Long-run performance of Asian venture-capital backed initial public offerings and the impact of venture capital reputation

(1)

Amsterdam Business School

June 2015

Master Thesis

Long-run performance of Asian venture-capital backed initial

public offerings and the impact of venture capital reputation

Program: MSc Business Economics - Finance

Author: Viet Dung Doan – 10824650

Supervisor: Dr. Ilko Naaborg

ABSTRACT

This paper studies the effect of venture capital participation on the three-year performance of Asian venture-backed initial public offerings (“IPO”) and compares that to the performance of their counterparts. Examining the period between 2001 and 2011, the empirical analysis shows that Asian venture-backed IPOs’ performance is not significantly different from that of Asian non-venture-backed IPOs but is better than the performance of non-Asian venture-backed IPOs. Furthermore, Asian venture-backed IPOs outperform the market when returns are value-weighted, implying that the excess returns of large IPOs are much higher than those of smaller IPOs. This study also examines the impact of venture capital reputation on Asian venture-backed IPOs’ three-year returns and finds a positive effect, although its magnitude is small statistically and economically.

(2)

1 Acknowledgements

I am grateful to many people who have helped me finalize my Master thesis in the past three months. First of all, I specially would like to express my gratitude to my supervisor, Dr. Ilko Naaborg, who has given me valuable advice on both my thesis and my career path. His endless guidance and supervision played an important role in my thesis progress and helped me learn new things beyond the topic I initially chose.

I also want to thank my family for their support not only in my Master program but also in my whole life. Without them, I would have much more difficulties in pursuing my dream.

Last but not least, I would like to thank all my friends in Amsterdam, who gave me a lot of help and encouragement during my thesis and also made my life in this city more meaningful in the past year.

(3)

2

The success of Alibaba’s initial public offering in 2014, the world’s largest IPO ever, has gained many attentions to venture capital (“VC”) industry in Asia. In the first half of 2014, Asia-Pacific region observed 217 IPO deals with a total value of $33.7 billion, accounting for respectively 36.90% (by number of deals) and 28.63% (by capital raised) of global IPOs (Ernst & Young, 2014). Moreover, China is the country with the second highest deal volume (108 deals in 2014), just behind the United States. Also in the same period, private equity (“PE”) and VC backed IPOs accounted for 33% of global volume and 64% of US IPOs deal volume. After the financial crisis, IPO activities have gradually risen again around the world in general and in Asia in particular.

Venture capital is a type of alternative investment focusing on start-up companies that have high potential but are in short of capital (Eckermann, 2006). Initiated in the United States and later spread to Europe, venture capital has only been existing in Asia for the last few decades. However, Asian venture capital industry has experienced a substantial growth over the years. In 2014, the number of Asian VC deals is estimated to be 591 deals with a total value of $11 billion, accounting for 20% of global venture capital deal (Preqin, 2014). The relative figures are even more impressive: in 2014, venture capital investment in Northeast and South Asia increases by approximately 60% compared to the previous year. Although investors have focused more on the US and European markets since the 2007 global crisis, current trend shows that venture capital industry in Asia still gains a lot of attention from other continents and is developing remarkably (Preqin, 2014).

A venture capital fund can exit from its investment in the entrepreneur through initial public offering, trade sale, secondary sale or in the worst case, through liquidation. Going public through an IPO is the exit option that often generates the highest value to venture capitalists (Gompers and Lerner, 2001); but one might question the welfare of venture-capital backed firms’ shareholders afterward, as well as the opportunity for a superior return. Additionally, among the VC industry there is a large discrepancy regarding size, age, past performance, etc. of venture capitalists. All of these factors can lead to differences in the performance of VC-backed IPOs. For that reason, this thesis examines the influence of venture capitalists’ reputation on long-run post-IPO performance of venture-capital backed firms.

Many studies have examined the local bias or similarity bias of investment (Coval and Moskowitz, 1999; Huberman, 2001), where investors tend to prefer investing in local firms or firms with some familiar characteristics with the attempt to reduce information asymmetry.

(5)

4

Similarly, Franke et al. (2006) and later Cumming and Dai (2010) confirm the existence of local bias in venture capital investment – most US VC firms show a strong preference for local investments although the magnitude may vary depending on each VC’s characteristics. Thus, it is worthwhile to pay attention to geographic uniqueness of venture capital investment as it may play a noticeable role in the decision making process of VC investors.

With regard to the analyzed region, the thesis focuses on Asia because most research has been done on firms in North America and Europe. Examining the US market, Ritter (1991) and Loughran and Ritter (1995) report the long-run underperformance of IPOs in overall while Brav and Gompers (1997) show the outperformance of venture-backed IPOs. Regarding the European market, a study from Bergström, Nilsson and Wahlberg (2006) shows that IPOs backed by private equity outperform non-PE-backed IPOs in all time horizons. There has been some research about Asian countries but they show inconsistent results regarding IPO long-run performance (Dawson 1987; Kim et al. 1995). Also, research on Japanese VC-backed IPOs (Hamao et al., 2000) points out their long-run underperformance, which is in contradiction to the conclusion on US VC-backed IPOs of Brav and Gompers (1997).

Asian firms and the emerging market they are operating in have many distinct and diverse characteristics (culture, regulation, politics etc.) that may result in different performance compared to the average of global venture-backed IPOs. According to Preqin (2014), the Asian venture capital industry is promising with the reopening of IPO market in China or the dynamism and integration of ASEAN region. Given the better outlook of this region’s countries as well as its less developed VC market, Asia should gain more attention from the academics. Regarding the variable of interest, reputation is taken into consideration since it is one of the most important factors that differentiate firms within VC industry. However, until recently most papers, while examining venture capital investment, treat VC firms indifferently. This method ignores the fact that reputation can bring many competitive advantages to firms (Shapiro, 1983). Some studies have paid attention to the impact of VC reputation (Nahata 2008; Hamza and Kooli 2011; Krishnan and Masulis 2011) and reported a significant influence on portfolio companies’ long run performance. Therefore, with the crucial importance of venture capitalists to young portfolio companies, it is interesting to study how VC firms with different reputations can affect post-IPO performance of firms that are backed by them.

Overall, the research aims to answer the question:

Do Asian venture-backed IPOs outperform or underperform their counterparts and is there

(6)

5

With this target, the thesis helps clarify the inconclusive opinions about Asian VC-backed IPOs’ performance. If Asian VC-backed IPOs outperform, investors can have a new way to build more profitable portfolio where a higher proportion of fund is allocated to Asian IPOs backed by highly ranked venture capitalists. Also, this outperformance can lessen the local bias of US or European VC firms and encourages them to pay more attention to Asian market. On the contrary, when the reverse is true, investors should carefully evaluate the risk of their investments in the Asian venture capital market. Another aspect should be considered is that entrepreneurs, based on the result regarding VC reputation, should seek for a reputable VC (even at the expense of more-discounted offer price and less favorable contract); and VC firms also need to pay more attention to retaining and improving their reputation in the market. With respect to the methodology, the thesis uses Ordinary Least Squares (OLS) regression with the inclusion of fixed effects to analyze Asian VC-backed IPOs that took place between 2001 and 2011. Three-year buy-and-hold abnormal return (BHAR) is used as the indicator for IPO long-run performance. The results are then compared to those of venture-backed IPOs in other parts of the world. Furthermore, the role of venture capitalist’s reputation on this performance is examined by the regression of post-IPO abnormal return on the level of VC reputation. A set of control variables is used, including firm’s age, IPO size, industry, country, IPO year, pre-IPO multiples, underwriter’s reputation, etc. Also, robustness tests are conducted to check the consistency of the results obtained from the main regressions.

The current study reports three main empirical findings. Firstly, Asian VC-backed IPOs do not significantly underperform Asian non-backed IPOs but they outperform non-Asian VC-backed IPOs over three years post IPO when value weighting is used. Furthermore, they perform better than the market when returns are value weighted. Secondly, when the reputation of the venture capitalist is taken into account, a more prestigious VC has a positive impact on IPO performance, but that effect is negligible in both economical and statistical terms. Finally, the results with regard to China, Japan and South Korea are mostly in line with the findings from aggregate Asian market, with the only exception of the significantly positive influence of VC reputation on the performance of Japanese and South Korean IPOs.

The thesis has the following structure. Section 2 is an overview of existing literature on (i) initial public offering and its long run underperformance phenomenon; (ii) venture capital, venture capital reputation and the influence on VC-backed IPOs; and (iii) these practices in Asian market. The third section presents the hypotheses and methodology of the thesis while

(7)

6

the next two show descriptive statistics of data, regression results and robustness tests. Finally, section 6 includes conclusion and discussion regarding the obtained results.

2. Literature review

2.1. Long run underperformance of initial public offering 2.1.1. Overview of initial public offering

Initial public offering (“IPO”) is the process where firms sell their stocks to the public for the first time. According to Berk and DeMarzo (2013), there are three mechanisms that companies can use to do their IPOs: firm commitment, best effort and auction. Firm commitment is the safest and the most common type of IPO where the underwriters back the IPOs by effectively buying all shares at a slightly lower price than offer price and then reselling them. Conversely, in best-effort IPOs, the underwriters do not guarantee but just promise to execute the IPOs for the highest benefit to the issuers. In this method, the investment bankers act more like brokers promoting the deals; hence firms might still face the risk that IPO volume would not be completely filled. The last way to conduct an IPO is through an auction, where the underwriter announces the IPO and the price is determined by market’s demand.

There are several reasons that motivate companies to go public through IPOs. By far the most obvious motivation is that firms need additional funding to do their businesses (Ritter and Welch, 2002). Public equity funding may be more costly compared to private funding or debt, but once companies reach certain stages in their life cycle, it is more optimal to do an IPO in which they can raise a higher level of capital from a wide range of investors. Ritter and Welch (2002) also mention another reason to do an IPO, that is, to create a public market for existing shareholders to sell their shares as a way to diversify or exit in the future.

Other benefits of IPOs are also substantial. By going public, companies increase their liquidity and lower their cost of funding. In order to be publicly listed, firms must satisfy all the requirements of the stock exchange, which signals a remarkable credibility to investors and lenders. Also, firms’ management benefits when IPOs determine market prices for their stock-based incentives (Draho, 2004).

On the other hand, initial public offerings are costly in many aspects. The direct cost of going public – underwriting fee, is about 7 percent of IPO proceeds (Chen and Ritter, 2000). But the total costs even go beyond that. When firms issue stock publicly, existing shareholders’

(8)

7

ownership is diluted. Additionally, firms must disclose private information and perform under strict supervision of the public, and thus lose their competitiveness to some extent. Therefore, Draho (2004) states that in many circumstances, especially for small firms, a public offering may be an expensive option.

2.1.2. Long run performance of initial public offering

The performance of IPOs has been an interesting topic for many studies. Examining US IPOs from 1975 to 1984, a study of Ritter (1991) concludes that on average, initial public offerings significantly underperform their peers (34.47% versus 61.86% average holding return) and young firms mostly underperform the market indices (NASDAQ and Amex-NYSE) over three years after their IPOs. Ritter refers to investors’ sentiment as a possible explanation for such underperformance: investors might be systematically over-optimistic about IPOs and firms also prefer going public when optimism is near its peak. In line with that study, Loughran and Ritter (1995) show that the five-year average annual return of IPOs in the period 1970-1990 is only 5% while comparable non-issuing companies have a return of 12% annually. Brav, Geczy and Gompers (2000) also conclude that 1975-1992 IPOs underperform by approximately 44% compared to S&P 500 over five years post IPO.

What are the possible reasons for such underperformance of initial public offerings? Ritter (1998) provides three hypotheses to explain this phenomenon. Firstly, IPOs underperformance may result from firms timing their issuance based on market optimism. Firms usually do their IPOs when investors’ sentiment is the most positive and firms’ performance is near its peak, which explains the later poor performance. Secondly, underperformance may be explained by the fact that buyers, who determine the market price, are often the most optimistic ones about IPO prospect. Therefore, when information asymmetry diminishes after IPOs, firm performance generally cannot be as high as the expectation of these optimistic investors, creating winner’s curse problem for them. A third explanation comes from IPO underwriters. These investment bankers may intentionally underprice IPO to create excess demand with the purpose to fill IPO volume and achieve high initial return. As a consequence, the long run return is lower, and companies with the highest first-day returns will generally have the lowest long-term returns.

There are also other explanations for IPO underperformance. Brav and Gompers (1997) state that this phenomenon may be caused by the problem of long-run performance test, i.e. its assumptions and the non-standard distribution of IPO returns. Alternatively, Jensen (1986) presents the agency cost hypothesis, where managers with excess cash flow from IPOs may

(9)

8

invest in negative, or not optimal, projects at the expense of shareholders, resulting in subsequent lower returns. Another reason is the “lottery effect” (Kahneman and Tversky, 1979), where investors may accept to buy a share with below-average return to have a small chance of gaining extraordinary abnormal return.

However, other studies doubt that underperformance is not an exclusive effect of IPO but also a result of the underlying factors behind equity issuance. Investigating the impact of size and market-to-book value, Brav and Gompers (1997) and later Brav, Geczy and Gompers (2000) conclude that IPOs perform no worse than non-issuing firms with comparable size and book-to-market ratios. Furthermore, underperformance is also associated with seasoned equity offerings (“SEO”), when firms decide to issue additional equity after their IPOs. According to Loughran and Ritter (1995), the average compounded annual return of SEOs is just 7% versus 15% average return of similar firm in a five-year period. Berk and DeMarzo (2013) try to explain equity issuance’s underperformance (both IPOs and SEOs) by the change in firm’s riskiness. Firms issue equity when they need funding for investment, which is equivalent to exercising their growth options. Since the growth options entail more risk than the investment projects themselves do, by exercising (and eliminating) these options, firms actually exchange the riskier ones (the options) with the safer ones (the projects). Hence, firms lower their betas and riskiness, which leads to lower subsequent returns.

2.2. Venture capital

2.2.1. Ways of equity financing for start-up firms

For start-up companies it is generally difficult and costly to raise capital because from the perspective of investors, there is a large uncertainty as well as a high chance of failure incorporated in newly-created firms. However, there are still several sources for these firms to raise external equity capital: angel investors, venture capital, private equity, institutional investors and corporate investors (Berk and DeMarzo, 2013).

Angel investors refer to individuals that invest very early in the companies’ life. Wetzel (1987) defines that business angels are individual investors with certain amount of asset and income who can take substantial risks over long horizon to have future large returns. At this stage the investment is very risky, so angels are typically friends or acquaintances of the entrepreneurs. And since the amount invested is sizable compared to the total capital of the startups, angel investors usually have considerable impact in managing and controlling the firms.

(10)

9

When firms grow bigger, they need to seek larger funding, potentially from venture capital or private equity funds. Private equity is a form of investment whose targets are privately-held companies or public companies which would then be turned into private ones. Similar to angel investing, PE funds often bring expertise and actively take part in the management of the start-ups. Venture capital investment can be considered an independent form or a subgroup of private equity investment, and hence is separately examined in the next part of this section.

Other sources of financing are institutional investors or corporate investors. Institutional investors, such as pension funds, endowments or insurance companies, may invest directly in start-up companies or indirectly through private equity or fund-of-funds. Corporate investors, normally operating in the same industry as the entrepreneurship, invest not only with financial incentives but also with ‘strategic’ purposes, for example to acquire the expertise or innovative ideas of the entrepreneur.

2.2.2. Venture capital industry

Venture capital is a form of investment into young, entrepreneur-led start-up firms, generally driven by technology innovation (Eckermann, 2006). As previously mentioned, venture capital can be considered an independent type or a form of private equity investment. A VC fund is usually structured as limited partnership with a general partner who manages and is liable for the fund and several limited partners who commit their money to the fund.

The history of venture capital industry goes back to mid-20th century when the first venture

capital firm - American Research and Development - was established in 1946 (Gompers and Lerner, 2001). However, it is not until 1958 that the first VC limited partnership – the most popular form of VC funds nowadays, was founded. The venture capital industry began to expand substantially in the late 1970s and early 1980s with a large amount of VC investment focusing on high technology companies. Currently venture capital is one of the fastest-growing industry, especially in the United States where several giant corporations were initially backed by VC such as Microsoft, Apple Inc. and Google.

Regarding venture capital investment, Gompers and Lerner (2001) divide the VC investment process into three stages: fund raising, investing and exiting.

Firstly, venture capital funds need to raise capital from their investors, or limited partners. However, the capital committed fluctuates over the years (Gompers and Lerner, 2001), which can be explained by the impact of capital gain tax variation. Also, it is affected by the economic conditions (Black and Gilson, 1998) since in a bull market, more start-up companies are

(11)

10

founded, which leads to the increase in the level of VC fund raised to capture the existing opportunities. Other macroeconomic variables also have a role in determining the activity of venture capital fundraising, including IPOs activity, labor market rigidity and market capitalization (Jeng and Wells, 2000).

The investing process involves choosing the right entrepreneurs to invest in and monitoring them over their growth phases. Start-up firms are highly risky due to their uncertainty and agency problem, but these risks can be reduced with VC investment (Gompers and Lerner, 2001). It is because venture capital funds not only invest but also actively monitor their portfolio companies and provide less-experienced entrepreneurs with expertise. The decreased risk level might also result from the screening process and diversification with which VC invest in a large number of risky firms (Baum and Silverman, 2004).

The last part of VC investment is cashing out, i.e. exiting from portfolio companies, which can take the form of an IPO and trade sale (if successful) or liquidation (if unsuccessful). According to Gompers and Lerner (2001), an initial public offering is the most profitable way to exit from a venture investment. However, as VC funds cannot exit immediately through IPOs because of the lock-up period, in many cases a trade sale, in which portfolio companies are purchased by corporate firms in the same industry, is preferred. Giot and Schwienbacher (2007) show that when the start-up firms reach a certain milestone (around 2.75 to 4 years since the founded date), their chance of exiting through IPO declines more quickly compared to the probability of a trade sale exit.

2.2.3. Advantages and disadvantages of venture capital

The benefits and costs of venture capital investment can be analyzed in two aspects: from the perspective of the venture capitalist and from the perspective of the funded company.

Regarding VC firms, it is observed that venture capital investment historically has higher return than traditional types of investment. Cochrane (2005) shows that in the period 1987-2000, the arithmetic average return to IPO or acquisition backed by venture capital is 698% without correcting for selection bias and still 57% after correction. Korteweg & Sorensen (2010) extend Cochrane’s model and report similar results regarding the abnormal return and risk due to substantial sample selection bias of venture capital investment. These characteristics can be explained by the illiquidity of private equity, the lack of diversification and the involvement of venture capitalists in firm’s management (Cochrane, 2005). Besides higher return, another advantage of VC investment is that when directly involving in the management process, VC

(12)

11

firms can actively monitor their investment and prevent the entrepreneurs from not using the resources efficiently for their own benefits.

In contrast, high risk of failure, illiquidity and lack of diversification are the main drawbacks of VC investment. VC return distribution is characterized with very high volatility, highly positive skewness and a few extreme positive observations (Cochrane, 2005), indicating the small probability of a successful VC investment. VC firms invest in a start-up company from the beginning of its life and remain their investment for several years before exiting. Thus, their investment is very illiquid until the portfolio company reaches certain stages of development and gains attention from the market. This illiquidity is also associated with the lack of diversification. A venture capitalist can diversify its portfolio by funding several start-up companies; however mostly it only focuses on a few areas that it has expert knowledge in. With regard to the backed companies, on the one hand, venture capital funding provides some advantages which are already mentioned in the previous parts, namely expertise and monitoring of VC firm and better access to the market, which are crucial in the early stages of the companies (Gompers and Lerner, 2001; Baum and Silverman, 2004). In addition, other benefits of VC are also backed by empirical results. Megginson and Weiss (1991) find that IPOs supported by venture capitalists have lower degree of underpricing and underwriter compensation, attract more reputable underwriters and auditors, and are able to happen at a younger age compared to non-VC backed firms. Sahlman (1990) concludes that venture capital solves the corporate governance problem (i.e. the problem that a firm’s manager may not act in the owners’ best interests) in start-up companies with extensive due diligence performed during the monitoring process.

On the other hand, newly-created companies funded by venture capital have to face some disadvantages. Since venture capitalists involve directly in management, conflicts of interest with entrepreneurs may appear (Ritter, 1998). The entrepreneurs lose some of their control over the firms and are partly dependent on the VCs in decision making process. Additionally, the VCs may have preferred benefits at the expense of other shareholders who invest in the companies later, such as the higher return from their preferred stocks.

2.2.4. Long-run performance of venture-capital-backed IPOs.

Private equity shares some common characteristics with venture capital and to some extent, venture capital is classified as one form of private equity investment. A study from Bergström, Nilsson and Wahlberg (2006) focuses on long run performance of private-equity-backed IPOs and points out that PE-backed IPOs outperform non-PE-backed IPOs across all time horizons.

(13)

12

Along with PE expertise, the reduced over-optimism as well as the less divergence of opinion due to more available information about firms backed by private equity are also offered as possible explanations for such outperformance.

A key article in this field from Brav and Gompers (1997) focuses on examining the influence of venture capitalists on the performance of their backed IPOs in the period 1972-1992. In this research, long run performance is calculated as five-year buy-and-hold stock return (equally weighted and value weighted) and then is compared to several benchmarks such as S&P 500, NASDAQ, NYSE/AMEX indices or comparable Fama–French industry portfolios. The authors conclude that although after going public firms achieve lower returns than market’s average, there are differences between venture-backed and non-venture-backed companies. The results show that returns on VC-backed IPOs are remarkably higher than those of non-VC-backed IPOs in the sample. One interesting point discovered in this paper is that differences in group’s performance are reduced when returns are value-weighted. Similar to Bergström, Nilsson and Wahlberg (2006), Brav and Gompers argue that the possible mentioned outperformance might be caused by the reputation and expertise of venture capitalists as well as the weaker effect of investors’ sentiment (which is measured by closed-end funds’ discounts) due to greater availability and certainty of information.

Focusing on European market, Bessler and Seim (2011) provide evidence for the better performance of VC-backed IPOs over both the market and non-VC-backed IPOs. However, the result also shows that positive abnormal returns only exist in two years post-IPO before turning to negative subsequently.

Based on the three mentioned studies, it seems that the advantages of venture capital investment (VC’s expertise and active monitoring, better access to credit, auditors and underwriters, etc.) outweigh the ordinary drawbacks of going public (information asymmetry, “peak” timing of IPO or managers’ suboptimal use of excess cash flow from IPO process) mentioned by Ritter (1998), resulting in the better performance of VC-backed IPOs.

2.2.5. Venture capital reputation

A problem with the above papers is that they treat venture capitalists homogeneously. Shapiro (1982) states that when information asymmetry exists, reputation of a firm is its valuable asset that signals to customers the quality of products or services it supplies. For a fragmented industry like venture capital, the importance of reputation is even more crucial (Krishnan and Masulis, 2011). From the perspective of venture capitalists, their offers to entrepreneurs are

(14)

13

more likely to be accepted and also accompanied by a larger discount with respect to more reputable VCs (Hsu, 2004).

Some studies aim to capture the difference in venture capitalists’ reputation and its impact on the portfolio company. Nahata (2008) and later Hamza and Kooli (2011) study the effect of VC reputation on survival chance or the likelihood of successful exit of IPOs. They come to the same conclusion being that firms backed by more prestigious venture capitalists tend to have more chance of survival and successful exit in the long term. Baker & Gompers (2003) and Krishnan et al. (2011) also find that firms funded by more prestigious venture capitalists typically exhibit higher degree of corporate governance, measured by the greater number of independent directors, and by the higher probability that the VCs retain their directorships and shareholdings in the funded companies three years after the IPO.

Krishnan and Masulis (2011) examine the effect of VC reputation on firms’ post-IPO performance in more detail. Reputable venture capitalists bring substantial benefits to portfolio companies, including implicit VC guarantees, expert advisory services and better access to credit or underwriters. Krishnan and Masulis use several ways to measure VC reputation such as three-year past market share of completed VC-backed IPOs, VC age or total capital managed by VC. The main finding of the paper implies that more reputable venture capitalists are associated with higher long-run performance of IPOs backed by them. The result is still correct after controlling for omitted variables bias and sample selection bias. In another study, Krishnan et al. (2011) use the same method and get similar results regarding the positive correlation of VC reputation and IPOs’ long-term performance.

To conclude, reputation of venture capital can bring many advantages to both VC firms and portfolio companies. According to Krishnan and Masulis (2011), for a reputable VC firm, the benefits include higher negotiation power with start-up companies, easier fundraising, higher fee structure and higher performance of its investments. For the portfolio company, being funded by a highly ranked VC firm brings along greater credibility, expertise and an implicit guarantee from the venture capitalist.

2.3. Geographic characteristic and Asian market

There are many evidences supporting the impact of geographic location on investment decision (Coval and Moskowitz, 1999; Huberman, 2001) and more specifically, in venture capital investment decision. A study from Cumming and Dai (2010) clarifies that venture capital firms show strong local bias – they prefer to make investments within their local areas or in their country rather than cross-border investments. The underlying reason is that VC firms have

(15)

14

more knowledge and expertise related to the area (or broader, the country) they are operating in and consequently, have less information asymmetry problem if they invest locally. The appearance of geographic bias makes it meaningful to examine the performance of VC investment in one particular area (i.e. Asia in this thesis) and compare to the corresponding performance in other areas.

With respect to the Asian region, not only there are currently just a few researches about Asian IPOs performance or venture capital industry but the results thereof are also inconsistent. Regarding Singapore and Hong Kong, Dawson (1987) finds that there is no significant difference between performance of IPOs and their comparable firms. For Japanese market, Cai and Wei (1997) point out the significant underperformance of IPO firms, which is in line with the findings of Ritter (1991) and Loughran and Ritter (1995).

Nevertheless, Kim et al. (1995) conclude that Korean IPOs outperform seasoned firms with similar characteristics in the period 1985-1989. Jelic, Saadouni and Briston (2001) report positive abnormal long-run performance of IPOs in Malaysia up to three years after going public. These findings are in contradiction to what research on US IPOs (Ritter, 1991; Loughran and Ritter, 1995) concludes.

In the field of venture capital, a study from Hamao et al. (2000) illustrates that Japanese VC-backed IPOs perform no better than other IPOs in the long run. Moreover, it finds that firms backed by securities companies’ subsidiaries perform considerably worse than other IPOs. Lee (2011) also concludes that VC participation has a negative impact on the three-year stock performance of Korean IPOs. Other studies of Wong (2004) and Xu (2006) similarly report the underperformance of venture-backed firms in Hong Kong and China IPO markets respectively. The diverging conclusions as well as differences compared to research conducted in the US or Europe may result from Asian business environment. According to Naqi and Hettihewa (2007), it is the economic and political diversity in Asian countries that leads to significantly different environments in which the firms operate. For example, Japanese IPOs are generally older, larger, are limited by more stringent listing requirements and probably more influenced by cultural environment (Matsuda, Vanderwerf and Scarbrough, 1994). Ownership structure is another difference, where firms in Asian countries are more likely to be family-controlled or state-controlled with poorer shareholder protection than firms in developed markets (La Porta, Lopez-De-Silanes and Shleifer, 1999).

(16)

15

Or regarding venture capital specifically, in Japan many VC firms are subsidiaries of securities firms or banks, which can lead to the conflict of interest when VC-backed IPOs are underwritten (Hamao et al., 2000). Even more, in Japan VC firms have a limited role in the management process of the start-up firms, hence the benefits of VC expertise may not be fully realized. Lee (2011) tries to explain VC-funded IPOs’ underperformance by the high initial returns of IPOs in hot market, causing poorer returns after that. Black and Gilson (1998) states that capital markets in Asian countries are generally bank-centered, while US capital market is stock-market centered. Also, in Asia the market is less transparent and regulatory is less protective to investors (Backman, 1995). These factors may lead to the difference of venture capital industry in Asia and the US.

From these dissimilarities, it is expected that the performance of Asian VC-backed IPOs would not be the same as concluded in existing literature focusing on US and European markets.

3. Hypothesis and methodology 3.1. Hypothesis

The objectives of this thesis are (i) to examine how the Asian venture-backed IPOs perform relatively to their counterparts that are either (a) in the same region but not VC backed or (b) in other geographical regions; and (ii) to test whether the findings from such examination can be enhanced by the reputation of venture capitalists.

Regarding the performance comparison, Brav and Gompers (1997) show better results from VC-backed IPOs compared to non-VC-backed IPOs, which implies venture-capital value-added in portfolio companies. In term of difference due to location, there is little research about Asian region and the conclusions are contradictable. Hamao et al. (2000), Wong (2004) and Xu (2006) present the long-run underperformance of VC-backed IPOs in different Asian markets, namely Japan, Hong Kong and China.

One suggestion for this mixture is the complicated effect of Asian business environment. On the one hand, VC industry is considered less developed in Asia (except for Japan and possibly South Korea) than in the US or Europe and thus there are areas for venture capitalists to exploit and get higher returns. Actually in recent years, Asia has achieved substantially higher growth than the world’s average and generated a brighter outlook from the perspective of investors (Preqin, 2014). On the other hand, as reported in the mentioned papers about Asian IPOs (Hamao et al. 2000; Wong 2004; Xu 2006), the immature, risky and less transparent IPO market

(17)

16

in Asia may probably lead to the underperformance of Asian VC-backed IPOs compared to their counterparts. The current research agrees with the latter effect and consequently proposes the following hypothesis:

Hypothesis 1: Asian VC-backed IPOs underperform (a) Asian non-VC-backed IPOs and (b)

other-location VC-backed IPOs.

The second objective of the thesis is to examine the effect of venture capital reputation on the post-IPO performance of venture-backed firms in Asia. Until recently only a few papers (Nahata 2008; Hamza and Kooli 2011; Krishnan and Masulis 2011) study that relationship (but only in US market); and this thesis examines it in more detail with the focus on Asian region. Reputation of the venture capitalists can bring considerable benefits to both the VCs and their backed firms (Krishnan and Masulis, 2011). With respect to the backed companies, these benefits include better expertise provided by VCs during management; better access to underwriters, auditors and creditors; or a more widely network to suppliers and customers. The hypothesis is based on the conclusion of aforementioned papers, i.e. being funded by highly prestigious venture capitalists enhances the companies’ post IPO performance.

Hypothesis 2: More reputable venture capitalists improve the long run performance of Asian

VC-backed IPOs. 3.2. Methodology

To answer the two mentioned hypotheses, ordinary least square (OLS) regression method is used to measure the influence that location (either Asia or others) and VC reputation have on post-IPO performance of venture-backed companies.

3.2.1. Main equations of regression analysis

To analyze the first hypothesis, OLS regression with dummy variables and interaction terms is used:

𝐵𝐻𝐴𝑅_𝑖 = 𝛽₀+ 𝛽₁𝑉𝐶_𝑖 + 𝛽₂𝐿𝑜𝑐𝑎𝑡𝑖𝑜𝑛_𝑖+ 𝛽₃𝑉𝐶_𝑖 × 𝐿𝑜𝑐𝑎𝑡𝑖𝑜𝑛_𝑖+ 𝐶𝑜𝑛𝑡𝑟𝑜𝑙_𝑖+ 𝑢_𝑖

VCi = 1 if the company i is venture-capital backed and 0 if otherwise. Locationi = 1 if the

company i is located (headquarters) in Asia and 0 if otherwise. To test whether Asian VC-backed IPOs underperform Asian non-VC-VC-backed IPOs, the model checks if β1+β3 is significantly smaller than zero. Similarly, if β2+β3 is significantly negative, it can be concluded that Asian VC-backed IPOs underperform VC-backed IPOs in other parts of the world. Finally,

(18)

17

Asian venture-backed performance can be compared to the benchmarks by examining the sum of β0, β1, β2 and β3.

For the second hypothesis, data sample is reduced to include only Asian VC-backed IPOs because the thesis aims to examine to the benefits of venture capital reputation with respect to Asian market. β1 is the coefficient of interest and its significance is tested. The regression equation is as follows:

𝐵𝐻𝐴𝑅_𝑖 = 𝛽₀ + 𝛽₁𝑉𝐶𝑟𝑒𝑝𝑢𝑡𝑎𝑡𝑖𝑜𝑛_𝑖 + 𝐶𝑜𝑛𝑡𝑟𝑜𝑙_𝑖 + 𝑢_𝑖

Both regressions use a set of control variables, which is introduced in the later part of this section. Firm, country and year fixed effects are included. Standard errors are clustered by year, industry and country since firms in the same industry or same country and go public in the same year tend to expose to similar risks and have correlation in their returns.

Data may face sample selection bias, that is, good firms are more likely to be discovered and funded by more reputable VCs. The observed relation between VC reputation and post-IPO performance can be the result of the portfolio company’s potential itself, or of more reputable VC having access to more promising business opportunities, or of both (Hsu 2004; Krishnan and Masulis 2011). For that reason, potential bias is controlled using Heckman two-step selection procedure, which is incorporated in STATA program. This procedure needs instrumental variables that are correlated to the independent variable (VC reputation) but exhibit no significant relationship with the dependent variable (buy-and-hold abnormal return). Based on Krishnan et al. (2011), two variables are chosen for that purpose: (i) the number of VCs funding the firm; and (ii) a dummy variable to show whether the headquarters of any VCs are in the same country with the funded companies’ headquarters. These variables are suitable because, as Krishnan et al. (2011) explain, (i) a reputable VC funding the entrepreneur can attract other VC for that firm, and (ii) VC firms tend to prefer nearby entrepreneur as they have more knowledge in their local areas and more reputable VCs have better chance to choose the funded firms near their locations. It is plausible that these variables do not affect post IPO stock return because VC ownership is likely to diminish after the offering and only the lead VC firm still has some controls over the entrepreneur.

3.2.2. Long-run performance measure

To measure the post-IPO performance of VC-backed firms, this thesis uses three-year buy-and-hold abnormal return (BHAR) of stocks. This indicator is widely used in extant research, such

(19)

18

as Ritter (1991), Loughran and Ritter (1995) or Brav and Gompers (1997). The period of three years is chosen so that the momentum effect of IPO on stock prices is minimized.

Buy-and-hold abnormal return is calculated as follows:

𝐵𝐻𝐴𝑅_𝑖 = ∏(1 + 𝑅_𝑖𝑡) − ∏(1 + 𝑅_𝑏𝑡)

𝑇

𝑡=1 𝑇

𝑡=1

That is, buy-and-hold abnormal return of stock i is the difference of its three-year cumulative return and the cumulative return of a benchmark during the same period. BHARs are calculated starting from the IPO first day close. Since global IPO data is assessed for the first hypothesis, the benchmark chosen here is MSCI World index, which consists of 1635 developed-country stocks and is widely used as a benchmark for global equity funds. For robustness check, along with MSCI World, two other indices (S&P 500 and Nikkei 225) are used so that IPO returns are also compared to a US index and an Asian-country (Japanese) index.

Buy-and-hold abnormal returns are then winsorized at 0.5% and 99.5% levels (Krishnan et al., 2011 use 1% and 99% levels) to limit the effect of outliers. The reason is that with venture capital investment, the return distribution is heavily skewed and the existence of extreme outliers may have large impact on the sample average.

In addition to buy-and-hold abnormal return, survival rate (i.e. the probability that a firm remains listed three years after IPO) is used to capture long-run performance (Hamza and Kooli, 2011) in robustness test. The hypotheses can be strengthen if firms’ chance of survival is affected in the same direction as buy-and-hold abnormal return is.

3.2.3. Venture capital reputation measure

The main variable of interest for the second hypothesis is venture capital reputation, which is measured by the market value of all IPOs backed by a VC as a proportion of total size of all VC-backed IPOs in the period of three years preceding IPO date (Krishnan et al., 2011). For example, for a venture-backed IPO that took place in 2011, the total value of all IPO backed by that VC is aggregated in three years 2008, 2009 and 2010. Then this figure is divided by the total value of all VC-backed IPOs during the same three years to result VC reputation variable. As Krishnan et al. (2011) explain, calculating VC reputation in this way helps capture all the information prior to IPO date, as a result, look-ahead bias is eliminated.

One potential problem is that a firm may be funded by not just one but a number of venture capital firms. In this case, each VC firm is given full amount of credit, which is in line with

(20)

19

Krishnan and Masulis (2011), since all VCs actively involve in the management of the portfolio company. Furthermore, an entrepreneurship is normally funded in several rounds, and it is difficult to isolate the impact of each VC funding in different time periods.

For the robustness check, the thesis use an alternative measure of VC reputation used in Krishnan et al. (2011), that is the age of VC firm, i.e. the number of years from VC’s founded year to IPO date. However, this measure has some disadvantages, because long time existence does not guarantee its performance or the perception from the market. Due to its limit, VC age is only used as a supplement for the main calculation of VC reputation mentioned above.

3.2.4. Explanatory control variables

In financial market there is substantial heterogeneity between companies. For that reason, this analysis uses a set of variables to control for the effects on firm’s performance that are not related to venture capital.

Underpricing

Underpricing is the change in stock’s first day close price relative to its IPO offer price. IPO underwriters may intentionally create a positive first day returns to attract investors and also for their own benefits. As suggested by Ritter (1998), higher degree of underpricing may be associated with lower long-run return of the stock.

Size and market-to-book ratio

Brav and Gompers (1997) discover the relationship between a firm’s size and book-to-market ratio and its long-run performance. Their result shows that post-IPO return of small and low book-to-market companies is not different from return of comparable non-issuing companies in the market, and that return varies when the firm’s size and book-to-market ratio change. Based on that study, the thesis includes the natural logarithms of firm size and market-to-book ratio at the time of IPO issue as control variables.

Offer price

IPO offer price is constructed by underwriters based on the demand of the market, or put differently, it is the unofficial bid of potential investors during book-building process. Therefore, it can be expected that the higher the offer price, the better outlook of the company from the market expectation; and hence the stronger return of the stock post IPO. However, this also depends on the initial size of the firm and the number of shares it wants to issue publicly.

(21)

20 The number of underwriters

Carter, Dark and Singh (1998) show that IPOs underwritten by more reputable investment bankers typically exhibit a lower degree of negative long-term returns than IPOs with less reputable ones. For that reason, this thesis uses the number of underwriters as a measure of the influence from the underwriters and includes it as a control variable in the analysis.

Other control variables

The regression analysis also includes fixed effect controls for the following characteristics: year of IPO, industry and country in which IPO firm operates. These variables help reduce the discrepancy in firms’ businesses. Regarding IPO year variable, Ritter (1991) illustrates that companies going public in heavy-volume years (such as the years just before financial crisis) exhibit the most significant underperformance. Additionally, the industry and country variables are used to account for the specific market condition of each firm, especially in this case where Asian countries have diverse characteristics in culture, regulatory or business practices.

3.2.5. Robustness checks

This study also performs some robustness tests to check for the consistency of the regression results. Along with MSCI World, two other indices (S&P 500 and Nikkei 225) are used for benchmarking so that the IPO returns are also compared to a US index and an Asian-country (Japanese) index. In addition to buy-and-hold abnormal return, survival rate (i.e. the probability that a firm remains listed on the exchange three years after its IPO) is used to measure long-run performance (Hamza and Kooli, 2011). With this dependent variable, a probit model is employed to measure the marginal effect on survival probability. For VC reputation, the venture capitalist’s age up to IPO date is employed as another measure. In addition, this study includes three analyses on respectively Japanese, Chinese and South Korean IPOs (where MSCI country indices are uses as benchmarks) to answer the research question within these three biggest economies in Asia.

4. Data and descriptive statistics 4.1. Data sample

To test the two stated hypotheses, this study covers a data sample of global IPOs that took place during the eleven-year-period between 2001 and 2011. The year 2001 is chosen to be the beginning year to minimize the effect of Asian financial crisis in the late 1990s. Moreover, the

(22)

21

data sample starting from 2001 may alleviate the effect of dot-com bubble that happened in 2000. And because the analysis needs three years to realize buy-and-hold abnormal returns, a sample period ending in 2011 is considered suitable for the observation. Furthermore, the analysis uses three-year IPO market share as an indicator for venture capital reputation; hence, information about IPO deals completed in 1998-2000 is also obtained.

The first step of the data collection is to extract IPO deals’ information from Thomson One database – an extensive database about IPO and financial transaction deals. Initially, there are a total of 29,783 completed deals found within the sample period 1998 – 2011. After removing duplicate data and IPOs which are not clearly specified as VC-backed or not, 22,662 deals remain in the data sample.

The next step requires matching companies’ information provided by Thomson One and Datastream. There are three popular identifiers that can be found with IPO data in Thomson One, namely ISIN, SEDOL and 9-digit CUSIP. Firms without any of these identifiers cannot be matched and thus are removed from the sample. After that, three mentioned identifiers are converted to Datastream code (DS code) for further uses.

In the third step, data about stock returns is collected from Datastream database. This study uses Total Return index as the indicator for stock return because this variable is already adjusted for dividends and changes in capital structure such as stock splits. Along with stock return, companies’ information missing in Thomson One is also collected from Datastream. For many variables such as first day closing price, unadjusted figures are extracted to be consistent with original data from Thomson One. At the end of the data collection progress, a sample of 15,341 IPOs (of which 11,597 IPOs occurred between 2001 and 2011) is drawn. Specifically for the second hypothesis, data sample is reduced to include only VC-backed IPOs in order to measure the influence of venture capital reputation. Furthermore, data about venture capitalists (VC firms) is required for this purpose. This kind of information can be collected from Thomson One database; however, there are a lot of VC-funded IPOs that do not have information about VC firms. Consequently, the sample for the second hypothesis is further narrowed, resulting in 267 Asian VC-backed IPOs and 517 non-Asian VC-backed IPOs.

4.2. Descriptive statistics

4.2.1. Long-run performance of Asian VC-backed IPOs

In the study period of 2001 – 2011, there are a total of 1,045 VC-backed IPOs from Asian firms, accounting for 20.22% of Asian IPOs (5,167 deals), 58.71% of VC-funded IPOs (1,780

(23)

22

deals) and 9.01% of global IPOs (11,597 deals) respectively. Table I describes the number of VC-backed IPOs and its proportion each year of the sample period. It can be seen that the years which experience the highest number of deals are 2005, 2006, 2007, 2010 and 2011. These five years are associated with about 78% of the total number of Asian VC-backed IPOs in the eleven-year period. While the large number of deals in 2005-2007 illustrates the expanding market with a very high growth rate, the 2010- 2011 figures probably show the recovery of venture capital and IPO market, at least partly, after the financial crisis.

Table II describes general statistics and correlation of different indicators of IPOs’ long run stock performance, namely three-year buy-and-hold return (BHR), three-year buy-and-hold abnormal return (BHARs) with MSCI World Index, S&P 500 or Nikkei 225 served as benchmarks, and the IPO survival (whether a firm remains listed after three years or not). There are three points worth noticing in this regard. Firstly, the first four variables have a very strong correlation of over 0.9, indicating that they can be replaced with each other. However their relation to the survival rate is loose with the correlations all below 0.03. Secondly, the data shows that during 2001 – 2011, Asian VC-backed IPOs exhibit both negative holding return and negative abnormal return in the period of three years post IPO. The average BHR is -13.61%; and after being standardized with benchmarks, the average return ranges from

Table I: Total number of Asian venture-backed IPOs (2001-2011)

Data sample consists of initial public offerings that take place worldwide between 2001 and 2011, which is collected from Thomson One database. The number of Asian VC-backed IPOs is computed each year and for the whole period and then is computed to the corresponding figures of larger groups, namely Asian IPOs, VC-backed IPOs and global IPOs.

Year Number of Asian

VC-backed IPOs

% of Asian IPOs

% of

VC-backed IPOs % of IPOs

2001 12 3.87% 22.22% 1.56% 2002 19 4.96% 33.93% 2.30% 2003 27 6.32% 43.55% 3.32% 2004 63 17.65% 36.84% 5.84% 2005 181 33.64% 62.63% 13.70% 2006 175 36.76% 56.45% 12.58% 2007 129 20.81% 50.59% 7.63% 2008 59 17.05% 77.63% 8.36% 2009 47 12.40% 74.60% 7.72% 2010 156 21.05% 73.93% 12.17% 2011 177 30.00% 75.97% 15.93% Total 1,045 20.22% 58.71% 9.01%

(24)

23

approximately -27% to -34%. In contrast, about 96% of Asian VC-backed IPOs remain listed after three years, which is higher than the average of general IPOs (89%). Thirdly, with the mean larger than the median, all return measures are positively skewed with maximum values of about 1,600%, even after they are winsorized to limit the effect of outliers. Therefore, it can be said that in general, a large percentage of VC-backed IPOs have negative returns, with the exception of very few firms exhibiting exceptional positive abnormal returns.

It is beneficial to zoom in on the average performances of IPOs within a same year, which are shown in table III. The main idea is that not only VC-backed IPOs but also other IPOs lag the market, since BHARs for all groups are negative in average and almost negative every year. This is in line with Ritter (1991), who finds an overall underperformance of IPOs. However, VC-backed IPOs seem to perform worse than the overall IPOs, contradicting the conclusions of Brav & Gompers (1997) and Bessler & Seim (2011). One potential reason is that the statistics here are of global IPOs while those two papers study US and European markets

Table II: Descriptive statistics of Buy-and-hold Return and Abnormal Return

Data sample consists of Asian venture-backed initial public offerings that take place between 2001 and 2011, which is collected from Thomson One database. Buy-and-hold return (BHR) and Buy-and-hold abnormal return (BHAR) are calculated over three-year period beginning from IPO date. BHR and BHAR are based on Total Return Index (RI) from Datastream database, which is adjusted for dividends and changes in capital structure. MSCI World Index, S&P 500 and Nikkei 225 are chosen as the benchmarks for abnormal return calculation. Returns are winsorized at 0.5% and 99.5% levels to reduce the effect of outliers.

Panel A: Correlation matrix of buy-and-hold return, buy-and-hold abnormal return and survival rate

BHR BHAR (MSCI) BHAR (S&P500) BHAR (Nikkei 225) Survival Rate BHR 1 BHAR (MSCI) 0.9667 1 BHAR (S&P500) 0.9681 0.9947 1 BHAR (Nikkei 225) 0.9364 0.989 0.9852 1 Survival Rate 0.0250 0.0118 0.0115 0.0048 1

Panel B: Summary statistics of buy-and-hold return and abnormal return

Mean Median Min Max Standard

Deviation BHR -13.61% -45.71% -99.86% 1611.19% 110.07% BHAR (MSCI) -31.91% -56.75% -176.72% 1560.86% 104.89% BHAR (S&P 500) -33.92% -57.16% -160.79% 1582.10% 106.06% BHAR (Nikkei 225) -27.58% -44.84% -194.98% 1547.53% 106.82% Survival Rate 0.9627 1 0 1 0.1896

(25)

24

respectively. Moreover, Asian VC-backed IPOs are the ones with the lowest returns over the sample period, especially when they are compared with Asian IPOs or IPOs in general. The data period can also be divided into two sub-periods: 2001-2004 when Asian VC-funded IPOs mainly outperform the market with positive BHARs and 2005-2011 when they severely underperform the market. This seems to be consistent with the financial crisis starting in 2007. Table III shows the average underperformance of Asian VC-backed IPOs in comparison with their counterparts; however, a more precise conclusion is drawn in the result section where other characteristics of the firms are taken into account.

It should also be noticed that firms in the same industry or from the same country are exposed to some similar risks and tend to have correlation in returns. Hence, table IV provides some statistics of Asian VC-backed IPOs divided into industries or nations.

There are 13 main groups of industry (according to Thomson One database) displayed in Panel A of Table IV, of which High Technology industry accounts for the highest number of IPOs (347 deals) – approximately 33.21% of total Asian VC-backed IPOs in the studied period.

Table III: Mean of IPOs’ returns and abnormal returns by year (2001-2011)

Data sample consists of initial public offerings that take place worldwide between 2001 and 2011, which is collected from Thomson One database. Buy-and-hold return (BHR) and Buy-and-hold abnormal return (BHAR) are calculated over three-year period beginning from IPO date. BHR and BHAR are based on Total Return Index (RI) from Datastream database, which is adjusted for dividends and changes in capital structure. MSCI World Index is chosen as the benchmark for abnormal return calculation. Returns are winsorized at 0.5% and 99.5% levels to reduce the effect of outliers. Returns of Asian VC-backed IPOs are computed each year and for the whole period and then are compared to the corresponding figures of larger groups, namely Asian IPOs, VC-backed IPOs and global IPOs.

Buy-and-Hold Return Buy-and-Hold Abnormal Return (MSCI)

Year Asian VC-backed IPOs Asian IPOs VC-backed IPOs Total IPOs Asian VC-backed IPOs Asian IPOs VC-backed IPOs Total IPOs 2001 81.46% 21.41% 16.03% 19.71% 74.74% 15.84% 10.71% 14.39% 2002 38.18% 9.26% 63.94% 34.44% -11.83% -34.48% 24.55% -8.30% 2003 83.29% 39.23% 54.20% 51.16% 15.59% -25.56% -8.47% -12.51% 2004 59.30% 63.63% 54.43% 72.67% 1.53% 6.75% -5.38% 14.45% 2005 -52.45% -6.33% -35.90% 2.28% -64.65% -20.21% -48.95% -13.02% 2006 -52.10% -29.99% -40.44% -30.25% -31.28% -9.19% -19.49% -9.60% 2007 -36.68% -16.94% -33.02% -22.85% -17.07% 2.57% -13.18% -3.05% 2008 -4.21% 23.36% 9.56% 10.59% -11.89% 19.58% 3.45% 7.31% 2009 -13.94% 5.36% 0.44% 10.14% -47.44% -27.96% -32.66% -23.75% 2010 -16.49% -11.48% -5.77% -5.44% -57.12% -51.90% -46.47% -45.80% 2011 27.70% 21.03% 25.95% 8.69% -18.77% -24.21% -19.62% -35.70% Total -13.61% 6.87% -4.85% 8.81% -31.91% -16.84% -22.03% -11.54%

(26)

25

Whereas, there are only 40 IPOs from financial firms (3.83%) but their total proceeds amount is $13,700 million – 14.16% of total Asian venture-backed IPOs value, indicating the average large size of financial IPOs. Most industries perform worse than the market, except for Government and Agencies group. Although having an abnormal return of 15.93%, it only consists of one IPO; therefore, its result is generally not representative. Consequently, it seems that the underperformance of VC-backed IPOs exists across the industries.

Panel B of Table IV presents data about VC-backed IPOs in 12 Asian countries. It is noticeable that Japan, China and South Korea have the highest numbers of IPOs – 406, 340 and 138 deals are completed in these countries respectively. Together they account for approximately 85% of total VC-funded IPOs (both in terms of the number of deals and the proceed amount) in Asia, indicating the highly focused venture capital industry in this continent. However, in these three nations, VC-backed IPOs underperform the market significantly (BHARs of -30% or lower), which contributes to the average negative performance in the whole Asia. Conversely, countries with smaller numbers of deals have better results. For example, Hong Kong VC-backed IPOs have an average BHAR of 77.86%, while for Indian firms, the figure is 51.34%. Venture-backed IPOs in Malaysia, Singapore and Thailand, three countries in the Southeast Asia region, also exhibit positive abnormal returns over the three-year post IPO period.

4.2.2. Venture capital reputation and VC-backed IPOs’ long run performance

As previously mentioned, due to the missing data about venture capitalists that back IPOs, the sample for the second hypothesis is considerably reduced. Therefore, statistics regarding abnormal return in the reduced sample are changed compared to those in the full sample. Table V describes the second equation’s main variables in the reduced sample. Regarding BHAR, although all statistics are different from those of the full sample, the main characteristics, which are the negative mean and median, positive skewness and highly positive outliers, remain unchanged. Furthermore, Asian VC-backed IPOs still seem to underperform non-Asian ones on average. In contrast, data of survival shows that Asian VC-backed IPOs have a higher chance of remaining listed after three years than non-Asian VC-backed IPOs. A description of two measures of VC reputation – IPO market share and VC Age, is also included in Table V. The correlation matrix shows the positive relationship between the two measures. With a correlation of just over 0.18, the relationship is quite loose and these variables cannot perfectly substitute each other. However, this is similar to the computed correlation in Krishnan and Masulis (2011), which is 0.14.

(27)

26

With regard to IPO market share, venture capitalists that back Asian IPOs have an average 3-year market share of 0.94%, which is significantly lower than its counterpart of non-Asian IPOs. Moreover, the median, which stands at 0.23%, is even much lower. This implies that VCs backing Asian IPOs tend to have less reputation and are less active than VCs backing non-Asian IPOs.

A similar conclusion can be drawn regarding the age of venture capitalists. The mean and median age of VCs that back Asian IPOs until IPO date are 15 and 12 years respectively; both numbers are smaller than the corresponding figures of VCs that fund non-Asian IPOs. It may seem surprising that the maximum age of a VC is 138 years while venture capital industry only appeared in the twentieth century; however, it is because some companies, such as Goldman Sachs & Co, have existed for a long time before entering the VC industry.

5. Empirical results

5.1. Long run performance of Asian venture-backed initial public offerings

To examine the long run performance of Asian VC-backed IPOs, their three-year returns are regressed on dummy variables indicating whether a firm is Asian-located and whether it is venture-backed. The returns are either equally-weighted or value-weighted based on the proceeds amount from IPO deals. A set of control variables includes underpricing level, the natural logarithm of company size at IPO date, market-to-book ratio at IPO date, IPO offer price, the number of underwriters as well as year, nation and industry fixed effects. The main regression equation is as follows:

𝐵𝐻𝐴𝑅_𝑖 = 𝛽₀+ 𝛽₁𝑉𝐶_𝑖 + 𝛽₂𝐿𝑜𝑐𝑎𝑡𝑖𝑜𝑛_𝑖+ 𝛽₃𝑉𝐶_𝑖 × 𝐿𝑜𝑐𝑎𝑡𝑖𝑜𝑛_𝑖+ 𝐶𝑜𝑛𝑡𝑟𝑜𝑙_𝑖+ 𝑢_𝑖

Table VI reports regressions where the dependent variable is either three-year buy-and-hold return or three-year buy-and-hold abnormal return. In the first two specifications, regressions use buy-and-hold return as the dependent variable while in the last four ones, buy-and-hold abnormal return is used.

The table shows contradicting results depending on whether control variables are included. In the columns (3) and (4), the coefficient of the Location dummy’s coefficient is not significant while the interaction term’s coefficient is significantly negative. However, the reverse is true regarding the other specifications, where interaction term’s coefficient is not statistically significant but Location dummy’s coefficient is. Similarly, including control variables changes

Long-run performance of Asian venture-capital backed initial public offerings and the impact of venture capital reputation

Long-run performance of Asian venture-capital backed initial

public offerings and the impact of venture capital reputation

Table of Contents