Venture Capital Incubation Period Underpricing and Firm Performance

Venture Capital Incubation Period, Underpricing, and Firm Performance NTU International Conference on Finance December 2006 Scott Hsu

Overview • Definition: Founding date VC incubation period First VC investment date IPO date • The findings on the effect of the VC incubation period: - Firms with a longer incubation period are less underpriced at the IPO stage - Firms with a longer incubation period are more likely to survive three years after the IPOs - Firms with a longer incubation period exhibit better post-IPO operating performance

Also important to study the determinants of the VC incubation period because. . • VCs play a dual role of “incubators” and “short-term investors” - Incubators: Hellmann and Puri (2002) - Short-term investors: “Google’s founders would have preferred to wait longer to do their IPOs, but had to rush it because venture capitalists, …, wanted to cash in” -The Economists, November 27, 2004 • It is thus important to find a measure to examine whether and under what circumstance VCs behave as “incubators” versus “investors”, and consequences that follow

The significance of studying the determinants of the VC incubation period and its impact • The conflicting role of VCs in incubating startups - Screening and monitoring: Jain and Kini (1995); Hellman and Puri (2002); Wang et al. (2003) - Certification in the IPO market: Megginson and Weiss (1991); Lee and Wahal (2004); Ljungqvist and Wilhelm (2003) - Most studies in this vein focus on the “binary comparison” between VCbacked and non-VC-backed firms. However, the “magnitude” of the VC effect could matter for firm success • A large proportion of IPOs were backed by VCs in recent years - 50. 33% of IPOs were backed by VCs in 2000; 34. 58% in 2001 - Thus, the investment strategy of VCs could have a profound impact marketwise

Data and methodology Data • 1502 venture-backed IPOs that went public from 1980 to 2003 • SDC Venture Xpert database • Supplemented by SDC New Issues database, CRSP, Compustat, and Jay Ritter’s firm founding dates and underwriter ranking data Methodology • OLS to study both the determinants and effects of VC incubation period • Robustness check: Heckman’s two step procedures and survival analysis

VC incubation period and investments over time Period Number of firms (percentage) Mean incubation period (years) Median incubation period(years) Mean of VC investments (thous. ) Median of VC investments (thous. ) 1980 -1990 342 22. 77% 3. 37 2. 72 18, 574. 02 8, 484. 85 1991 -1998 697 46. 40% 4. 58 3. 82 26, 860. 45 16, 722. 93 1999 -2000 381 25. 37% 3. 38 2. 95 59, 416. 11 41, 687. 29 2001 -2004 82 5. 46% 5. 03 3. 97 77, 888. 46 49, 098. 20 Total 1502 100. 00% 4. 02 3. 28 35, 951. 32 20, 199. 43

Univariate result: VC incubation period and IPO underpricing High incubation period Mean (Median) Low incubation period Mean (Median) Non-bubble period 11. 79% (5. 71%) 15. 10% (8. 54%) Bubble period (1999 -2000) 57. 68% (21. 61%) 82. 97% (38. 71%) High (Low) incubation period group: Firms whose incubation periods are ranked in the top (bottom) quarter of the whole sample

Determinants of the VC incubation period Hypothesis Market timing (VCs as short-term investors) Literature Lerner(1994) Pagano et al. (1998) Michelacci and Suarez(2004) Information asymmetry Gompers (1995); Cumming and need of monitoring Mc. Intosh(2001) (VCs as incubators) Prediction Market-to-book ratio (-) Research intensity (+) Hightech dummy (+) Early stage dummy(+)

Determinants of the VC incubation period Dependent variable: VC incubation period (year) (1) (2) (3) (4) (5) (6) Coefficient S. E Intercept 4. 43** 1. 97 4. 72** 1. 97 4. 39** 1. 96 5. 28*** 1. 98 4. 92** 1. 97 5. 21*** 1. 97 Industry market-tobook ratio -0. 31*** 0. 06 -0. 33*** 0. 06 -0. 32*** 0. 06 -0. 33*** 0. 06 -0. 31*** 0. 06 -0. 32*** 0. 06 Industry research intensity 0. 34* 0. 20 0. 23 0. 20 0. 35* 0. 20 0. 23 0. 20 0. 24 0. 20 Industry tangibility ratio -1. 32 2. 01 -0. 68 2. 01 -1. 50 2. 00 -0. 82 2. 01 -0. 90 2. 01 Herfindahl Index -1. 07** 0. 51 -1. 22** 0. 51 -0. 92* 0. 52 Industry capital intensity -10. 25*** 3. 35 -11. 23*** 3. 36 -10. 04*** 3. 42 Lead VC age > 4 -0. 44** 0. 18 -0. 45** 0. 18 -0. 48*** 0. 18 Industry fit -0. 20 0. 21 -0. 23 0. 21 Hightech dummy 0. 61*** 0. 17 0. 55*** 0. 17 0. 57*** 0. 17 0. 54*** 0. 17 0. 563*** 0. 17 0. 51*** 0. 17 Early stage dummy 1. 95*** 0. 26 1. 94*** 0. 26 1. 91*** 0. 26 1. 97*** 0. 26 1. 930*** 0. 26 1. 93*** 0. 26 Middle stage dummy 0. 74*** 0. 29 0. 75*** 0. 29 0. 73** 0. 29 0. 76*** 0. 29 0. 73** 0. 28 0. 74*** 0. 28 Adj R-Sq 0. 0713 0. 0735 0. 0766 0. 0764 0. 0812 N 1474 1474

Firm Survival and the VC incubation period • Definition of non-survivors: Firms that are delisted within 3 years after the IPO for reasons other than mergers and acquisitions Number of observations Mean (Median) VC incubation period (Years) Survivors Non-survivors 1168 288 4. 15 (3. 36) 3. 63 (2. 82) T-test 2. 53**

Firm survival probability and the VC incubation period

IPO underpricing and the VC incubation period • The IPO underpricing decreases with the length of the VC incubation period • The effect of firm age at IPO on the underpricing mainly come from the incubation period, not from the period before the VCs’ investment

IPO underpricing and firm age at IPO (1) (2) (3) (4) (5) (6) coefficient S. E Intercept -1. 64 13. 9 6 0. 53 14. 76 -4. 39 14. 24 2. 17 16. 18 3. 80 16. 75 -1. 44 14. 67 Log of firm age at IPO -8. 17** 3. 17 -9. 60*** 3. 43 -7. 97** 3. 17 -8. 33** 3. 32 -9. 12** 3. 56 -7. 41** 2. 94 Underwriter ranking 2. 91** 1. 43 2. 69* 1. 44 2. 89** 1. 43 3. 04** 1. 46 2. 99** 1. 49 2. 61** 1. 21 Share overhang 1. 58*** 0. 29 1. 54*** 0. 29 1. 62*** 0. 29 1. 57*** 0. 29 1. 54*** 0. 29 1. 62*** 0. 28 Ln(firm assets) -0. 34 1. 66 -0. 19 1. 66 -0. 34 1. 77 Ln(firm sales) 0. 65 1. 26 0. 23 1. 38 Lead VC age >4 -1. 10 4. 42 0. 21 4. 19 Industry fit 12. 94*** 4. 99 13. 25*** 4. 67 Hightech dummy 10. 09*** 3. 82 9. 91** 3. 93 10. 39*** 3. 81 9. 88** 4. 10 9. 70** 4. 21 9. 06** 3. 80 Bubble dummy 66. 59*** 4. 37 67. 65*** 4. 52 65. 38*** 4. 39 68. 11*** 4. 52 69. 01*** 4. 67 67. 60*** 4. 30 Number of IPOs 0. 00 0. 01 0. 01 Market-to-book -1. 40 1. 79 -1. 38 1. 84 -1. 01 1. 64 Research intensity -7. 93* 4. 71 -7. 93 5. 43 -8. 38* 4. 45 Herfindahl Index 13. 65 13. 56 7. 71 13. 75 13. 82 12. 07 Early stage dummy -0. 55 6. 55 -0. 29 6. 65 0. 29 5. 98 Middle stage dummy -7. 00 7. 06 -7. 40 7. 11 -6. 22 6. 47 Adj R-Sq 0. 2800 0. 2871 0. 2837 0. 2877 0. 2911 N 932 878 932 915 862 994

IPO underpricing and VC incubation period (1) (2) (3) (4) coefficient S. E Intercept -2. 49 13. 7 -1. 52 14. 43 -5. 17 14. 02 -0. 30 15. 70 1. 11 16. 23 -4. 15 14. 27 Log of VC incubation period -8. 32** 3. 49 -8. 80** 3. 59 -8. 05** 3. 48 -9. 82*** 3. 77 -10. 19*** 3. 90 -8. 93*** 3. 39 Log of firm age at first VC investment date -3. 27 2. 05 -4. 03* 2. 21 -3. 21 2. 05 -2. 49 2. 25 -2. 84 2. 36 -2. 18 2. 10 Underwriter ranking 3. 11** 1. 42 2. 89** 1. 44 3. 10** 1. 42 3. 15** 1. 45 3. 11** 1. 48 2. 68** 1. 21 Share overhang 1. 58*** 0. 29 1. 55*** 0. 289 1. 62*** 0. 29 1. 55*** 0. 29 1. 53*** 0. 29 1. 60*** 0. 28 Ln(firm assets) -0. 67 1. 65 -0. 52 1. 64 -0. 53 1. 76 Ln(firm sales) 0. 25 1. 25 -0. 10 1. 36 Lead VC age >4 -1. 19 4. 41 -0. 14 4. 18 Industry fit 12. 56** 4. 97 13. 20*** 4. 67 Hightech dummy 9. 77** 3. 80 9. 68** 3. 92 9. 72** 3. 80 9. 44** 4. 08 9. 36** 4. 19 8. 91** 3. 79 Bubble dummy 67. 30*** 4. 34 68. 33*** 4. 50 66. 14*** 4. 36 68. 80*** 4. 49 69. 64*** 4. 64 67. 80*** 4. 29 Number of IPOs 0. 00 0. 01 0. 00 0. 01 Market-to-book -1. 93 1. 79 -1. 97 1. 84 -1. 32 1. 66 Research intensity -7. 48 4. 65 -7. 72 5. 34 -7. 97* 4. 46 Herfindahl Index 10. 80 13. 53 4. 97 13. 71 12. 22 12. 09 5. 05 7. 09 5. 28 7. 24 5. 21 6. 53 -3. 87 7. 14 -4. 11 7. 20 -3. 48 6. 59 Early stage dummy Middle stage dummy (5) (6) Adj R-Sq 0. 2817 0. 2884 0. 2851 0. 2842 0. 2905 0. 2914 N 938 884 938 921 868 995

The level of post-issue ROA and VC incubation period Dependent variable Mean ROA (year 0 to year 3) ROA(year 0) ROA(year 1) ROA(year 2) ROA 3(year 3) Coefficient S. E Intercept -37. 40*** 7. 87 -21. 57*** 6. 60 -40. 61*** 10. 60 -55. 07*** 9. 81 -41. 86*** 10. 10 Log of VC incubation period 5. 65*** 1. 82 6. 67*** 1. 52 6. 10** 2. 44 6. 03*** 2. 25 4. 80** 2. 33 Log of firm age at first VC investment date 4. 65*** 1. 21 5. 15*** 0. 98 5. 00*** 1. 60 3. 82** 1. 52 5. 62*** 1. 57 Ownership retention 15. 19*** 4. 85 2. 30 4. 35 15. 477** 6. 93 24. 71*** 6. 28 14. 39** 6. 29 Natural log of IPO offer size 8. 09*** 1. 44 6. 89*** 1. 18 9. 40*** 1. 92 9. 35*** 1. 80 7. 60*** 1. 85 Hightech dummy 3. 32* 2. 00 5. 96*** 1. 67 4. 44 2. 71 3. 60 2. 51 0. 28 2. 60 Bubble dummy -20. 04*** 2. 83 -22. 86*** 2. 24 -25. 42*** 3. 71 -28. 48*** 3. 51 -10. 96*** 3. 65 Industry median market-to-book ratio -3. 54*** 0. 86 -4. 40*** 0. 71 -4. 46** 1. 15 -2. 54** 1. 07 -2. 78** 1. 11 Industry median research intensity -13. 17*** 2. 30 -9. 77*** 1. 96 -17. 08*** 3. 14 -11. 57*** 2. 97 -14. 65*** 2. 95 Herfindahl Index -1. 37621 6. 46 -8. 01 5. 27 -6. 87 8. 52 10. 52 8. 11 8. 63 8. 35 Early stage dummy -9. 85*** 3. 67 -8. 35*** 2. 95 -9. 90** 4. 80 -11. 72*** 4. 53 -8. 15* 4. 76 Middle stage dummy -0. 74 3. 64 -0. 70 2. 95 -1. 20 4. 79 -1. 00 4. 50 0. 68 4. 72 Adj R-Sq 0. 1945 0. 2187 0. 1287 0. 1391 0. 1034 N 953 1300 1207 1086 963

Robustness Check 1. The effect of VC participation or merely natural maturation? A further investigation 2. Does there exist a potential selection bias of venture financing? -Heckman’s two-step procedure 3. A survival analysis of the determinants of the VC incubation period

“Age” effect is more pronounced for venture-backed firms Dependent Variable ROA (year 0) Ind. Variables S. E Intercept -9. 38 *** 1. 50 -33. 59 Ln(IPOAGE) 2. 91 *** 0. 41 VC dummy -22. 06 *** 1. 84 VC*Ln(IPOAGE) 5. 94 *** Underwriter ranking 2. 84 Ln(proceeds) Hightech Dummy Bubble Dummy *** 4. 83 -35. 18 3. 48 *** 1. 31 3. 10 -21. 75 *** 6. 01 0. 80 6. 56 ** 2. 59 *** 0. 22 3. 45 *** 0. 27 0. 43 2. 33 -1. 14 0. 76 -1. 83 *** 1. 06 -28. 11 0. 2327 4222 coefficient ROA(year 2) S. E -21. 90 Adjusted R-Square Number of observations coefficient ROA(year 1) S. E *** 5. 50 -25. 62 *** 4. 05 ** 1. 49 3. 54 *** 1. 08 -31. 49 *** 6. 86 -24. 36 *** 5. 05 9. 25 *** 2. 93 6. 83 *** 2. 16 0. 70 4. 29 *** 0. 79 3. 36 *** 0. 58 * 1. 40 1. 05 1. 60 -0. 26 1. 17 2. 48 -2. 11 2. 80 -3. 11 2. 05 *** 3. 50 *** 4. 02 *** 2. 95 -27. 73 0. 0430 3963 coefficient ROA(year 3) -13. 30 0. 0453 3525 coefficient 0. 0513 3074 S. E

Underpricing and VC incubation period: Heckman’s twostep procedure (1) (2) Variables coeff. S. E Intercept 9. 99 12. 04 17. 43 14. 57 Log of VC incubation period -8. 31*** 3. 18 -9. 43*** 3. 41 Log of firm age at first VC investment date -2. 84 1. 97 -2. 16 2. 13 Underwriter ranking 2. 37** 1. 19 2. 43** 1. 21 Share overhang 1. 78*** 0. 29 1. 75*** 0. 29 Hightech Dummy 8. 36** 3. 59 7. 73** 3. 84 Bubble Dummy 63. 22*** 4. 38 63. 64*** 4. 47 Market-to-book ratio -1. 77 1. 64 Research intensity -9. 07** 4. 57 Herfindahl Index 11. 75 11. 95 Early stage dummy 2. 84 6. 51 Middle stage dummy -3. 68 6. 56 -9. 37** 4. 71 Lambda -6. 47 4. 46 Selection Model Industry dummies included State dummies included Firms were founded after 1980 IPO year dummies 0. 66*** included 0. 06 0. 65*** included 0. 07

Level of post-issue ROAs and VC incubation period: Heckman’s two-step procedure ROA (year 0) ROA (year 1) Variables coefficient S. E Intercept -7. 38 6. 59 -26. 92*** 9. 21 -43. 33*** 10. 03 -33. 87*** 10. 86 Log of VC incubation period 6. 55*** 1. 44 6. 56*** 1. 99 4. 30** 2. 17 2. 77 2. 37 Log of firm age at first VC investment date 4. 76*** 0. 91 4. 80*** 1. 29 3. 65** 1. 44 4. 50*** 1. 58 Ownership retention 0. 67 3. 86 13. 26** 5. 32 19. 73*** 5. 66 12. 85** 5. 99 Log of IPO size 3. 08*** 1. 13 6. 32*** 1. 61 6. 04*** 1. 79 4. 49** 1. 94 Hightech Dummy 5. 45*** 1. 59 3. 33 2. 23 5. 28** 2. 44 2. 54 2. 65 Bubble Dummy -19. 60*** 2. 10 -23. 34*** 2. 99 -21. 17*** 3. 31 -3. 86 3. 65 Market-to-book ratio -4. 50*** 0. 66 -5. 84*** 0. 94 -3. 64*** 1. 03 -4. 17*** 1. 11 Research intensity -8. 90*** 1. 80 -10. 99*** 2. 50 -8. 85*** 2. 81 -13. 33*** 2. 95 Herfindahl Index -9. 33* 4. 92 -11. 89* 6. 94 11. 61 7. 79 13. 67 8. 55 Early stage dummy -5. 27* 2. 71 -4. 79 3. 82 -5. 36 4. 24 -3. 51 4. 75 Middle stage dummy 1. 15 2. 71 2. 32 3. 82 1. 72 4. 21 4. 25 4. 68 Lambda -1. 29 1. 78 -0. 48 2. 44 3. 11 2. 57 5. 12* 2. 75 ROA(year 2) ROA(year 3) Selection Model Firm's industry dummies included Firm's state dummies included whether Firms were founded after 1980 Firm's IPO year dummies 0. 70*** included 0. 06 0. 70*** included 0. 07 0. 69*** included 0. 07 0. 65*** included 0. 07

Determinants of VC incubation period-A survival analysis Weibull distribution (1) (2) Cox Proportional Model (3) (4) (5) (6) (7) coeff. S. E coeff. S. E Intercept 1. 73*** 0. 49 1. 60*** 0. 49 1. 92*** 0. 48 1. 77*** 0. 49 - - - Industry marketto-book ratio -0. 07*** 0. 01 -0. 06*** 0. 01 -0. 07*** 0. 01 0. 09*** 0. 02 0. 08*** 0. 02 0. 09*** 0. 02 Industry research intensity 0. 07 0. 05 -0. 08 0. 07 -0. 05 0. 07 -0. 06 0. 07 Industry tangibility ratio -0. 42 0. 50 -0. 25 0. 51 -0. 49 -0. 32 0. 50 0. 53 0. 71 0. 34 0. 72 0. 48 0. 71 Herfindahl Index -0. 21* 0. 12 -0. 26** 0. 12 0. 23 0. 17 0. 24 0. 18 Industry capital intensity -1. 89*** 0. 69 -1. 90*** 0. 68 2. 19** 1. 00 1. 93* 1. 02 Lead VC age > 4 -0. 13*** 0. 04 -0. 12*** 0. 04 0. 18*** 0. 06 Industry fit -0. 07 0. 05 -0. 08 0. 05 0. 09 0. 07 Hightech dummy 0. 13*** 0. 04 -0. 16*** 0. 06 Early stage dummy 0. 43*** 0. 06 0. 44*** 0. 06 -0. 48*** 0. 10 -0. 47*** 0. 10 Middle stage dummy 0. 16** 0. 07 -0. 13 0. 11 -0. 11 N 1473 1395

Conclusion • This paper finds that the strategic choice of VCs as “incubators” versus “short-term investors” determines the length of the VC incubation period, which has positive impacts on firms: - Less underpricing Higher surviving probability - Better operating performance • This paper suggests that VCs play a positive role in incubating startups • This paper provides implications on the issues of corporate governance—the internal conflicts of large shareholders could affect the performance of firms