ETH Transfer

Embed or link this publication

Popular Pages

p. 1

eth transfer linking science and business ingvi oskarsson and alexander schläpfer the performance of spin-off companies at the swiss federal institute of technology zurich


p. 2

foreword dear reader innovation is one of the key factors in the success of the swiss economy this is why transfer of research results knowhow and technology from academia to industry is important and has only become more so over the years eth zurich works constantly in pursuit of its mandate to conduct research that benefits society creating spin-off companies is one of the most successful ways to transform scientific discoveries into products that meet the market needs of today and the future by sharpening its focus on spin-off creation in the past decade eth zurich has been able to establish an environment that fosters spin-off generation resources are and have been invested to raise the entrepreneurial awareness of students and to stimulate and support the transfer of eth zurich technologies into market-competitive spin-offs after such an intensive period of investment and growth now seems a good time to reflect on the results of the past ten years we are very fortunate that the two authors alexander schläpfer and ingvi oskarsson bringing with them the professional edge of the masters in finance program from the london business school elected eth zurich to be their case study on the success and economic impact of spin-offs eth zurich has nurtured the founding of 130 spin-offs in the last 10 years the spin-offs have had a direct impact on the local economy creating more than 900 direct jobs and a total pretax income of chf 43 mio in 2007 this resulted in about chf 18 mio of annual tax income the total investment in these companies is close to chf 170 mio with an estimated pooled internal rate of return irr of more than 43 percent these results are very encouraging they strengthen our conviction to continue with our efforts in support of spin-off creation and bolster our commitment to keep spin-offs market-relevant if not cutting-edge this study in book form is the first publication to provide indepth analysis of eth zurich spin-offs we hope that this book will be yet another means of encouragement for students to become entrepreneurs and to found their own companies and for universities to strongly support their own spin-off programs it is hoped too that this study will convince swiss policy makers to continue with or to even increase their efforts which have turned switzerland into a thriving place for entrepreneurs and new company generation and the promising location of many more to come finally we hope that this case study will entice much appreciated national and foreign investors to have an even closer look at the swiss spin-off portfolio and to deepen their investment in it we would like to thank the two authors for their enormous enthusiasm and surprising achievement that went far beyond the typical masters thesis our sincere gratitude is extended to zürcher kantonalbank zkb cti startup and the swiss private equity corporate finance association seca for their financial support that allowed publication of this study and the sharing of its exciting insights sincerely silvio bonaccio head of eth transfer gerd scheller spin-off manager eth transfer publisher eth transfer editors ingvi oskarsson alexander schläpfer dr marjan kraak dr gerd scheller content design cover design corporate communications eth zurich photo studio monte rosa prof andrea deplazes departement architecture eth zurich printed by neidhart schön ag circulation 3 500 © eth zurich september 2008 3


p. 3

table of contents the performance of spin-off companies at the swiss federal institute of technology zurich thesis for the masters in finance program msc finance at london business school supervised by prof francesca cornelli ingvi oskarsson and alexander schläpfer table of contents list of graphs tables and illustrations 1 executive summary 2 background and problem 3 data and description 3.1 data 3.2 description 4 literature review 5 analysis 5.1 failures and survival 5.1.1 spin-off survival compared to other universities 5.1.2 spin-off survival compared to all start-up companies in switzerland 5.2 job creation 5.2.1 job creation compared to other universities 5.2.2 job creation compared to all start-up companies in switzerland 5.3 vc/angel backing and exits 5.3.1 vc/angel backing compared to other universities key differences in the spin-off process ­ uk universities vs eth zurich 5.3.2 vc/angel backing compared to all start-up companies in switzerland 5.4 return on equity 5.4.1 methodology 5.4.2 equity value created 5.4.3 returns on equity the quality of returns in individual spin-offs 5.4.4 returns compared with vc funds in the us and in europe 5.4.5 returns compared to swiss stock market returns 5.4.6 returns compared to other university spin-offs 6 conclusions 6.1 the value and benefits to the economy as a whole 6.2 benchmarking performance with other universities 6.3 what process improvements do the spin-offs want to see 6.4 should eth invest in its own spin-offs 6.5 recommendations appendix 1 ­ graphs and tables appendix 2 ­ valuation list of abbreviations references about the authors 4 5 6 7 7 7 8 11 11 12 12 13 13 14 14 16 18 19 20 20 21 21 21 22 24 25 26 26 27 28 28 29 31 33 35 36 39 4 5


p. 4

list of graphs tables and illustrations 1 executive summary list of graphs tables and illustrations graph 1 graph 2 graph 3 graph 4 graph 5 graph 6 graph 7 graph 8 graph 9 graph 10 graph 11 graph 12 graph 13 graph 14 graph 15 graph 16 graph 17 graph 18 graph 19 graph 20 graph 21 graph 22 table 1 table 2 table 3 table 4 table 5 table 6 table 7 table 8 table 9 table 10 table 11 illustration 1 illustration 2 eth zurich spin-offs total population numbers by year and sector year-on-year rate of change in eth zurich spin-off creation and swiss new company incorporations from 2000 to 2007 timed survival rates for eth zurich spin-offs and for all new incorporations in switzerland and in the canton of zurich timed survival rates by major sector for total polulation of eth zurich spin-offs and for all new incorporations in switzerland and in the canton of zurich eth zurich spin-offs total population jobs created per sector in total and per surviving spin-off eth zurich spin-offs total population jobs created until dec 31st 2007 by year of spin-off incorporation surviving spin-offs only swiss start-ups and eth zurich spin-offs average numer of jobs created in each company over the first 5 years of operation eth zurich spin-offs vc/angel backed companies by year of incorporation eth zurich spin-offs total population with and without vc/angel backing analysis by sector eth zurich spin-offs sample of 82 equity raised by round and source eth zurich spin-offs total population exits accomplished by vintage comparison of spin-offs sectorial distribution uk universities vs eth zurich eth zurich spin-offs sample of 82 equity raised by year and provider eth zurich spin-offs sample of 82 method used for equity valuation eth zurich spin-offs sample of 82 absolute returns and money-multiples by sector eth zurich spin-offs return on equity by investor category ­ pooled irr eth zurich spin-offs vc returns in each of 24 investments eth zurich spin-offs distribution of founders returns in vc-backed and non-vc backed spin-offs returns of us and european vc funds 1998 ­ 2007 returns of us and european vc funds by year of fund creation ­ 1980 ­ 2006 eth zurich spin-offs sample of 82 job and equity value creation by vc-backed and non-vc backed firms eth zurich spin-offs sample of 82 of positive responses to the question `through which of the following measures could eth further improve its technology transfer performance eth zurich spin-offs composition of total population and survey sample eth zurich spin-offs `timed failure rates by vintage survival rates of university spin-offs in various countries and three universities eth zurich spin-offs sample of 82 equity raised eth zurich spin-offs vc/angel backing and exits ­ comparison with uk universities eth zurich spin-offs vc/angel backing and exits from vc backed spin-offs only ­ comparison with uk universities comparables used for valuation european and us venture capital fund performance 1969/80 to 2007 vs returns observed in eth zurich spin-offs international venture capital fund performance for funds created 1998 to 2006 vs returns observed in eth zurich spin-offs capm for expected vc-returns in the swiss market eth zurich spin-offs total population estimated personal income tax revenues the funding gap the generic technology transfer process 28 31 11 12 15 16 17 33 22 23 25 32 9 18 31 13 14 14 15 32 16 32 17 19 20 32 21 21 22 23 24 26 7 31 12 1 executive summary commercialising university technology by creating spin-off companies is a widely practised method of technology transfer today nevertheless there still seem to be some doubts about how effective this method actually is and whether it justifies the build-up in universities of dedicated resources to pro-actively support the creation of such spin-offs with data from 130 eth zurich spin-off companies created from 1998 to 2007 and detailed financial information obtained by questionnaire from a subset of 82 spin-offs we looked at three principal questions in our study 1 how successful these spin-offs are compared to all start-up companies in switzerland and compared to other university spinoffs internationally 2 whether the creation of such spin-offs appears to be beneficial to the local economy and 3 whether a comparison to other university spin-off programs could identify potential areas of improvement for eth zurich we were able to demonstrate that eth zurich spin-offs have significantly higher survival rates create more jobs attract more vc/angel investments and provide higher returns on equity than the average of all swiss start-up companies created over a similar time period compared specifically to spin-offs from leading uk universities the eth zurich spinoffs show higher survival rates a slightly lower job-creation a significantly lower proportion of venture capital vc or business angel backing but higher average investments per spin-off that receives backing and similar returns on equity vc/angel backing appears to be the key factor of growth and value creation as vc/angel-backed spin-off companies create significantly more jobs grow faster and founders experience significantly higher returns and lower failure rates if they are backed by vc s or angels than if not with a raw pooled return of 37.5 p.a before fees and carry the vc s/angels that have invested in a `hypothetical fund of eth zurich spin-offs made returns significantly higher than even the top quartile of us and european vc funds over the last decade and outperformed the swiss market index by over 2000 bps p.a during the same time period on basis of a simple capm we estimate that this portfolio of eth spin-offs has experienced abnormal returns in the range of 20 ­ 25 p.a with annual revenues of approx chf 250 million the 130 spin-offs have to date created close to 1500 direct and indirect jobs and generate annual personal and corporate income tax revenues to local and federal government of an estimated chf 18 million p.a not directly quantifiable benefits include the formation of innovation clusters and the attraction of highly qualified students and faculty to eth we have finally identified a set of recommendations for eth zurich mainly aimed at improving the ratio of vc/angel backing among their spin-offs 6 7


p. 5

2 background and problem 3 data and description 2 background and problem the swiss federal institute of technology zurich eth zurich is among the world s leading science schools and has been ranked in 2007 in the top 30 of universities in the world both by the shanghai jiao tong university s survey and by the times higher education supplement1 with a budget of over 1 billion swiss francs chf a year it provides higher education to more than 12000 students at bachelor and post-graduate level and conducts cutting edge research with close to 5000 staff fte including phd s the focus of its teaching and research is in natural sciences and mathematics engineering sciences system-oriented sciences earth environmental agriculture and food sciences construction and geomatics as well as in specialized areas of management and social sciences e.g technology management besides teaching and research eth zurich considers the transfer of its technology to a wider application and commercialisation in industry and education as third major element of its mission it expanded its technology transfer activities in the early 1990s created a specialised group to manage the patenting and transfer of its technology in 1995 and formalised ­ in 2005 ­ the group s status as a distinct unit eth transfer with a dedicated budget and reporting to the vp for research eth zurich utilises the following methods of direct technology transfer 1 research collaborations with industry or educational institutions 2 technology licensing and 3 spin-off creation while its first documented spin-off 2 company was incorporated in 1973 it is only over the last decade that eth zurich has been putting a stronger emphasis ­ and resources ­ on supporting the creation of such companies as for any other new initiative that it has been launching eth is interested in knowing whether the spin-off-program is meeting its objectives i.e the commercialization of its research and the creation of jobs for its graduates as well as for others and shows a satisfactory performance after it had undertaken in 2004 a review of survival rates and jobcreation in its spin-off companies eth transfer is now looking to conduct a more in-depth study into the performance and wider economic impact of its spin-offs the objective stated for this new study therefore is twofold first to benchmark the performance of the eth spin-offs with data samples from other universities and to determine potential areas of improvement where there is a clear difference in performance and second to demonstrate ­ to the extent possible ­ the value and benefits created by these spin-offs to the economy as a whole 3 data and description the basis for this study is a total population of 130 eth zurich spin-off companies that have been created in the 10-year period between jan 1st 1998 to december 31st 2007 3 eth transfer has given us full access to its spin-off database as well as the data collected in the frame of the 2004 study furthermore we had access to the paper documentation for each spin-off company available at eth transfer s offices on campus together with our own research of company websites the swiss company registry and telephone interviews with a large number of founders we were able to verify ­ for the total population of 130 ­ survival rates the number of employees the number of companies with vc/angel participation as well as the number of exits ipo or trade-sale furthermore we conducted a survey by questionnaire ­ in particular to obtain specific data on equity funding and financial performance of the spin-offs after a trial round with 8 selected spin-offs we sent out a questionnaire on march 5th 2008 to 115 companies and received ­ after follow-up ­ 74 responses with complete data in addition we have been able to fully research the equity funding of 8 companies that have either gone into bankruptcy or ceased commercial activity giving us a total of 82 valid returns i.e a 63.1 response rate in table 1 appendix 1 we show that the composition of this sample correlates very closely with the composition of total population in terms of outcome survival bankruptcy/inactive sector representation and vintage representation and support this with a chi-square-test and correlation factor for each criterion finally in each case applicable we have determined the specific parameters standard error based on the sample size and mention the significance level of our findings separately for our analysis we have then contrasted the data from eth zurich with data available in a number of other studies and publications most notably in the following · two studies on the performance of uk university technology transfer offices and spin-off companies published by library house in 2007 holi et al 2007 and franklin et al 2007 3.1 data · the statistics on spin-offs published annually by the high· · · er education founding council for england he ­ bci surveys for the years 99/00 through 05/06 the swiss federal office for statistics annual data on new company incorporations in switzerland as well as a specific study on new company survival and job creation rates in switzerland swiss federal office for statistics 2008 the european venture capital associations s annual publication on pe/vc investment performance and activity evca 2008 and imperial innovation s annual reports and ipo prospectus 3.2 description as shown in graph 1 below the total population of 130 eth zurich spin-offs is composed of 10 annual vintages or cohorts in each of which between 9 and 21 new spin-off companies have been incorporated graph 1 eth zurich spin-offs total population numbers by year and sector 25 20 of spin-offs 15 10 5 0 it material sciences chemicals 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 biotech and pharmaceutical electrical electronics consulting services medtech diagnostics mechanical avionics micro nanotech sensors analytics source data from eth spin-off database 1 2 27th in the overall ranking by shanghai jiao tong 42nd in thes overall ranking but 18th in thes `natural sciences and 13th in `technology categories for the purpose of this study we use eth zurich s definition of the term spin-off `a spin-off company of eth zurich is a newly founded company by eth employees or graduates based on research results of eth zurich this definition is widely supported by academic literature and congruent with scott shane s shane 2004 definition of `a new company founded by current or former members of a university to exploit a piece of intellectual property created in that university uk academic literature sometimes refers to spin-offs as `spin-outs however following the same definition as above in contrast `start-up s are all newly created companies whether with or without university technology or university members participation 3 following our strict definition of `spin-off we have excluded a hand-full of start-ups i.e newly created companies not using eth zurich technology created by eth zurich graduates in the same time period a comparison with data on new company incorporations in switzerland from 2000 to 2006 shows that the year-on-year rate of change in new spin-off creation is fluctuating in the same direction as all new company incorporations although due to the much smaller base the fluctuations are much more pronounced among the spin-offs see graph 2 appendix 1 this suggests that the variation in new eth spin-off incorporations may be ­ to a good extent ­ influenced by the prevailing economic climate in switzerland 8 9


p. 6

4 literature review 4 literature review 4 literature review it/technology 26 and biotech 16 are the two sectors with the largest and most regular representation in spinoff creation while medtech chemicals and the various engineering sciences each make up between 5 and 8 see table 1 appendix 1 for details consulting and services 20 is a collection of service businesses in various sectors such as architecture/construction geology/geophysics meteorology hydrology health and business/technology management it and biotech are the only two sectors in which spinoffs have been created every year and ­ until 2005 ­ these two sectors were actually driving the variation in number of spin-offs from year to year as of 2006 the number of non-it and non-biotech spin-offs has started to increase above the previous average of 6 ­ 7 per year as explained above the data of our survey sample of 82 spin-offs questionnaire returns from which we derive the financial performance indicators for this report is ­ to a large extent ­ congruent with the total 130 spin-off population see table 1 appendix 1 for details although there are examples of university technology spinoffs dating as far back as to the 18 th century shane 2004 mentions the example of professor johannes pickel who started in 1784 an acetic acid production based on his discoveries at würzburg university in germany skepticism whether universities should engage in creating spin-off firms to commercialise technology have been prevailing in academia ­ as well as among the institutions determining the universities budgets ­ still far into the 1970s in the us shane 2004 and probably even longer in europe in the 1970s some of the leading science universities in the us started to experiment with policies to promote spin-off creation and the 1970s saw also the first creation of universitylinked venture capital funds shane 2004 along with these new dynamic academia started taking an interest in `academic entrepreneurship and since the 1980s technology transfer as a whole and spin-off creation in particular has been widely researched shane 2004 o shea et al 2005 the area of particular interest for our study is the performance measurement and the definition of performance indicators for spin-off success many publications focus on reviewing the success of universities technology transfer program as a whole and measure success predominantly in the number of spin-offs created per year among the publications that deal more specifically with the performance of the spin­off companies the most frequently used measure of success is survival rates shane 2004 references 7 different publications and we have separately reviewed data from 4 further studies mustar 1997 lawton smith 2006 leung and mathews 2006 clayman and holbrook 2006 although these studies cover different time-spans and ­ likely ­ periods with varying economic conditions the stated university spin-off survival rates are in the range of 70 ­ 90 and consistently higher than the survival rates of nonuniversity start-up companies shane 2004 we also looked extensively at literature appraising vc investment performance most authors do not use `survival as a benchmark but rather look at the four possible outcomes of portfolio investments ipo `trade-sale `still in portfolio and `failure and develop ­ with empirical data ­ models based on the competing probabilities of these outcomes dean and giglierano 1990 cochrane 2004 metrick 2007 we have included a more detailed review of the data found in these articles under the heading `failure rates in the next chapter a second frequently used performance indicator is `employment or `job creation particularly the higher education governing bodies and national or regional enterprise development agencies understandably focus on this aspect of spinoff creation and sponsor studies such as the ones by uk s library house already referenced above holi et al 2007 and franklin et al 2007 or the unico survey unico 2001 but also academia uses this metric shane 2004 lawton smith and ho 2006 and ­ although numbers and measuring approaches vary widely there is a general consensus that university spin-off companies create more direct jobs than the average small business founded in the same country they also seem to create more qualified jobs and ­ as new technology companies tend to cluster ­ contribute more to local economic development through the creation of indirect jobs shane 2004 shane also shows that the transfer of technology to spin-offs creates more jobs than the licensing of technology to large existing corporations a third measure often discussed in literature is the spin-offs ability to obtain angel or venture capital investments shane and stuart 2002 wright et al 2006 and ­ if so ­ the amount of venture capital funding received lawton smith and ho 2006 unico 2004 connected to that is the number or percentage of successful exits through ipo which ­ as shane and stuart 2002 show ­ itself is strongly correlated to the amount of venture capital funding received although vc s seem to have a greater concern with the quality and ­ in particular ­ the lacking experience of management teams in university spin-offs than with other companies they are backing wright et al 2006 and are therefore reluctant to back university spin-offs at seed start-up stage the evidence in literature points to a significantly higher proportion 20 to 40 times more of university spin-offs being able to obtain venture backing than the average small business shane 2004 wright et al 2006 spin-offs tend to get backed however at a later stage resulting in a so-called `funding gap in the seed/start-up stage as per the following illustration illustration 1 the funding gap stage source demand per round supply pre-seed founders/ff $20-50k seed/start-up business angels $50-250k early angels/vc $2000k expansion vc $5000k funding gap source adapted from sohl 2003 at this point business angels are the major source of equity capital and provide in the range of $50 000 to $250 000 per investment round which can secure 12 ­ 18 months of operation sohl 2003 once the spin-offs have vc backing they manage to raise larger amounts and ­ not surprisingly ­ a substantially higher 10 times more proportion of university spin-off companies experience an ipo than other start-ups lockett and wright 2005 access to venture financing is therefore a key determinant of growth and value generation for new technology-based spin-off firms and new ventures that do not attract vc funding in the initial years are unlikely to do so in the future wright et al 2006 shane and stuart 2002 so why do `academic enterprises perform better than the average start-up company de coster and butler 2005 have empirically shown that university spin-offs have normally an advantage with a better protected competitive position cutting edge technology and ip protection through patenting and they manage to satisfy better the market demand de coster and butler explain the latter observation by the spinoffs conducting more systematic market research and preparing better business plans indicating that the support services available to university spin-offs are more effective than those available to other start-ups mustar 1997 attributes the success to a generally more extensive support system research funding access to university laboratories incubators and the network of academic and non-academic contacts and shane 2004 sees a distinct value in the `university brand that is normally associated to spin-off companies 11 10


p. 7

4 literature review 5 analysis 5 analysis the fourth measure is financial returns there is very limited literature specifically on returns in university spin-offs but the financial performance of vc investments has been the focus of several academic studies with the general premise being that venture capital investments have historically displayed high average returns and high risk standard deviation but observations of returns vary widely chen baierl and kaplan 2002 analysed data gathered by venture ecomomics for a 40 year period from 1960 through 1999 but focused on liquidated funds their conclusion is that venture capital has had an annual arithmetic average return of 45 with a standard deviation of 115.6 over this period the geometric average return compounded average is estimated around 13 the correlation between vc and publicly traded equity is estimated to be close to zero 0.04 the realised median annual irr among the 148 funds they studied is only 8.5 and the average is 9.99 the maximum annual irr they observed is 74 and the minimum is ­72 kaplan and schoar investigated the performance of 765 private equity funds from the venture economics database from 1987 through 2000 on average they found that lbo-fund returns net of fees are slightly higher than those of the s&p 500 while vc-fund returns are lower on an equal-weighted basis but higher than the s&p 500 on a capital-weighted basis they concluded that these results combined with previous evidence on private equity fees however suggest that ­ on average ­ both types of private equity returns exceed those of the s&p 500 gross of fees ljungkvist and richardson 2003 analysed a sample of 73 mature funds established from 1981 through 1993 and found an average irr of 19.81 and a standard deviation of 22.29 moreover they observe a 5 ­ 8 annual return above s&p 500 and 2 ­ 6 above the nasdaq composite for these funds cochrane studies data from the ventureone database from 1987 to june 2000 and investigates vc returns based on the economics of individual investments in portfolio companies he reports a mean log return of 15 for the whole dataset compared to 15.9 for the s&p 500 over the same period the standard deviation of the log return is 89 much larger than the 14.9 standard deviation of the s&p 500 log return over the same period this indicates that venture returns are very volatile but he also finds that later stage vc deals have less volatility than early stage deals furthermore cochrane s model estimates 12 the beta for vc fund returns at 1.7 and arithmetic returns gross of fees with a highly positive alpha 32 per year over his sample period artus et al investigated the performance of european private equity from 1985 through 2002 from a dataset from thomson venture economics their study included a calculation of the internal rate of return based on cash flows of 201 funds the funds which were selected had either been liquidated or had a small residual net asset value lower than 12 they report an average irr for venture funds of 10.6 and an excess-irr return compared to the msci europe of 4.4 for the average european private equity fund including buy-out funds the final area of literature that we reviewed is around the question whether universities should take equity stakes in their spin-offs or not in a paper that has been widely referenced lerner 2005 takes a strong stance specifically against universities taking large stakes in spin-offs he mainly argues along the principal-agent dilemma that vc investors want the entrepreneurs to take a substantial stake and that ­ if a third party takes such a large stake ­ management s incentives are diluted another argument he puts forward is that universities technology transfer offices tto often act as trusted intermediary introducing to vc s good new business opportunities if tto s now themselves become investors and compete for good deals their role as `honest broker will be undermined shane 2004 takes a contrarian s view arguing that creating spin-offs ­ and taking an equity stake in compensation for licensing the universities intellectual property ip is a more profitable way of transferring technology than licensing the technology to established companies feldman et al 2002 see three major advantages in universities taking equity it 1 provides the university with an option on the patents true commercial value 2 it aligns interests of university and entrepreneurs towards the common goal of commercialising the technology and it 3 may serve as `certification function that provides a signal to other investors and to the market that the university is confident of the technology s value of the total 130 eth spin-offs incorporated since 1998 9 have been liquidated and a further 6 have ceased commercial activity although they were still registered with the company registrar by december 31st 2007 we defined `commercial activity for the purpose of this study as 1 the company having employees either full or part-time and 2 regular revenues of chf 10 000 or more per year the population s aggregate failure rate4 is therefore 11.5 15 of 130 spin-offs the average time to failure for these 15 companies was 3.75 years with extremes of 9 months at the low end and 10 years in maximum 7 of the 15 spin-offs have gone out of business during their first 2 years of activity while the remaining 8 were active for between 3 and 10 years in terms of vintages 1999 had the highest failure rate with 31 5 of 16 as of dec 31st 2007 followed by 1998 with 22 2 of 9 11 73 of the 15 failures have been incorporated in the period from 1998-2001 see table 2 below this can partly be explained by simple statistics i.e if time-to-failure ­ ttf ­ values are more or less evenly distributed from 1 to 10 years then the number of failure events will be higher in the older vintages where more ttfvalues can occur and partly with the above average absolute number in new spin-off creation in the years 1999 and 2000 table 2 eth zurich spin-offs `timed failure rates by vintage failure rate after vintage 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years 10 years 1998 1999 0 6 0 6 13 0 13 0 25 0 25 0 31 0 11 31 11 31 22 5.1 failures and survival 2000 2001 0 2002 0 6 2003 2004 10 0 0 6 2005 0 20 0 0 12 2006 0 2007 0 11 8 20 0 0 12 average 0 2 0 6 11 8 20 0 0 12 8 20 0 20 12 10 20 12 7 8 9 13 16 18 21 22 source own computation with data from survey and separate research 4 for the purpose of this study we distinguish between two different methods for calculating failure rates the `aggregate failure rate is calculated for a sample/population of companies created over a series of years where the number of liquidated/out-of-business spin-offs by the end of the study period s last year is divided by the total number of spin-offs created over the study period this method does not take into account the age of the companies the `timed failure rate refers to the percentage of businesses liquidated within a specific number of years from their incorporation and therefore better reflects the age factor from a sector point of view biotech 4 companies 27 of failures and the engineering sciences electrical/electronics 2 13 material sciences 2 13 are slightly over-represented considering their proportion in the total sample while the it spin-offs 3 20 are slightly underrepresented counter to intuition there is no clear evidence of the `bursting of the tech-bubble as only 2 of the 11 failed spin-offs incorporated in the 1998-2001 period were it businesses none of the total 20 spin-offs in medtech chemicals and sensors analytics have failed in 4 27 of the failed spin-offs we have found evidence of vc/angel participation this is in line with the overall level vc/angel participation in 26 or 34 of the 130 total population and also the sub-samples failure rate in vc/angel investments of 12 is equal to the populations overall failure rate hence vc/angels have not had a particularly good or bad hand in selecting spin-offs to back although there has been much academic research into the topic no precise estimates of failure rates in early-stage vc investment seem to be available metrick 2007 ­ in his extensive study of data from sand hill econometrics she database ­ has determined that after 5 years from incorporation 6.3 of venture-backed start up firms had failed 33 had experienced an exit ipo or trade sale and 60.7 were still in the vc s portfolio after 10 years 14.3 had failed 61.2 of companies had experienced an exit and 24.6 were still in vc portfolios however she has labeled companies for which no status is available as `still private i.e still in the vc portfolio and metrick 2007 suggests that ­ as vc s ordinarily exit all their investments after 10 years normal fund life ­ most of the 24.6 listed as `still in vc portfolios may actually be failures that the vc omitted to report he therefore determined the `timed failure rate after 10 years likely to be in the range of 30 ­ 40 dean and giglierano 1990 in their study of 38 silicon valley based venture capital funds report an average failure rate presumably calculated as aggregate failure rate of between 15 and 16 however with large standard deviations of 18 percentage points in single round investments and 13 percentage points in multiple-round investments finally mason and harrison 2001 report a failure rate aggregate failure rate of 34 in their study of 127 earlystage investments by uk business angels in all above studies `failure is defined as total loss of investment 13


p. 8

5 analysis 5 analysis 5.1.1 spin-off survival compared to other universities the aggregate survival rate for eth zurich 1998-2007 spinoffs is 88.5 115 out of 130 as shown in table 3 below we have compared this to spin-off survival rates published in 5 studies that we were able to access in our extensive research of academic publications these studies all use the same method of calculating the aggregate survival rate and were conducted at either national or university level again eth compares very favorably particularly when considering that the only value higher than eth s survival rates northern ireland is based on a quite small sample study durations of 8-10 years and even beyond are reasonably comparable as failure rates start to level off as of the 7 th year of company existence table 3 survival rates of university spin-offs in various countries and three universities country survival rate 68 73 period 1980-2000 1995-2003 1997-2004 years 21 9 8 sample source size n 3376 shane 2004 301 clayman and holbrook 2006 56 leung and mathews 2006 92 shane 2004 canada usa lawton smith and ho 2006 provide evidence by comparing oxford university s 80 aggregate survival rate with the 71 timed 3-year survival rate for uk businesses measured for all businesses incorporated in 2002 in the case of eth zurich spin-offs this difference is far more pronounced the swiss federal office for statistics 2008 published the 1 to 5 year `timed survival rates of all companies newly incorporated in the years 2000 to 2004 for switzerland as a whole and for the canton of zurich where 111 of the 130 spin-offs have their registered domicile we have compared this data to exactly the same vintages of eth zurich spin-offs and graph 3 below shows that eth zurich s spin-offs have a survival rate that is between 16 percentage points for year 1 after incorporation to 44 percentage points for year 5 after incorporation higher than the ones of all new companies in switzerland and in the canton of zurich taking instead the average survival rates of all eth spin-off vintages would show an only minimally different result i.e 40 percentage points difference in year 5 after incorporation graph 3 timed survival rates for eth zurich spin-offs and for all new incorporations in switzerland and in the canton of zurich 100 of companies surviving eth zurich 2000 ­ 2004 5.2 job creation hongkong france netherlands sweden 79 83 84 87 1984-1992 1984-1987 1960-1993 1984-1995 9 4 n ireland university usa ­ mit 94 34 100 mustar 1997 30 shane 2004 17 shane 2004 12 uk ­ oxford 80 81 88 1980-1996 1994-2002 1998-2007 17 10 9 134 shane 2004 130 own survey 90 80 70 60 50 40 1 2 3 years after incorporation 4 5 eth ­ zurich 83 lawton smith and ho 2006 source own compilation from sources indicated above all new incorporations switzerland 2000 ­ 2004 all new incorporations canton zurich 2000 ­ 2004 eth zurich s 130 spin-offs incorporated since 1998 have so far by december 31st 2007 created employment opportunities for a total of 918 persons on average every spin-off has therefore created 7.1 jobs5 if we include those spin-offs that went out of business or 7.98 jobs if we consider only the `surviving 115 spin-offs from a sector perspective the most jobs in absolute numbers were created in it 258 followed by biotech/pharmaceutical 126 and sensors analytics 121 while it and biotech/pharmaceutical are among the sectors with the highest number of spin-offs created the job creation in sensors analytics is largely due to the success of one particular spin-off ­ which is clearly visible in graph 5 below where sensors analytics is also the sector with the highest number of jobs created per spin-off spin-offs in the consulting and services group have created among the fewest jobs per company indicating that many of the businesses in this group are what may be called `life-style businesses i.e businesses through which the founders seek the independence of being self-employed but do not necessarily create a large amount of other jobs however as we will show later the variation in levels of job creation by sector seems to be driven rather by a difference in venture capital backing for specific sectors than by sector specific manpower requirements or job-creation dynamics it would therefore be erroneous to conclude that specific sectors show a genuinely higher job-creation pattern and therefore merit a stronger support in view of creating more employment opportunities graph 5 eth zurich spin-offs total population jobs created per sector in total and per surviving spin-off company 300 total jobs created 200 100 0 25 20 15 10 5 0 bi ot ec source data from survey and separate research the low survival rate in the us ­ where some of the most successful university spin-offs have been created ­ raises however the question whether a high survival rate is actually desirable or whether too strong a focus on creating `surviving spin-offs does not eliminate some of the potentially very successful ventures that may not look so promising or too risky 5.1.2 spin-off survival compared to all start-up companies in switzerland as already discussed in our literature review above university spin-offs seem to have higher survival rates than the average newly incorporated small businesses shane 2004 14 source own comparison with data from the swiss federal office for statistics a sector-specific comparison for biotech/healthcare it and `other services see graph 3 in appendix 1 consistently confirms these markedly higher survival rates of eth zurich s spin-offs versus the total of new incorporations in switzerland over the same time period 5 5.2.1 job creation compared to other universities data published annually by the uk s higher education funding counsel he-bci survey 2007 show that the 1145 surviving university spin-offs in the uk have created ­ until 2006 ­ 16 225 jobs i.e 14.2 on average per spin-off 6 this ratio of job creation has been gradually improving from 12.9 jobs/spin-off since 2002 likely reflecting the improving macro-economic environment during this period but also the increasing average age of companies in the sample data from other countries show large differences and shane 2004 references six studies that estimate university spinoff job creation in a range of 4.8 jobs in the university of twente nl to 83 jobs per spin-off in a countrywide analysis of the us based on a study by the association of university technology managers that covers spin-offs over a 20-year time period from 1980-1999 these differences may be explained by the variance in time-spans the samples cover i.e the longer the time-period the higher the number of employees on average as the number of employees and company age are positively correlated in the three studies referenced by shane that cover comparable 9 ­ 10 year timespans each spin-off has created 4.8 10 and 10.6 jobs in this chapter `jobs signifies the number of persons employed rather than full-time equivalents fte 6 hefce uses the same definition of spin-off as suggested in footnote 2 se ph rv ice ar m s ac eu tic al s el ch ec em tr ica ica l ls el ec tr on ics m at er it ia m ls ec cie ha nc ni ca es m l ed te av ch io ni cs di ag m no icr st o ics na se no ns te or ch s an al yt ics ns ul co h an d tin g per sector total per spin-off created rhs aveage no of jobs created per spin-off 15


p. 9

5 analysis 5 analysis respectively in the case of eth zurich s spin-offs the positive correlation between age and job-creation is clearly visible in graph 6 below where the older vintages 1998 ­ 2002 have ­ on average ­ created 13.6 jobs and the younger vintages 2003 ­ 2007 4.5 jobs each spin-off has created on average 1.8 new jobs per each year of its life graph 6 eth zurich spin-offs total population jobs created until dec 31st 2007 by year of spin-off incorporation surviving spin-offs only jobs created per spin-off company 250 total jobs created 200 150 100 50 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 30.0 25.0 20.0 15.0 10.0 5.0 0.0 average 3.7 jobs have been created during the first five years among the 2000 ­ 2004 vintages while the average here is still higher than the 3.0 of all swiss start-up companies the difference is not statistically significant t-stat=0.7 standard error of 1.0 n=51 graph 7 swiss start-ups and eth zurich spin-offs average numer of jobs created in each company over the first 5 years of operation jobs created in each company average y1 to y5 5 4 3 2 1 0 all swiss start-ups 2000 ­ 2004 eth spin-offs 2000 ­ 2004 eth spin-offs 2003 ­ 2007 graph 8 eth zurich spin-offs vc/angel backed companies by year of incorporation angel backed vc s/angels have contributed almost 91 of the total equity funding requirements for all 82 spin-offs in our sample 50 40 30 20 10 25 20 of spin-offs 15 10 5 0 1998 1999 2000 without vc/angel participation with vc/angel participation with vc/angel participation rhs table 4 eth zurich spin-offs sample of 82 equity raised total equity raised chf non-vc/angel-backed vc/angel-backed total founders and others 6 419 242 vc/angels total 162 987 658 169 406 900 6 419 242 15 552 124 153 854v776 9.2 9 132 882 153 854 776 90.8 equity raised per spin-off chf non-vc/angel-backed 58 vc/angel-backed 24 founders and others 380 537 110 677 vc/angels 6 410 616 total 6 791 152 110 677 2001 2002 2003 2004 2005 2006 2007 0 year of incorporation source data obtained in survey source own data obtained from survey and separate research vintage total until 31/12/2007 per spin-off incorporated rhs per spin-off per year of life rhs source swiss federal office for statistics eth database own data collection source own data obtained in survey and separate research 5.2.2 job creation compared to all start-up companies in switzerland in the previously mentioned set of data of all new company incorporations in switzerland published by the swiss federal office for statistics the average surviving start-up company has created 3.7 jobs after 5 years since eth zurich conducted a job count in 2004 we have two vintages to compare the 5-year job creation the surviving spin-offs incorporated in 2000 had ­ on average ­ created 15.0 jobs by 2004 and the ones incorporated in 2003 had created 4.0 jobs on average by 2007 as shown in graph 6 above there are significant differences between the eth spin-off vintages nevertheless comparing the average job-creation over each of the first 5 years since company incorporation we can say with 99 confidence that eth spin-offs of the vintages 20042007 average 4.5 jobs standard error of 0.36 n=64 have created more jobs than the average swiss start-up company average 3.0 jobs n=37569 in the vintages 2000 ­ 2004 given that this difference may to a certain degree be influenced by the prevailing economic cycle we have also considered the data of eth s previous study according to which on 16 in our detailed survey of 82 spin-offs the `sample of 82 spinoffs we have further observed that each job is ­ on average ­ 0.81 full-time equivalent fte and that 42 of all spin-off employees are eth graduates 5.3 vc/angel backing and exits among the total population of 130 eth zurich spin-offs we have found evidence of venture capitalist or business angel backing in 34 companies 26.1 with a total of 80 investment rounds from a `vintage perspective as shown in graph 8 below the years 2000 and 2004 have the highest number and proportion 40 of vc backed companies while 2007 is ­ so far ­ the year with the lowest proportion 14 but we will see later that vc s tend to invest on average only after two years from incorporation meaning that the vc-backing in the years after 2005 may still improve looking at the sectors in graph 9 in appendix 1 biotech 10 vc/angel backed spin-offs it 7 and material sciences 4 appear to be the most popular sectors in absolute while in relative terms biotech chemicals material sciences and medtech seem to stand the highest chances of obtaining vc angel backing in only 9 7 of the 130 spin-offs have the founders sold a major part of their stake and 6 of these were venture backed the vc s exit rate is therefore 17.7 with one 2.9 ipo and five 14.7 trade-sales compared to the 5-year exit-rate of 33 observed by metrick 2007 in his study of almost 12 000 vc investments mentioned above these values seem low the average time from incorporation to exit for the founders was 5.54 years among our detailed survey sample of 82 spin-offs 24 were vc/angel-backed7 for these 82 companies we obtained detailed data on 50 different financing rounds according to this data shown in table 4 below venture capitalists and angels have provided equity totaling almost chf 153.9 million the founders their family and friends as well as other investors have contributed an additional chf 9.1 million equity to venture-backed spin-offs and chf 6.4 million to the non-vc nevertheless there is a noteworthy time delay before vc s or angels start backing spin-offs and we have observed among 20 venture backed companies which have not managed to raise vc or angels money at seed stage that it took them ­ on average ­ 723 days almost 2 years from their last financing round by founders or others to obtain their first venture/angel investment ­ evidence of the funding gap discussed earlier this gap becomes even clearer when looking at the different rounds we observed in our sample of 82 spin-offs in graph 9 below in a total of 96 seed and start-up rounds the spin-offs raised chf 7.7 million of which chf 5.6 million 74 from founders friends and family and only chf 1.2 million 16 from angels and vc s others ­ mainly zürcher kantonalbank zkb ­ help to close some of that gap 8 interesting is also the surprisingly high level of founders friends/families participation in a-rounds with total investments of chf 5.0 million 65 of these a-round investments by founders are made without vc/angel-backing and have an average size of chf 187 075 however those founders and family friends who have managed to raise a vc a-round seem to take the opportunity to invest along with actually higher average investments chf 275 025 in general it is striking how `wealthy and keen to invest the founders personal networks in switzerland are according to shane 7 to be specific of those 24 spin-offs 13 had venture capital-backing 8 had angel-backing and 3 both 8 zkb invests in both equity and convertible debt included here are only straight equity investments 17


p. 10

5 analysis 5 analysis 2004 the average us entrepreneur raises only approx $50 000 of equity from own savings family and friends and uk numbers appear to be in a similar magnitude mason and harrison 2002b important for this study is however the indication that the founders personal networks seem to bridge some of the funding gap that is left open particularly by business angels graph 10 eth zurich spin-offs sample of 82 equity raised by round and source 70 60 chf million 50 40 30 20 10 0 seed/start-up round a 36 1 850 369 round b 15 3 553 462 round c 8 3 209 377 round d 6 1 647 348 round 3 2 088 496 table 5 eth zurich spin-offs vc/angel backing and exits ­ comparison with uk universities institution total spin offs 130 67 30 with vc/angel backing 34 26.2 18 26.9 trade sale 8 1 8 1 6.2 1.5 1 0 5 0 ipo 0.8 0.0 2.1 0.0 table 6 eth zurich spin-offs vc/angel backing and exits from vc backed spin-offs only ­ comparison with uk universities institution total with vc/angel trade sale ipo spin offs backing vc backed only vc backed only 130 67 34 26.2 18 26.9 5 14.7 0 8 0.0 1 0 2.9 0.0 eth zurich 1998 ­ 2007 eth zurich 2001 ­ 2006 20 uk universities 2001 ­ 2006 university of cambridge university of oxford imperial college 233 137 58.8 24 20 66.7 3.4 eth zurich 1998 ­ 2007 university college london university of edinburgh 29 18 75.0 26 9 19 65.5 0 3.3 eth zurich 2001 ­ 2006 15 57.7 6 66.7 1 0.0 2 22.2 0 3.4 1 1 4.2 20 uk universities 2001 ­ 2006 university of cambridge university of oxford imperial college 233 0.0 0 3.4 30 137 58.8 0 0.0 24 20 66.7 0.0 university college london university of edinburgh 29 18 75.0 1 5.8 source own survey and data from hori et al 2007 others zkb vc/angel founders ff 26 9 19 65.5 0 5.0 5 15 57.7 6 66.7 1 0.0 0 3.6 2 33.3 0 5.3 1 0.0 make up for almost 30 of eth zurich s spin-offs but are not main-stream areas of vc investment interestingly as shown in graph 8 in appendix 1 eth zurich scores an above average 40 vc/angel backing in its material sciences spin-offs meaning that the area of real shortage of vc/angel investments may be the engineering/industrials i.e electrical engineering electronics mechanical engineering avionics and sensors analytics graph 12 comparision of spin-off s sectorial distribution uk universities vs eth zurich lifesciences 1 5.6 0.0 0 5.3 0 0.0 0.0 source own survey and data from hori et al 2007 of rounds 96 average chf 79 864 source own data obtained in survey 5.3.1 vc/angel backing compared to other universities holi et al 2007 have conducted a very detailed survey of the technology transfer and spin-off activities of 20 uk universities with data covering a six-year period from 2001 to 2006 as shown in table 5 below on average close to 60 of the 233 spin-offs created over these 6 years have obtained vc angel backing and this level of backing is even higher among the leading uk universities such as oxford cambridge imperial college and ucl compared to the average of the 20 uk universities as well as of the 5 leading ones shown in the table 5 below the level of venture backing in eth zurich s spin-off is significantly lower by 14.6 standard errors compared to the 5 leading universities as for exits the picture is similar in a comparable time-period 2001 ­ 2006 eth zurich s total population of spin-offs experienced significantly less exits by trade-sale and ipo than the uk university spin-offs did as most of eth zurich s exits have concerned the spin-off vintages 1998 ­ 2000 see graph 10 in appendix 1 in order to eliminate a potential bias arising from having a significantly larger proportion of non-vc backed spin-offs we re-set table 5 above to show the proportion of exits specifically from vc-backed spin-offs see table 6 below and note that ­ particularly for the 2001 ­ 2006 period covered by the uk study eth zurich s vc-backed spin-offs have not yet experienced any exit while the uk samples of similar size have all experienced at least one exit there seems an apparent problem with finding routes to exit for the eth zurich spin-offs at least for ipo s this may be explained by the specifics of the swiss capital market i.e the lack of a separate exchange with a streamlined admission process for small growth stocks but with sufficient liquidity and international clientele such as aim in the uk from an eth point of view the important question that this low proportion of vc/angel backing raises is whether the spin-offs lack access to sufficient vc/angel equity funding or whether a large proportion of them do not posses the characteristics that would make them interesting for vc/angel investment from the data above it is obvious that eth zurich creates a higher number of spin-offs per annum than even the large uk universities such as cambridge and oxford ­ as a matter of fact more than double a comparison of the spin-offs distribution by sector 9 in graph 12 below with data from another library house survey commissioned in 2005 by the british venture capital association bvca study 2005 shows that the uk universities spin-off activity is focused predominantly on life sciences 46 and it 39 while eth zurich s spin-offs are far more diverse life sciences it communications and ­ more recently ­ cleantech are the four major areas of vc investment over the last years evca 2007 however only half of eth zurich s spin-offs are part of these categories furthermore almost 20 are in the category `others that barely attracts vc investments and in which ­ as we have determined earlier ­ a certain proportion of spin-offs are `life-style businesses that are not seeking venture capital backing by definition this leaves materials and engineering sciences which it materials engineering/industrials telecom others 0 10 20 uk universities 30 eth 40 50 source bvca 2005 and data from eth s spin-off database 9 to fit with the categories in the bvca study we have grouped eth zurich s sectors biotech medtech and chemicals to `life sciences electrical eng mechanical eng and sensors to `engineering/industrials and micro/nanotechnology with material sciences to `materials these factors however cannot explain the whole difference eth still has a comparatively lower level of vc backing in the core sectors of vc investment i.e biotech 40 of spin-offs are vc/backed and it 20 one other area that may provide an explanation is the spin-off process itself in the scope of this study we cannot perform a detailed analysis of this process comparing eth with the leading uk universities but we can briefly highlight some of the key differences that we observed in the spin-off support process under a separate header below before that we briefly look at the average size of vc/angel investments attracted by eth zurich s spin-offs to that attracted by the uk universities here eth zurich is fairing actually very well its average of chf 6.4 million per vc/angel backed spin-off is almost 20 above the £2.3 million chf 5.3 million average for the 20 uk universities and only two of them have attracted more institutional equity funding per spin-off than eth zurich cambridge with an average of £5.5 million chf 13.0 million and ucl with an average of £4.3 million chf 10.1 million 19 18


p. 11

5 analysis 5 analysis key differences in the spin-off process ­ uk universities vs eth zurich the leading uk universities have entrusted their technology commercialisation activities to separate legal entities ­ e.g oxford university to isis innovation ltd imperial college to imperial innovations group plc and cambridge university to cambridge enterprise ltd ­ which are either owned by the university or ­ as in the case of imperial innovations ­ are public listed on aim with the university holding a minority interest these dedicated organisations manage all aspects of technology transfer following the 6-stage process and generating the outputs as outlined in the following illustration illustration 2 the generic technology trantfer process sourcing of ideas ip protection supporting proof of concept licencing ip agreements spin-off incubation investment growth and exit output inventions patents poc projects ip agreements spin-offs financial returns source adapted from imperial innovations 2007 part of these organisations mandate is their management of university challenge funds through which the universities invest in spin-offs at proof-of-concept stage and/or at seed stage while their dedicated technology transfer organisations act as investment managers imperial innovations further manages two funds for third party investors the carbon trust incubator fund and money made available under the waste resources management program of the uk government the university challenge funds each have a capital of £ 4 ­ 6 million chf 9 ­ 14 million and were allocated to the universities starting in 1999 by the uk department of trade and industry as part of an initiative to encourage spin-off creation challenge fund investments can reach £ 60 000 chf 140 000 per spinn-off at proof-of-concept stage and £ 250 000 chf 600 000 at seed-stage and each of these funds currently holds between 50 and 80 equity investments in spin-offs cambridge enterprise isis innovations imperial innovations 2007 this system of university challenge fund is widely applied across uk universities and explains why ­ on average ­ 80 of new spin-offs every year 20 have equity participation of various levels by their university he-bci 2002 ­ 2007 a striking feature of these university technology commercialisation organisations is their incubation support activities and ­ in particular ­ the active search and recruitment of experienced managers and non-executive directors to lead the spin-offs in which they hold a stake imperial innovations for example continuously develops a pool of individuals that will potentially take leadership roles in their future spin-offs they do this through an `entrepreneur in residence program as well as through supply relationships with recruitment firms imperial innovations 2007 many of imperial college s new spin-offs are lead by experienced managers ceo s sales and bd while graduates normally take technical roles cambridge and oxford have similar programs and maintain close relations to `serial entrepreneurs with whom they have successfully worked in previous spin-offs a final important task is the development of vc/angel relations both oxford and cambridge actively manage membership-based networks of potential investors isis angels network cambridge enterprise venture partners to which they regularly pitch new investment opportunities they frequently hold events introducing research trends and new technologies bringing researchers managers and investors together isis spinners venturefest cambridge university technology venture conference imperial innovations ii ­ after it has raised £25 million in its ipo in 2005 ­ is essentially acting as venture capital investor itself it invests into imperial spinoffs that have gone through its incubation program as well as in other companies however in its investments it systematically seeks syndication by other vc investors and ­ in 2007 ­ managed to attract 2£ of third party investments for every 1£ of its own money imperial 2007 ii therefore relies on close relations to the vc/angel community as much as oxford and cambridge do in eth zurich the spin-off support process historically focused on developing ip licensing agreements with the spinoffs and facilitating relations with providers of infrastructure technopark eth it department etc research funding through cti and ­ to a limited extent ­ equity capital eth zurich has not invested into its spin-offs but has granted small loans chf 50 000 to 100 000 which ­ in two instances ­ it converted to equity as part of a capital restructuring more recently eth zurich has increased its focus on spin-off support and now also provides consultancy business plans access to business networks/investors and has started to take small equity stakes at start-up stage eth does however rarely invest cash but rather designs its technology licensing terms in a way that the equity stake partly compensates for the license fees payable there is no actual university challenge fund at eth10 but ­ together with other swiss universities ­ eth teams can take part in a bi-annual business plan competition organised by mckinsey with eth zurich and sponsored by 19 large swiss corporations the pricemoney of total chf 150 000 the winner receives chf 60 000 is awarded to the teams personally independent of them later creating a business or not separately ­ cti ­ the swiss innovation promotion agency in the federal department of economy provides research/proof-of-concept grants business coaching entrepreneurial training through venturelab and supports a private initiative of a membershipbased network of business angels vc firms and other investors which organises pitching and networking events on a fairly regular basis cti s services are open to any start-up in switzerland in 2000 mckinsey and eth zurich initiated the creation of venture incubator vi partners a venture capital firm to support university spin-off s as well as other promising startup companies with capital coaching consulting and networks and specifically with the objective to close an early stage funding gap they observed its vi partners fund raised chf 101 million from 10 blue-chip companies based in switzerland with the aim to invest into university spin-offs and non-university start-ups over the last few years as many other vc firms in europe vi partners however appear to be shifting their investment focus to later expansion stage.11 another noteworthy source of start-up stage financing in this context is zürcher kantonalbank zkb who offer mostly mezzanine/convertible loans of chf 100 000 to max 500 000 size that convert at a pre-determined valuation at zkb s option this facility is part of a start-up business support program under which zkb invest every year a total of chf 10 to 13 million 10 11 5.3.2 vc/angel backing compared to all start-up companies in switzerland dr maurice pedergnana the secretary general of the swiss private equity association seca estimates that local and international venture capital firms and business angels invest into swiss start-ups on average between chf 50 ­ 60 million in seed and start-up rounds and chf 70 ­ 80 million in expansion rounds per annum seca s annual reports seca annual reports show that an average of 120 firms per annum have received seed/start-up/expansion funding over the last 5 years and ­ although not all seca members may have reported their transactions the number of start-up companies that receive vc/angel backing per year is certainly lower than 200 indicating that the average investment per startup may be just under chf 1 million eth zurich s 3-4 spin-offs per year that obtain vc/angel backing therefore manage to raise more than 6 times as much vc financing as the average vc/angel-backed swiss company does in fact the total of eth spin-offs may attract around 20 ­ 25 of all vc/angel investments in switzerland ­ if we linearly extrapolate 12 the data received in our survey as shown in graph 13 graph 13 eth zurich spin-offs sample of 82 equity raised by year and provider 50 40 chf million 30 20 10 0 others eth zkb vc angels founders ff 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 source data from own survey 12 the establishment of such a fund has been proposed by eth transfer in 2006 according to venturexpert 4 of the 5 last investments by vi partners were in expansion rounds endoart apr.08 silentsoft/feb.08 nemerix/sep.07 ganymed pharmaceuticals apr.07 obviously years 1998 and 1999 are not representative of the average equity raised by eth zurich spin-offs as the data sample starts with 1998 and ­ as mentioned before ­ spin-offs on average take 2 years to raise a first vc/angel round the average vc/angel investment per year from 2000 to 2007 is chf 18.9 million in our sample we have 71 of vc-backed companies represented which could mean that all vc/angel backed eth zurich spin-offs may have received equity investments of chf 26 ­ 27 million per annum 21


p. 12

5 analysis 5 analysis furthermore the sharp increase in vc/angel funding raised by eth zurich spin-offs from a low in 2005 also seems to counter the trend of diminishing seed/early-stage investments in switzerland seca 2007 and in europe as a whole evca 2007 comparing performance in terms of exits by ipo and trade sale will likely not be very conclusive as the events are too few on both sides switzerland experiences every year between 8 and 12 ipo s and around 200 m&a transaction seca 2007 5.4.1 methodology we assessed the financial performance of the spin-offs in our sample of 82 by identifying the ex-post return to equity invested each company s equity was evaluated individually and we treated ­ in a slight simplification ­ all stock as common ordinary stock in our questionnaire we had requested information about all financing events trade sales and equity offerings as well as the key p&l performance indicators and balance sheet values for the financial year 2007 our calculation of the irr is based on the equity investments made in each company since its incorporation and the present value of the equity at date of exit or ­ if no exit occurred ­ on dec 31st 2007 based on comparables financing event or cost of investment more specifically the equity value for the 82 spin-offs was determined as follows · exits one spin-off has experienced an ipo on swx and the stock price at the end of the first day of trading serves as basis for our valuation 5 further spin-offs have gone through a trade sale in which at least a substantial part of the equity was purchased by a third party in this case our valuation is based on the purchase price paid for the equity stake at the date of sale · financing events the valuation of 13 companies is based on a recent financing event ­ usually a significant equity investment by a venture capital firm according to bvca guidelines we use the implied post-money valuation of the latest financing round as long as that financing round has not taken place more than 18 months prior to dec 31st 2007 · comparables for 29 companies which have stable operations and steadily growing turnover and profits but have 22 5.4 return on equity · · neither gone through a trade sale recent financing event nor an ipo we use multiples to estimate the firm s value the comparables utilised were p/e ev/ebitda ev/ebit and ev/revenues we used average industry multiples for companies listed in switzerland as obtained from the capital iq database and applied them to the current performance indicators fy2007 this rather conservative approach instead of forward multiples and projected earnings reduces the risk of exaggerated valuations but we also believe this approach is more in line with industry s fair-value guidelines moreover as the spin-offs valued are all private companies we applied a 30 liquidity discount to the value of their equity for reference the bvca recommend a minimum liquidity discount of 25 an overview of the multiples used is in table 7 in appendix 2 cost 26 companies for which none of the information above is available or appropriate we use the cost of the initial investment and subsequent investments all paid-in equity as benchmark zero value the 8 spin-offs that either have gone bankrupt or have ceased commercial operations are assigned a zero value giving a negative return to equity-holders spin-off a fast growing profit generating biotech company as it has made considerable progress since its latest vc round we utilise industry-specific multiples to value it 5.4.2 equity value created over the period of 10 years from 1998 to year-end 2007 the 82 eth zurich spin-offs attracted chf 169 million of equity investments by the founders of the companies vc firms or angel investors the absolute and accumulated return on these investments calculated with the methods described above amounted to chf 650 million at year-end 2007 representing a money-multiple of 3.84 over an average investment period of approx 3.7 years this absolute return is driven by few large `caps i.e the spin-off with the largest valuation makes up for 36 of the total the top 3 for 73 and the top 10 for 91 in terms of sectors the highest absolute return was generated in biotech and pharmaceutical followed by electrical engineering electronics and medtech diagnostics see graph 15 in appendix 1 as noted earlier 91 of equity investments in absolute numbers were made by vcs and business angels with chf 484 million a much smaller part 75 of the absolute returns accrued to vcs while founders and other investors claimed chf 166 million 25 62 of these returns have effectively been realised through an exit the average ownership stake of vc s at the time of valuation exit or year end 2007 was 51 of equity on a fully diluted basis including preferred stock and convertible debt ­ there is however considerable variance among the companies as the standard deviation of ownership stake of vc firms is 27 our data reveals as might be expected that the ownership stake of venture capitalists increases as the companies go through more financing rounds the average stake of vc s at exit trade sale or ipo was 61.3 5.4.3 returns on equity using the methodology described above our calculations result in a 43.33 pooled internal rate of return irr for the sample of 82 eth spin-offs we calculated separately the returns for founders and for vc s/angels as per graph 16 below non-vc-backed founders experienced an irr of 40.5 while the vc-backed founders with 78.1 almost double vcs who made most of the absolute returns experience an irr of 37.5 ­ the lowest relative returns these returns are somewhat dependant on one large and highly successful transaction without it the total irr would be 25.1 and the vc s irr 20.0 only graph 16 eth zurich spin-offs return on equity by investor category ­ pooled irr 80 60 irr 40 20 10 0 founders and others non-vc backed 58 founders and others vc/angels vc backed 24 total vc backed all 82 70 50 30 source data from own survey graph 14 eth zurich spin-offs sample of 82 method used for equity valuation 35 30 25 20 15 10 5 0 number of spin-offs the quality of returns in individual spin-offs when examining the distribution of the irrs for each investment in the sample of eth zurich spin-offs as shown in graph 17 below we notice that among the 24 companies in which they have invested vcs/angels experience 100 and higher irr in 6 instances between 10 and 100 irr in 5 instances and negative returns in 5 instances in a further 8 instances vc/angel returns are zero because we valued the investment at the level of the most recent vc-round meaning that returns can still go either way in the future the `fattail to the right hand side obviously is the reason for the high positive pooled irr the maximum irr observed in vc investments is 887 graph 17 eth zurich spin-offs vc s returns in each of 24 investments vc returns irr in each of 24 investments 1000 800 600 400 200 0 ­200 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 n 82 comparables cost/book value financing event trade sale ipo zero value source data from survey the `pooled irr for the portfolio of 82 spin-offs is calculated by combining/pooling all investments and the valuation i.e investments are added as negative cash flows to the year in which they were made while exits are added as positive cash flow to the year they occurred and the equity value of the other spin-offs is added to 2007 in appendix 2 we have attached an example of the valuation of an individual n 24 source own calculations with data from survey 23


p. 13

5 analysis 5 analysis more surprisingly the founders of vc/angel backed companies experience in 8 of instances a negative irr on their investment and none of them was actually a total loss in 30 of their investments the irr is between zero and 100 and in 62 above 100 with a maximum observed irr of 4887 this seems to provide quite clear evidence that the founders downside risk in spin-offs with vc-backing is very small while the upside potential for extraordinary returns is virtually unlimited in contrast non-vc backed founders experience relatively more events of zero or negative irr as graph 18 below makes clear graph 18 eth zurich spin-offs distribution of founders returns in vc-backed and non-vc backed spin-offs 70 of sub-sample 60 50 40 30 20 10 0 -75 -75 -50 -25 0 50.01 -25.01 -0.01 24.99 founders irr 25 49.99 50 74.99 75 99.99 100 vc backed non vc-backed the performance of spin-offs which were backed by various vc funds it is also common to benchmark fund performance by the so-called dpi ratio i.e realised multiple distributions of cash or stock to investors/paid-in-capital the so-called rvpi i.e unrealised multiple residual value/paid-in-capital and tvpi i.e total value multiple dpi+rvpi/paid-in-capital 13 table 8 below compares pooled irr and dpi of eth spin-offs with the averages for the venture capital industry based on data collected by evca14 and by ourselves from venturexpert as follows · pooled irr and average dpi of 695 european venture funds established 1980 ­ 2006 however the pooled cash flows used for the irr calculations are from 1980 until june 2007 · pooled irr and average dpi of 1204 us venture funds established 1969 ­ 2006 however the pooled cash flows used for the irr calculations are from 1969 until june 2007 as the data from venturexpert is net of carried interest the vc s share of profits we adapt the returns on investment experienced by the vc s by calculating theoretical returns to limited partners lp s on the hypothetical `fund of spin-offs this calculation assumes an average 20 carried interest on every deal and results therefore in a lower irr and dpi venturexpert data is also net of gp fees however venturexpert deducts the fees from the investment cash flows and therefore actually slightly overstates the fund returns to lp because of the lower base we therefore do not adjust for fees table 8 european and us venture capital fund performance 1969/80 to 2007 vs returns observed in eth zurich s spin-offs pooled irr europe top 10 european venture funds 16.5 5.0 average dpi 1.55 0.61 standard deviation 28 46 na na numbers of funds 695 1204 120 1 1 70 n 82 this comparison appears to indicate that the sample of eth zurich spin-offs generates an above average return however certain caveats need to be considered the data above is collected over a long period of time including several business cycles with quite different circumstances in financial markets the high standard deviation especially for the us data also indicates a wide variance between individual fund performance and average fund performance over different time periods we have therefore grouped funds into 10 cohorts based on their performance and it becomes evident that the performance of the eth zurich sample of spin-offs is slightly lower than the top 10 percentile of all vc funds in the us but clearly superior to returns experienced on by the top 10 percentile of funds in europe during the last three decades we then narrowed this comparison to the time period in which the spin-offs were incorporated 1998 to 2007 in order to eliminate a possible bias arising from different economic conditions prevailing in earlier periods as shown in table 9 below we looked at pooled irr for the period of 1998-2007 of all us european uk and swiss vc funds founded between 1998 until 2006 table 9 international venture capital fund performance for funds created 1998 to 2006 vs returns observed in eth zurich s spin-offs pooled irr european venture funds -1.4 2.9 average dpi 0.26 0.49 1.06 0.14 0.29 0.31 0.31 2.21 1.77 0.5 numbers of funds 439 467 46 16 63 6 1 1 2 43 comparison to the average and the top 10 percentile of funds in the vc industry created over the same period however two important aspects should be kept in mind this data covers the period of the boom and subsequent bust also it includes funds established in recent years which predictably have mainly had capital outflows and only few exits yet15 how dramatic the impact of the bubble burst was on cumulative returns to us and european venture funds is shown in graph 19 below we used pooled cashflows from the same set of data as used in table 9 above not only did valuations change dramatically when the bubble ended but investments in venture capital had been at all time high during the boom one of the reasons why the performance observed in the sample of eth zurich spin-offs is so strong compared to venture funds in europe and the us is probably the fact that only 2 spin-offs or 11 of the total equity investments in that period represent ` type of companies that lost significant value when the bubble burst graph 19 returns of us and european vc funds 1998 ­ 2007 100 accumlated irr 80 60 40 20 0 ­20 2 3 05 06 /3 1 12 99 1 9 1 0 1 0 1 0 1 04 12 /3 /3 /3 12 /3 12 /3 /3 /3 /3 1 /3 1 1 1 07 01 8 0 us vc funds european vc funds source own calculations with data from survey the obvious explanation is `selection bias i.e the notion that vcs/angels will back only firms that promise the potential of creating extraordinary returns but there will also be an element of `self-fulfilling prophecy in the sense that vc-backed firms do get more financial and better management/boardroom support than non-vc-backed firms and ­ therefore ­ they stand a better chance to thrive 5.4.4 returns compared with vc funds in the us and in europe the method described above ­ the pooled irr ­ is the most common method in the venture industry to compare returns on invested capital of venture capital funds even though the portfolio of eth spin-offs does not represent investments of one vc fund average fund performance calculated by the same method present a convenient and interesting comparison to benchmark the overall performance of these spin-offs against the performance of the venture industry particularly 24 12 europe top 10 us top 10 us venture funds switzerland vc funds switzerland top 10 uk vc funds uk top 10 10.1 19.2 -1.2 -0.6 -3.6 6.2 source data from venturexpert eth zurich spin-offs gross vc returns eth zurich spin-offs theoretical lp returns 37.5 30.0 us top 10 us venture funds 37.9 37.5 30.0 15.8 3.45 2.21 1.77 1.16 eth zurich spin-offs gross vc returns eth zurich spin-offs theoretical lp returns na na source venturexpert source evca institute and venturexpert 13 14 david bernard evca institute benchmarking private equity performance november 2007 david bernard evca institute benchmarking private equity performance november 2007 the cyclicality of the venture capital business does indeed show in this comparison but actually to the favour of eth zurich s spin-offs the pooled irr and realised multiple for the sample of eth zurich spin-offs is extraordinarily high in the next analysis in graph 20 below shows the average pooled irr for us and european funds calculated separately for each vintage year of funds if we compare the sample of eth zurich spin-offs to these funds by vintage year we see that the pooled irr for funds established in the early and mid 90 s in the us is comparable to the performance of those eth zurich spin-offs that were backed by venture capitalists however the spin-offs do much better than the funds established in the late 90 s presumably ­ again ­ because of the 15 although they are valued along the same fair value guidelines applied by us 12 12 12 12 12 25


p. 14

5 analysis 5 analysis bubble and its devastating impact on returns to venture capital investments particularly in the us this entire analysis shows how cyclical the venture capital industry is and also helps to understand why many venture funds have over the last few years scaled down their early stage investment activity and started to focus on expansion stage or buyout transactions 3i being a prominent recent example when it announced in march 200816 that it was withdrawing from seed and early-stage investments altogether graph 20 returns of us and european vc funds by year of fund creation ­ 1980 ­ 2006 100 80 average pooled irr 60 40 20 0 -20 80 19 81 19 82 19 83 19 84 19 85 19 86 19 87 19 88 19 89 19 90 19 9 19 1 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 0 20 1 0 20 2 03 20 0 20 4 0 20 5 0 20 6 07 19 us vc funds european vc funds source data from venturexpert as a conclusion of this analysis of vc returns we can say that a fund created in 1998 and which had systematically invested in eth zurich s spin-offs would have outperformed the average us and european venture capital funds by a margin one of the main reasons for this significant out-performance is likely the comparatively higher diversification among the eth zurich spin-offs in terms of technologies as shown in graph 14 in appendix 1 and ­ in particular ­ the relatively small exposure to the ` bubble 5.4.5 returns compared to swiss stock market returns the smi is switzerland s key equity index it represents about 85 of the free-float capitalisation of the swiss equity market because the smi is considered to be a mirror of the overall swiss stock market it is used as the underlying index for numerous derivative financial instruments consequently we use the smi index total return to compare the performance of the sample of eth zurich spin-offs to returns to capital invested in listed stocks over the same time period in the irr model described above we pooled all investments made in each year and returns were measured as value of stock at exit recent financing event or at year-end 2007 for our comparison with the stock market performance we use this model to estimate the returns as if the same amounts had been invested at the same time in the smi index instead of in eth zurich spin-offs we use the smi year-end closing prices and returns are measured for each company in the year of exit if occurred or at year-end 2007 this calculation results in an internal rate of return of 10.31 by year-end 2007 the observed irrs in eth zurich ­ spin-offs outperform the smi total return index by 27.23 gross or 19.68 if calculated against lp theoretical returns to verify these large abnormal returns we modelled the realised expected returns for vcs in the swiss market over the period 1998-2007 with a capm we used the annual returns of the smi total returns and the 10-year swiss treasury bill rates as risk-free rate as we do not have sufficient data to regress for betas ourself we test the model with two vc-fund betas observed in academic literature the 1.7 as estimated by cochrane 2005 and the 3.2 outlined by driessen et al 2008 who explain their higher beta with the fact that their time-series includes the years 2000 ­ 2003 which apparently amplified the covariance with market as per table 11 below our calculated average expected return er for beta 1.7 is 8.49 and for beta 3.2 is 13.42 giving further credibility to our hypothesis above that the eth zurich spinoff portfolio contains alpha in the magnitude of 20 ­ 25 table 10 capm for expected vc-returns in the swiss market year 1998 1999 2001 0.0270 rf smi ­ tr 7774.10 2000 2002 2003 2005 2007 2004 2006 average return 0.0304 8808.80 rm lin 1.7 e r with beta 3.2 0.0392 0.0338 10315.77 9439.00 0.1331 0.0715 0.2074 0.0320 8228.60 0.0265 6043.47 -0.2023 0.0929 0.1003 0.3665 0.0274 7306.29 -0.2656 -0.3676 0.1305 0.1621 0.0210 0.0251 10481.00 7707.93 0.2090 -0.4738 -0.7218 0.2110 0.0550 0.3367 -0.9202 0.0293 12371.88 0.3598 0.0743 0.6104 standard deviation 0.0051 2.92 12199.29 -0.0140 0.1866 0.1804 0.5969 0.1156 6.20 -0.0442 0.3194 0.2891 1.1051 8.49 -0.1091 0.6041 0.5221 13.42 source data from swx and swiss national bank 5.4.6 returns compared to other university spin-offs very limited data seems to exist publicly on returns to capital invested in university spin-offs consequently we are not able to offer a general comparison of the performance of eth spin-offs with similar university initiatives however we believe that a comparison with imperial innovations ii the technology transfer organisation of imperial college is both relevant and interesting ii was established in 1986 with the aim to protect and exploit commercial opportunities arising from the research base of imperial college primarily in the fields of science engineering and medicine in 2006 ii was publicly listed on aim but still has an exclusivity agreement until 2020 with imperial college to commercialise intellectual property that is developed within imperial college s research departments at the time of ipo ii had equity holdings in a portfolio of 58 spin-off companies 21 of which are early stage and 37 at a more advanced stage approximately 60 of its spin-off companies are focused on the engineering and technology sectors.17 according to the prospectus published in 2006 the fair value accounted for on the basis of bvca guidelines of ii s portfolio was approximately £19 million and £31.5 million in july 31st 2005 and 2006 respectively on july 31st 2007 the fair value of the portfolio was £53.7 million18 and by 31st january 2008 it was £51.7 million the publicly disclosed cash flow statements in the prospectus and ar 2006 and 2007 also reported the investments made in its spin-off companies and proceeds from the sale of investments by using the 2005 valuation of the portfolio subsequent cash flows before deduction of management fees and carry and the valuation by january 31st 2008 we calculated an irr of 29.4 during this period this is of course only a view over a 3.5-year time-window that does not account for the value creation prior to july 31 2005 but it provides a rough benchmark however what we find particularly interesting is that ii seems to experience similar above average returns to eth zurich s spin-offs with a similar portfolio of technologies i.e mainly bioscience and engineering technologies it should also be noted that according to the 2006 and 2007 financials 40 and 34 respectively of the fair value of investments was related to imperial s shareholding in ceres power which is a listed company 17 16 financial times march 26th 2008 18 imperial innovations group plc prospectus july 2006 imperial innovations group plc annual report 2007 26 27


p. 15

6 conclusions 6 conclusions 6 conclusions the data collected in our survey and our separate research provides strong evidence that eth zurich s spin-offs are more successful than normal start-up firms and are strongly beneficial to the local economy in line with observations internationally eth zurich s spin-offs have significantly higher survival rates create more jobs attract more vc/angel investments and provide higher returns on equity than swiss start-ups on average among the spin-offs themselves those with venture capital backing investments significantly outperform those without in terms of job and value-creation as graph 21 below makes quite obvious the bubble size gives an indication of the spin-offs valuation per 31st december 2007 graph 21 eth zurich spin-offs sample of 82 job and equity value creation by vc-backed and non-vc-backed firms 90 number of employees 80 70 60 50 40 30 20 10 0 0 1 2 3 4 5 6 7 8 9 10 11 vc-backed not vc-backed valuation size chf 20 mio of companies since its ipo until june 2007 imperial is a uk leader in the field of technology transfer and commercialisation of technology the founders entrepreneurs experience even higher returns than the vc s/angels and significantly higher returns if they obtain vc-backing for their company than if not in exchange the founders give up majority control in their companies as we have seen vc s average stake at exit was 61.3 and the prospect of loosing control over `their spin-off appears to be a concern to many young founders along with the fear of giving away that control too `cheaply however the difference in achievable returns is so dramatic that giving up control should really not be an issue unless the founders seek to create what we have called a `life-style business over a 7 year period ­ vc-backed founders entrepreneurs make over 5x higher capital gains in absolute than their nonvc backed peers i.e 57x their initial investment vs 11x vc angel investments therefore appear beneficial to all parties involved making a very clear case for eth zurich trying to attract more vc/angel interest in its spin-offs years since incorporation source own compilation with data from survey furthermore vc/angel investors experience considerably higher returns than the average vc fund during the last decades in europe and the us as we have demonstrated with data from venturexpert returns are comparable to the performance of the top 10 percentile of us vc funds we reviewed several studies in academic literature that have investigated the financial performance of vc investments and vc funds the performance of our sample `portfolio has been considerably higher than the average returns reported in all of these studies our calculation shows an out-performance of the smi total return index by 19 ­ 27 and abnormal returns in the range of 20 ­ 25 during the 1998 ­ 2007 period eth zurich spin-offs returns are comparable to and actually seem to exceed the returns in imperial innovation portfolio 28 as shown above the 130 eth zurich spin-offs incorporated in the 10-year period from 1998 to 2007 have created direct employment for a total of 918 persons or close to 8 jobs per surviving spin-off in comparison the average swiss start-up company creates ­ over 5 years which is the average age of the eth spin-offs ­ less than half as many jobs 3.65 many of these jobs are highly qualified over 40 of the spin-off employees are eth graduates and offer part-time employment the average job equals 0.81 fte requiring a flexible well-educated and self-motivated workforce while providing very interesting career development opportunities in a highly dynamic work environment with total revenues of close to chf 250 million19 and personnel cost of an estimated chf 100 million including social contributions eth zurich s spinoffs out-source approx chf 120 million of goods and services and are likely to have caused the creation of at least 6.1 the value and benefits to the economy as a whole additional 500 indirect jobs20 at their suppliers and service providers a second major economic benefit is wealth creation the average team of spin-off founders ­ along with family friends and other private investors ­ invest equity of chf 110 000 in their spin-off if they have no vc or business angel backing them and can hope ­ on average ­ for a return of 40.5 on their investment resulting in a capital gain of chf 490 000 over a 5 year period with vc/angel backing that gain would be 4 times more over the same period obviously not every founder can expect to realise this level of returns but surprisingly about half can expect even higher returns and only approx 12 risk a negative return in total the 130 spin-offs have so far created capital gains to the investing entrepreneurs of at least chf 150 million21 of which however only 60 were realised at the same time vc s and business angels have made ­ in total ­ capital gains of over chf 350 millions based on the data received in our survey we have attempted to quantify the tax-revenues to the swiss federal cantonal and communal authorities generated by the 130 eth spinoffs and estimate ­ for 2007 ­ taxes paid of close to chf 18 million resulting from employees personal income taxes of chf 8.6 million see table 11 in appendix 1 and corporate income taxes of chf 9.3 million on an estimated total pre-tax income of chf 43 million22 this does not include taxation of institutional investors profits and gains of those private investors that may be subject to capital gains tax apart from their direct economic impact in job and investor wealth creation as well as tax revenue generation eth zurich s spin-offs also have a noteworthy indirect economic and social impact for example as catalyst of high-tech cluster formation and in the attraction/retention of world-class faculty to eth zurich itself to name two examples in a long list of likely benefits mentioned by shane 2004 a detailed discussion of these benefits would however lead beyond the purpose and scope of this study 6.2 benchmarking performance with other universities 20 21 22 19 total revenues in sample of 82 is chf 169 million we have linearly extrapolated this chf 120 million 250m ­ 100m ­ 30m profits of outsourced services goods x 40 100 000 chf per employee with 90 confidence average founders capital gain per spin-off is chf 1.83 million with a standard-error of chf 0.51 million total positive pre-tax income in our sample of 82 companies is chf 27.2 million we have linearly extrapolated that to reflect the total population of 130 spin-offs and applied a 21.5 corporate tax rate applicable for the canton of zurich concluding on our comparison of eth s zurich spin-off program to the ones of other institutions of higher education and ­ in particular ­ to those of leading uk universities for which we have found a large amount of relevant data we can say that the eth spin-offs are among the best in terms of survival rates but possibly create slightly less jobs ­ although the variance in data from other universities does not allow for a clear-cut conclusion furthermore a significantly lower proportion of eth zurich s spin-offs manage to obtain vc/angel backing than in the uk but those who do get backed attract 20 higher investments than the average uk spin-off a comparatively lower proportion of the total 130 spin-off population have experienced an exit through tradesale and those spin-offs that have obtained vc/angel backing are also being exited later than their peers in the uk along with the exceptionally high rates of return for both entrepreneurs and vc/angel investors this seems to indicate that the spin-offs that do get backed by venture capital are of high quality and attractiveness it is therefore surprising that only a relatively low proportion of eth zurich s spin-offs get vc/angel backing counter to no-arbitrage theory that would suggest these abnormal returns will attract more vc angel interest and the ensuing competition for good deals would drive down returns a possible explanation is ­ of course ­ the limited attractiveness of the swiss vc market to international investors partly for regulatory reasons23 partly for lacking exit routes ­ in particular a liquid small cap exchange however we have identified two other possible factors that may explain the low level of vc/angel backing the first one being the portfolio of sectors that eth zurich s spin-offs cover only 54 of eth zurich s spin-offs are created in sectors lifesciences it that can be regarded as mainstream area of investment focus for vc s a further 10 is in material sciences and manages to attract a reasonable level of vc interest the remaining areas ­ in particular engineering sciences ­ seem underserved by vc s and angels however one should not conclude that spin-offs in these areas are not desirable as demonstrated by one specific spin-off 23 a london-based vc mentioned for example the taxation of stock-options grants to management as a major concern 29



no comments yet