2 67 Eng. Services Engineering services 5 37 Others Other tech 0 5 Manufacturing 1 10 Manuf. Energy – env. 0 7 Energy ‐ Env Internet 177 102 Without VC Internet IT & SW 125 93 All IT & SW Telecom 107 34 VC‐backed Electronics 35 43 Telecom Semiconductor 65 26 Electronics Computers 18 8 Semiconductor Medtech 51 27 Computers Biotech 36 18 Total 630 574 Medtech Biotech 0 2 4 6 8 10 12 14 16 18 20

13.
Figure 3: Status of WI companies by field of activities (N=2140) 100% 90% 80% 70% 60% 50% Unknown 40% Public 30% Private M&A 20% Ceased 10% 0%Table 4: Founders by company (including professors) Stanford founders All Founders including one professor by company Companies Individuals Companies Individuals 0 62 0 1 2’203 2’203 140 140 2 300 600 49 52 3 113 339 38 43 4 29 116 9 16 5 12 60 5 6 6 5 30 1 1 7 1 7 8 1 8 1 2 9 1 9 Total 2’727 3’372 243 260 Unique names 2’711 167

14.
Figure 4: Histogram of time between activity at Stanford and spinoff foundation (N=142). 100 90 80 70 60 50 40 30 20 10 0 ‐7 ‐6 ‐5 ‐4 ‐3 ‐2 ‐1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 36 YearsFigure 5: Histogram of time between activity at Stanford and start-up foundation (N=2523) 250 200 150 100 50 0 ‐17 ‐15 ‐13 ‐11 ‐9 ‐7 ‐5 ‐3 ‐1 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 Years

15.
Figure 6: Average time between activity at Stanford and foundation of start-ups (by field) Total 8.86 Consumer goods 9.81 Finance 13.04 Non tech services 12.63 Eng. Services 9.97 Manuf. 9.59 Energy ‐ Env 11.95 Internet 6.95 IT & SW 7.68 Telecom 9.37 Electronics 8.8 Semiconductor 7.11 Computers 6.43 Medtech 7.05 Biotech 5.71 0 2 4 6 8 10 12 14Table 5: Value Creation by Serial FoundersData on non-serial VC-backed M&A Public Ceased1739 Number Average Number Average Number Average 474 $36081’020 253 $520000000 102 $4929000000 370Data on serial VC-backed M&A Public Ceased988 Number Average Number Average Number Average 386 $39132000 220 $624000000 55 $5955000000 2321st comp VC-backed M&A Public Ceased445 Number Average Number Average Number Average 147 $28466000 120 $900000000 30 $11934000000 922nd comp VC-backed M&A Public Ceased445 Number Average Number Average Number Average 202 $42042000 93 $617000000 20 $3371000000 1073rd comp VC-backed M&A Public Ceased128 Number Average Number Average Number Average 57 $54251000 18 $277000000 6 $2324000000 394th comp VC-backed M&A Public Ceased46 Number Average Number Average Number Average 23 $38867000 13 $165000000 3 $1109000000 12