Financial Market Sustainability in a Dual-Track System: Venture Capital and Startups’ Speed of Passing

Abstract

The government’s intervention under the approval system seriously affects the healthy and sustainable development of the financial market. An IPO is an important way for a venture capitalist (VC) to gain income, which impacts the efficiency of resource allocation in the capital market. From the perspective of resource allocation efficiency, this paper compares the influence of venture capital on the IPO process of startup enterprises under registration and approval systems. The findings are as follows: (1) after the trial registration system, the speed of passing and listing of VC-owned startup enterprises can be significantly accelerated. (2) Venture capitalists can accelerate the startup enterprises’ speed of passing by sending directors to startup enterprises and improving the level of risk disclosure, which is only significant under the registration and issuance system. (3) Further research shows that VC-supported startups perform better after listing. (4) VCs can help startup enterprises to choose hot season listing, which has a good timing effect. The conclusion of this text study is still robust after using propensity score matching (PSM) and Heckman to eliminate endogeneity. The conclusion of this study provides a theoretical basis and empirical support for emerging market countries to promote market-oriented reform.

1. Introduction

China is an emerging market economy, and its stock market has experienced 30 years of rapid development and dazzling achievements. However, the examination and approval system or the approval system has anomalies such as administrative intervention, long audit queues, and high underpricing rates, which have been criticized by market investors [1,2]. In addition, under the approval system, a high break rate of new listings and IPO suspension events occasionally occur, seriously influencing the stability and sustainability of financial markets. Therefore, reform of the registration system has become an inevitable choice in the development of China’s stock market. The registration and issuance system takes information disclosure as its core [3], substitutes formal examination for substantive examination of reporting enterprises by regulatory authorities, returns a judgment on the merits and demerits of listed enterprises to market players, reduces the distortion of stock price caused by administrative intervention and possible rent-seeking behavior, improves audit efficiency, and opens up financing channels for enterprises to go public. At the same time, the registration and issuance system improves the delisting system, increases penalties for enterprises that make false disclosures, promotes a virtuous circle of the stock market, and prevents the phenomenon of “Bad Money Drives Good Money out of Circulation” [4]. With the change in China’s marketization, the audit mode and efficiency of the registration system have always been the focus of practical circles, and it is also a new topic that academic circles can dig deeply into.

A venture capitalist is a long-term participant in the capital market [5,6]. Through detailed prior investigation, high-growth and high-risk startup enterprises are selected as investment objects. While providing financial support to startup enterprises, venture capitalists actively participate in the daily operation and investment decision-making of startup enterprises, provide human and intellectual support at different growth stages of startup enterprises, promptly push mature enterprises to the market, and realize investment return and reputation building [7,8]. However, European and American scholars have found that some young VCs are eager to establish an industry reputation and will quickly push immature startups to the market using financial whitewashing and false disclosure to help startups cover up their inferior performance until they exit after listing. This radical and short-sighted (grandstanding) behavior affects the stable operation of the capital market [9,10,11,12,13]. However, on the contrary, Zhang and Liao [14], Wang et al. [15], and Song et al. [16] found that the IPO underpricing rate of enterprises held by VCs is lower. However, the relevant research samples in China are based on the approval system, and regulators conduct substantive audits on the IPO enterprises to judge the development potential of the enterprises after listing and, thus, have a certain degree of policy and industry tendencies; therefore, the regulators also have a certain endorsement role in the approval system audit. However, the registration system in Europe and America focuses on formal audit, focusing on improving the degree of information disclosure, simplifying the issuance procedures, and improving the issuance efficiency, and does not make substantive judgments on the quality of enterprises. Therefore, the question is, after the trial of China’s registration system, does the short-sighted and radical (grandstanding) behavior of venture capitalists in European and American studies also exist in China’s capital market, that is, does venture capital push high-quality enterprises to the market or immature enterprises to the market for reputation building?

This paper studies the efficiency of a venture capitalist in resource allocation under the changing degree of marketization, taking the listed companies of China’s National Equities Exchange and Quotations (NEEQ) as samples. The results are as follows: (1) after the trial registration system, the VC-owned startup enterprises have a higher probability of success and faster passing and listing, which verifies the confirmation function of a VC. (2) After the trial registration system, the VC sends directors to startup enterprises and by improving the level of risk disclosure helps enterprises increase their passing speed, which verifies the supervision effect of venture capital. (3) The proportion of VC-owned startups listed in the IPO hot season is higher, which verifies the timing effect of the VC.

The main contributions of this paper are as follows: Firstly, it expands the relevant research on the gradual reform of capital markets in emerging market countries. The institutional background of the existing literature is usually a single registration system or approval system [11,12,14]. There are a few scenarios where the registration system and the approval system go hand in hand in the same economy. This paper compares the differences, advantages, and disadvantages of different IPO issuance systems in the particular period of China’s gradual market-oriented reform, the coexistence of a registration system and an approval system, which enriches the diversity of IPO-related research. Secondly, the existing European and American literature shows that under the registration system, some younger VCs may push immature startups into the market to build their reputation, and there are “grandstanding” behaviors [9,11,12]. On the contrary, as per the existing literature in China, a VC can significantly reduce the first-day underpricing rate of startups under the approval system of substantive audit, and there is no grandstanding behavior. However, because of the differences in the market environment and institutional culture in different economies, the research conclusions at home and abroad do not support each other. This paper argues that, after the trial of the registration system in China, a VC accelerates the IPO process of startup enterprises and reduces the underpricing rate and stock price volatility of startup enterprises, meaning there is no grandstanding behavior. Thirdly, this paper broadens the research margin of the role of a VC in the IPO market. The existing research mainly focuses on the role of the VC in the listing performance of enterprises [17], such as the underpricing rate on the first day after listing, stock price volatility, excess return rate, corporate governance structure, performance change, and financial status [18,19,20,21,22,23], but there is little research on the role of the VC in the “game” between startups and regulators when applying for listing. From the perspective of resource allocation efficiency, this paper studies the influence of the VC on the IPO process of startups by comparing the registration system and the approval system and verifying the certification and supervision functions of the VC. Finally, this study pro-vides a theoretical basis and empirical support for emerging market countries to promote the reform of the issuance system and the sustainable development of the financial market. The existing research on the reform effect of issuance policy in emerging market countries and the real development of audience groups is still scarce and, therefore, can not provide a theoretical basis and empirical support for policy makers in emerging market countries. This paper takes China’s gradual market-oriented reform as the research background. It compares and analyzes the influence of venture capital on the IPO process of startup enterprises under the conditions of an approval system and a registration system. The research conclusion has reference significance for the market-oriented reform and sustainable development of emerging market countries [21].

2. Literature Review and Research Hypothesis

A venture capitalist usually provides financial support to unlisted growth-oriented startups with great development potential, supplemented by management participation investment, to exit through IPOs, M&A, etc., and realize a high return on investment [22]. As one of the main exit channels of a VC, an IPO has always been a concern for academic circles. Megginson and Weiss [7], who introduced venture capitalism into the IPO field in the early stages, pointed out that a startup enterprise supported by a VC performs better in the market after launching an IPO; the listing cost is lower, and the role of the VC agrees with the certification hypothesis. Barry et al. [24] found that the VC actively plays a supervisory role in the IPO process of startup enterprises, effectively reducing the degree of information asymmetry between IPO enterprises and market investors, thus reducing their initial underpricing rates. SØRensen [25] found that a high-reputation VC could improve a startup’s probability of being listed. Nahata [8] found that high-reputation venture-backed startups go public faster and are more productive after they go public. Foreign scholars’ research on VCs mainly focuses on the degree of financial manipulation of startup enterprises [26], the improvement of production and operation ability [27], the short-term and long-term market performance after listing [28,29], and the role of a VC’s reputation [30,31], exit methods, and exit benefits [5].

Compared with mature capital markets, such as Europe and America, China started its venture capitalism late as an emerging market country. However, with rapid economic development, the number and volume of VCs in China have multiplied, and academic research on Chinese VCs has become increasingly abundant. Wang and Yang [32] confirmed that intermediaries could alleviate the information asymmetry between startups and market investors under the examination and approval system. Chen et al. [33] found that startups supported by VCs and high-reputation VCs have lower underpricing rates and a higher exit probability of VCs after listing. Fu et al. [34] found that high-quality VCs under the approval system can speed up the successful listing of startups, which confirms the value-added effect of VCs. Zeng et al. [35] found that venture capitalism and its characteristic variables under the approval system can improve the probability and listing speed of startups. In addition, there is a dual sponsorship system under the approval system [36], for example, for the opening of high-speed rail [37] and the association of the Development and Examination Committee [38,39]. Media exposure has an impact on the listing audit of startups [40]. With the trial of the registration system in China’s capital market, research on the registration system has become increasingly abundant. However, most of the existing literature focuses on the inquiry system in the science and technology innovation board [41,42], and the Shell resource value [43] rarely compares the differences between venture capital and startup listing progress under the approval system and the registration system.

2.1. Authentication Effect of VC

As an active equity investor, a VC actively participates in the production and operation activities of an IPO enterprise after investment. By helping the startup enterprise to formulate reasonable business development strategies, improve production efficiency, broaden sales channels, standardize financial systems, and enhance innovation ability, the VC can improve the comprehensive competitiveness of the IPO enterprise, and through its network, the startup enterprise can be helped to hire high-quality underwriters and audit institutions for listing counseling [44]. First, as an “insider”, a VC can more easily obtain the internal information and actual operating conditions of the startup enterprise by sending directors; a high-reputation VC has a richer IPO exit experience and a better understanding of the audit standards and priorities of the Development and Audit Committee (Listing Committee), which can help the startup enterprise to clarify its own development experience in a more organized way in terms of prospectus writing and feedback, reduce the information acquisition cost of regulators, and alleviate the degree of information asymmetry between the startup enterprise and regulators [27]. Second, as an “outsider”, by receiving the effective information released by the VC, the regulator has a better understanding of the actual operating conditions, financial quality, and future development prospects of the startup, which reduces the difficulty of obtaining information and the accountability risks caused by the performance reversal or delisting of the startup after listing [8,25]. In addition, a high-reputation VC has richer experience and networks of contacts, which can help the IPO enterprise obtain better resources and have a better under-standing of the audit rules and the style of the Development and Audit Committee (Listing Committee). At the same time, reputation is a long-term accumulated intangible asset, and a high-reputation VC pays more attention to maintaining the existing good reputation, which is more likely to prevent a startup enterprise from doing financial manipulation and displaying other fraudulent behaviors to avoid reputational damage caused by any investigation of the startup enterprise in the future, which will affect the future fundraising and exit rate [11].

The current approval system has many problems, such as administrative intervention, industry tilt policy, an opaque audit process, a substantive review of startup enterprises, the narrow IPO channel, and frequent queuing phenomenon under the approval system make it difficult to give full play to the certification effect of the VC [44]. However, after the reform of the registration system, regulators only formally examine the application materials of startup enterprises and no longer use administrative planning and industry policies as the basis for judging whether enterprises are listed or not; as a result, regulators recognize the role of VCs, thus speeding up the audit speed of startup enterprises. To sum up, the following assumption is made:

Hypothesis 1 (H1).

After the trial of the registration system, the VC-owned startup enterprises have a higher probability of meeting and increasing the speed of passing and listing.

2.2. Supervision Effect of VC

A VC plays an active role in the development and governance of a startup enterprise. The introduction of a VC can improve the governance structure of the startup enterprise, adding value to it [45,46,47]. A VC provides funds and has a more significant impact on the governance structure of a startup [48]. The principal–agent theory holds that the utility functions of the principal and agent are different. The principal pursues the maximization of his or her wealth. In contrast, the agent pursues the maximization of his or her salary and allowance income, luxury consumption, and leisure time, which will inevitably lead to a conflict of interests between them [49]. There will also be a principal–agent relationship between VCs and startup enterprises. The VC invests in a startup enterprise as the owner of capital, and the startup enterprise operates the entrusted property and creates profit as the agent. The two parties agree upon the principal–agent relationship through investment agreements. This principal–agent relationship will also lead to moral hazard and adverse selection. To avoid adverse selection, moral hazard, and control agency risk, the VC will send representatives to the board of directors of the startup enterprise to participate directly in business decisions [48]. First, the rich industry experience of the VC can help startup enterprises to make strategic plans, standardize financial disclosure systems, and broaden marketing channels, and improve the comprehensive competitiveness of startup enterprises. Second, the intervention of the VC optimizes the ownership structure of the board of directors and improves the efficiency of investment decision-making of the board of directors. Third, the VC’s rich IPO counseling experience and extensive social network can help startup enterprises avoid possible mistakes in the IPO audit, reduce listing costs, and improve the speed of passing. In addition, the formal audit of the registered track pays more attention to the material compliance of the reporting enterprise. This enterprise is not affected by the administrative intervention, regional development balance, and industry development policies in the approved track. Therefore, this paper argues that compared with the approval system, the registration system to send directors to a startup enterprise can effectively improve its governance level and comprehensive competitiveness and accelerate its speed of passing. To sum up, the following assumption is put forward:

Hypothesis 2 (H2).

After the trial of the registration system, a board of directors dispatched by the VC to the venture capital startup company can significantly accelerate the passing speed.

Information disclosure is the core of the registration system, and the prospectus is an important carrier of enterprise information disclosure. The goal is to require enterprises to provide more information to help investors understand the core competitiveness of enterprises. The information disclosure mechanism of the prospectus aims to improve the quality of information disclosure and reduce the degree of information asymmetry between investors and enterprises [50]. In the United States, the review of IPO materials is the responsibility of the Securities and Exchange Commission (SEC). The SEC usually sends inquiries to companies with low profits, high business complexity, and aggressive tax avoidance [51,52], requiring detailed disclosure of relevant details in the prospectus. The IPO audit in China is mainly the responsibility of the exchange. Existing literature shows that the prospectus disclosure mechanism of Chinese exchanges has played an active role in information disclosure supervision [53] and improved the quality of information disclosure [3]. Specifically, the information disclosure mechanism can reduce the degree of management of earnings of enterprises; improve audit quality; inhibit the purchase behavior of internal control opinions of enterprises, enhance CSR [54]; identify potential risks in the process of listing audit; reduce bid–ask spreads, analysts’ earning prediction errors, and analysts’ optimism after inquiry; and improve the enthusiasm and prediction accuracy of performance forecast but increase the listing cost of the said companies [55]. According to past IPO experience, a VC improves the completeness and detail of risk warnings in the prospectus to reduce the rounds and duration of inquiries for startups and then speed up the meeting of startups [56]. In addition, the formal examination of the registration system depends more on the voluntary disclosure degree of startup enterprises. On the one hand, more detailed voluntary disclosure of startup enterprises can effectively reduce information asymmetry; on the other hand, detailed disclosure of information can reduce the information acquisition cost of regulators, reduce work intensity, and improve examination and approval efficiency. To sum up, the following assumption is made:

Hypothesis 3 (H3).

After the trial of the registration system, the more detailed the risk information disclosure degree of VC-owned startups, the faster the speed of passing.

3. Study Design

3.1. Sample and Data Selection

The sample was selected from companies that had declared IPOs from 2014 to 2019 and were listed on the NEEQ. The companies that had declared IPOs to the CSRC and passed the review of the Stock Issuance Review Committee (SIRC) during the sample period were referred to as the passing sample, while startup enterprises that had completed listing after the meeting were called the listing sample. The data on the IPO and the listing data were collected manually from the websites of CSRC, the NEEQ, the Shanghai Stock Exchange, and the Shenzhen Stock Exchange. The data on the shareholdings of venture capitalists were collected manually by searching the prospectuses (filings) of IPO companies on the website of the CSRC, and other financial data were obtained from the Wind database. After excluding companies with missing data, we received 299 samples of companies with passed IPOs and 200 samples of listed companies. All continuous variables were tail-shrunk by a two-sided 1% quantile to eliminate the effect of outliers.

Table 1. Annual distribution of the sample.

Annual	VC	Sample Passing		Sample Listing
		Approval	Registration	Approval	Registration
2014	With VC	0	0	0	0
	NO VC	0	0	0	0
2015	With VC	4	0	1	0
	NO VC	1	0	2	0
2016	With VC	1	0	1	0
	NO VC	1	0	0	0
2017	With VC	21	0	9	0
	NO VC	24	0	10	0
2018	With VC	17	0	10	0
	NO VC	39	0	12	0
2019	With VC	17	9	8	7
	NO VC	25	20	19	7
2020	With VC	11	31	15	22
	NO VC	28	45	26	32
2021	With VC	0	0	2	2
	NO VC	0	5	5	10
Total		189	110	120	80
VC accounted for (%)		30.10	36.36	36.67	38.75

shows the annual distribution of the sample. As the registration system was piloted on the STAR and GEM boards one after another after 2019, in all, 189 companies (63.21%) filed IPOs under the approved system board, and 110 companies (36.79%) filed under the registration system board in the passing sample. In the listing sample, 120 companies (60%) declared IPOs under the approved system, and 80 companies (40%) declared IPOs under the registered system. In addition, the number of companies filing IPOs in the registered segment increased significantly and exceeded the number of companies filing in the approved segment in 2020.

3.2. Model and Variable Definitions

In this paper, the passing probability of a startup enterprise adopts logistic regression. OLS regression is used for the speeds of passing (Speed 1) and listing (Speed 2) of the startup enterprise. Drawing lessons from the research of Chan et al. [57] and Jin et al. [37], this paper constructs the following (mode l) to test the influence of VC shareholding and its characteristic variables on the probability and speed of passing and listing.

$${\mathrm{I}\mathrm{P}\mathrm{O}}_i={\mathsf{\alpha }}_i+{\beta }_i\times {VC}_i+{\gamma }_i\times {Control}_i+{\epsilon }_i$$

(1)

3.2.1. Dependent Variable

In this paper, logistic regression is used for the probability that an IPO company will pass (Pass) and OLS regression is used for the speeds at which the startup enterprise was passed (Speed 1) and listed (Speed 2).

Drawing on Allen and Faulhaber [58] and Xiong et al. [40], the dummy variable Pass is used for whether a startup enterprise passes or not: 1 is assigned if the startup enterprise passes the audit by the issuance committee (listing committee) and 0 is assigned if it does not.

Drawing on a related study by Zeng et al. [35], the speed of passing (Speed 1) refers to the number of days between the date the IPO application of a startup enterprise is accepted and the review and approval date of the issuing and examination committee (listing committee) and the speed of listing (Speed 2) refers to the interval between the review and approval date of the issuing and examination committee (listing committee) and the gap between the date the company is listed and the date the issuing committee clears the company.

3.2.2. Independent Variable

We draw on Wu et al. [59] to define a VC. The dummy variable of VC support (VC), which refers to the period from the entry of shares before the restructuring of the startup enterprise until one year before the IPO meeting, is assigned a value of 1 if there is VC support in the top 10 shareholders and a value of 0 if not. Because establishing a joint-stock company is the starting point of a company’s IPO planning, its purpose is to bring the company’s organizational structure and shareholding structure more in line with the listing requirements of the CSRC to prepare the company’s listing plan. The reasons for confirming whether there is VC shareholding at the time of shareholding reformation are as follows: (1) the establishment of the change of a joint-stock company is the starting point of an enterprise’s listing planning, and its purpose is to bring the company’s organizational structure and shareholding structure more in line with the listing requirements of the CSRC to prepare the company’s listing plan; (2) a VC enters a startup enterprise before the establishment of the shareholding system, and the VC has a stronger supervisory and management effect on the startup enterprises, a better understanding of internal information, such as the substantive operation and financial status of the startup enterprise, and stronger motivation to promote the listing of the enterprise; and (3) after the startup enterprise is listed on the NEEQ, more equity investment capital will directly hold shares in the startup enterprise through the stock transfer system, but this equity investment capital tends to hold a smaller percentage of shares, holds shares for a longer period of time, and does not participate in corporate governance and other phenomena; therefore, this paper excludes the equity investment capital that enters after the restructuring interference with the empirical results.

3.2.3. Mechanism Variables

In this paper, the analysis mechanism is divided into three parts. (1) Drawing on relevant studies by Dyer et al. [60], we use text analysis techniques to measure risk disclosure (Risk), i.e., the frequency of words such as “risk”, “uncertainty”, “volatility”, and “instability” in the prospectus of a startup enterprise (meeting drafts) as a percentage of the total number of words in the prospectus (meeting drafts). (2) A dummy variable for the presence of directors and supervisors of VCs (Board) is given the value of 1 if a VC has a director and a supervisor in the startup enterprise; otherwise, it is given a value of 0.

3.2.4. Control Variables

In this paper, to control the influence of other factors on the regression results, the control variables include (1) firm size (Size), profitability (ROE), financial leverage (Lev), growth capacity (Growth), profitability of sales (ROS), and net profit growth rate (NI), drawing on the relevant studies by Dai et al. [36]; (2) shareholding structure (Bshare): the shareholding ratio of the largest shareholder of the startup enterprises one year before the IPO meeting (listing); (3) shareholding balance (Bal): the sum of the shareholding ratio of the second-to-tenth-largest shareholders of the startup enterprises one year before the IPO meeting (listing) divided by the shareholding ratio of the largest shareholder; (4) equity balance (Bal): the sum of the shareholding ratio of the second-to-tenth-largest shareholders divided by the shareholding ratio of the largest shareholder one year before the IPO; and (5) industry dummy variable (Industry): determined according to the industry code of the CSRC.

4. Empirical Results

4.1. Descriptive Statistics

Table 2. Variable definitions and descriptive statistics.

Variables	Mean	Median	Sd	Max	Min
Pass	0.676	1.000	0.468	1.000	0.000
Speed1	428.306	385.000	237.0266	1607.000	66.000
Speed2	89.111	77.000	62.185	27.000	210.000
VC	0.350	0.000	0.477	1.000	0.000
Risk	0.003	0.003	0.001	1.000	0.000
Board	0.274	0.000	0.446	1.000	0.000
Reputation	0.220	0.000	0.415	1	0
Size	8.874	8.838	0.329	10.049	8.112
ROE	0.211	0.195	0.116	0.712	−0.332
Lev	0.354	0.348	0.154	0.847	0.035
Growth	1.127	0.184	15.242	263.755	−0.447
ROS	0.113	0.073	0.479	3.987	−4.153
NI	0.353	0.261	0.868	5.620	−8.334
Bal	1.629	1.142	5.726	99.000	−0.977
Bshare	0.425	0.401	0.171	0.999	0.009
Reg_IPO	0.368	0.000	0.483	1.000	0.000

shows that the mean value of the probability of passing (Pass) was 67.6%, and the median was 100%. The mean value of the speed of passing screening (Speed 1) was 428.306 days, and the median was 385 days. The mean value of the speed of listing (Speed 2) was 89.111 days, and the median was 77 days. The standard deviations of both the speed of passing screening and the speed of listing were large. This indicates the existence of startup enterprises with a slower speed of passing and significant differences in the speed of passing among different startup enterprises. The mean value of VCs with VC holdings was 0.35, indicating that 35% of the startup enterprises in the sample had VC holdings. In addition, 27.4% of the VCs in the sample assigned directors and supervisors to startup enterprises.

4.2. Univariate Tests

Table 3. Univariate and Pearson test.

Panel A: Univariate Test of VC
		Pass		Speed1		Speed2
		Mean	Differences	Mean	Differences	Mean	Differences
Full sample	With VC	0.762	−0.133 **	382.721	70.150 **	73.549	25.006 ***
	No VC	0.628		452.872		98.555
Registration	With VC	0.794	−0.113	266.333	143.269 **	70.166	28.333 ***
	No VC	0.681		409.603		98.500
Approval	With VC	0.741	−0.143 *	451.409	24.761	76.024	22.567
	No VC	0.598		476.170		98.591
With VC	Registration	0.794	−0.603	266.333	5.671 ***	70.166	0.769
	Approval	0.741		451.409		76.024
Panel B: Pearson correlation coefficient analysis
	VC	Pass	Speed1	Speed2	VCnum	Fshare	Syn
VC	1.000
Pass	0.137 **	1.000
Speed1	−0.141 **	0.210 ***	1.000
Speed2	−0.195 ***	0.313 **	0.147 **	1.000
VCnum	0.799 ***	−0.005	−0.099 *	−0.161 **	1.000
Fshare	0.703 ***	−0.015	−0.130 **	−0.186 **	0.653 ***	1.000
Syn	0.577 ***	0.108 *	−0.112 *	−0.113	0.788 ***	0.474 ***	1.000

Note: * p < 0.1, ** p < 0.05, *** p < 0.01.

, Panel A, presents the univariate tests of VC shareholding. First, in the full-sample test, startup enterprises with VC holdings had a significantly higher probability of passing the meeting (76.2%) than those without VC holdings (62.8%). Startup enterprises with VC holdings passed the meeting (382.721 days) and listing (75.549 days) significantly faster than those without VC holdings (452.872 days) and listing (98.555 days). Second, for startup enterprises filing IPOs in the registered segment, those with VC backing passed the meeting significantly faster (266.333 days) than those without VC backing (409.603 days). In the sample of startup enterprises with VC backing, those filing IPOs in the registered segment passed the meeting significantly faster (266.333 days) than those in the approved IPO system (451.409 days). Finally, the speed to market of startup enterprises filing IPOs in the registered system segment with VC support was significantly faster (98.5 days) than that of companies without VC support (70.166 days). Hypothesis 1 was initially tested.

4.3. Variable Correlation Coefficient Analysis

Table 3 (Panel B) shows the correlation coefficient matrix between the main variables. There was a significant positive correlation between VC and the probability of passing (Pass) (r = 0.137, p < 0.05). There was a significant negative correlation between VC and the speed of passing (Speed 1) and the speed of listing (Speed 2) (r = −0.141, p < 0.05; r = −0.195, p < 0.01). There were significant negative correlations between the number of VC (VCnum), lead VC shareholdings (Fshare), and VC co-investments (Syn) with the speeds of passing (Speed 1) and listing (Speed 2).

5. Regression Analysis Result

5.1. Main Regression

The natural logarithm of the speeds of passing the meeting (Speed 1) and listing (Speed 2) was taken in the regression to reduce the interference of their excessive standard deviation in the regression results.

As shown in

Table 4. Main regression analysis.

	Full Sample			Registration		Approval
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	Pass	Speed1	Speed2	Speed1	Speed2	Speed1	Speed2
VC	0.683 **	−0.297 ***	−0.242 ***	−0.569 ***	−0.383 ***	−0.132	−0.157
	(2.172)	(−3.720)	(−3.251)	(−3.210)	(−3.605)	(−1.591)	(−1.600)
Size	3.471 ***	0.525 ***	−0.272 *	0.437	−0.694 ***	0.360 ***	−0.065
	(4.404)	(3.935)	(−1.817)	(1.229)	(−3.594)	(2.995)	(−0.326)
ROE	13.103 ***	1.015 ***	−0.600	0.510	−0.910	1.004 **	−0.149
	(4.276)	(2.805)	(−1.403)	(0.815)	(−1.452)	(2.402)	(−0.265)
Lev	−4.042 ***	−0.583 **	0.538 *	1.115 **	1.255 ***	−0.369	0.184
	(−3.081)	(−2.008)	(1.886)	(2.181)	(2.987)	(−1.460)	(0.462)
ROS	0.738	0.068	0.113	0.066	−0.075	−0.653 **	−0.680
	(0.694)	(0.478)	(0.641)	(0.318)	(−0.295)	(−2.090)	(−1.528)
NI	−0.825	−0.096	−0.115	−1.128 **	−0.007	0.313	0.541 *
	(−1.028)	(−1.054)	(−1.018)	(−2.543)	(−0.044)	(1.312)	(1.661)
Growth	0.017 **	−0.001	0.001	−0.002	0.003 *	−0.660 **	−0.977 **
	(2.545)	(−1.308)	(1.104)	(−0.718)	(1.858)	(−1.975)	(−2.323)
Bshare	−1.539	−0.097	−0.004	0.588	−0.486	0.015	0.225
	(−1.075)	(−0.436)	(−0.007)	(0.640)	(−0.467)	(0.083)	(0.334)
Bal	−0.257	−0.003	−0.045	0.199	−0.085	−0.004 ***	−0.014
	(−1.062)	(−1.632)	(−0.521)	(1.530)	(−0.623)	(−3.524)	(−0.146)
Reg_IPO	0.439	−0.444 ***	0.015
	(1.419)	(−4.935)	(0.197)
Constant	−3.344 ***	2.319 *	6.905 ***	2.153	10.738 ***	2.927 ***	4.980 ***
	(−4.334)	(1.938)	(5.223)	(0.594)	(6.029)	(2.793)	(3.013)
Industry	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	299	299	200	110	80	189	120
R2/Pse-R2	0.207	0.331	0.106	0.338	0.289	0.099	0.080

Note: industry-fixed effects and year-fixed effects are included; t-values in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01, same later.

, first, in the full-sample regression, column (1) tests the effect of the presence or absence of VC holdings on the probability of startup enterprises passing the meeting and the results show that the probability of startup enterprises passing the meeting with venture capital holdings was 68.3% higher than that of companies without VC holdings and was significantly positive at the 95% level. Column (2) tests the effect of the presence or absence of VC holdings on the speed of startup enterprises passing the meeting. The results show that startup enterprises with VC holdings passed the meeting in 29.7% fewer days than those without VC support and were significantly negative at the 99% level. Column (3) tests the effect of the presence of startup enterprise holdings on the speed of startup enterprises. The number of days to IPO for companies with VC holdings was fewer by 24.2% than for companies without VC support and was significantly negative at the 99% level. The results in columns (1), (2), and (3) validate Hypothesis 1, indicating that VC-holding startup enterprises can release certification signals to the outside world, diminish the degree of information asymmetry between startup enterprises and regulators, and increase the probability of firms passing the IPO screening process and speed up passing and listing.

Second, to explore the impact of differences in listing systems on the IPO process of startup enterprises, this paper divides the sample into two registration system and approval system subsamples. When companies file IPOs in the registered system, companies with VC support can shorten the time required to pass the meeting by 56.9% and the speed of listing by 38.3% compared to companies without VC support, as shown in columns (4) and (5). Both are significantly negative at the 99% level. Finally, as shown in columns (6) and (7), for startup enterprises filing IPOs in the approved segment, VC holdings do not significantly shorten the speed of passing and listing of these enterprises. The results in columns (4) and (5) are consistent with the venture capital certification hypothesis proposed by Megginson and Weiss [7] and Barry et al. [24]. The results indicate that regulators better recognize the certification effect of VC holdings under the registration system and market investors, which accelerates the speed at which IPO firms pass the IPO review and are listed. However, under the approval system environment, the certification effect of VC holdings is difficult due to excessive administrative intervention and complicated approval procedures, which affect the speed at which startup enterprises pass the IPO review and are listed.

5.2. The Mediating Effect of Risk Disclosure on the Speed of Passing

In this paper, we use the method of Wen and Ye [61] to test the mediation effect. The test model is as follows:

$${\mathrm{R}\mathrm{i}\mathrm{s}\mathrm{k}}_i={\alpha }_i+{\beta }_i\times {VC}_i+{\gamma }_i\times {Control}_i+{ϵ}_i$$

(2)

$${\mathrm{S}\mathrm{p}\mathrm{e}\mathrm{e}\mathrm{d}1}_i={\mathsf{\alpha }}_i+{\beta }_i\times {VC}_i+{\delta }_i\times {\mathrm{R}\mathrm{i}\mathrm{s}\mathrm{k}}_i+{\gamma }_i\times {Control}_i+{\epsilon }_i$$

(3)

As shown in Panel A of

Table 5. The impact of risk disclosure on the speed of startup enterprises’ passing.

Panel A: Intermediary Test
	Full Sample				Registration
	(1)	(2)	(3)	(4)	(5)	(6)
	Risk	Speed1	Risk	Speed1	Risk	Speed1
VC	0.000 ***	−0.213 ***	0.000 ***	−0.203 ***	0.001 ***	−0.303 **
	(3.172)	(−2.828)	(3.460)	(−2.753)	(2.846)	(−2.419)
Risk		−276.387 ***		−222.740 ***		−289.512 ***
		(−4.787)		(−4.334)		(−3.658)
Size			−0.000 **	0.447 ***	−0.000	0.501 **
			(−2.420)	(3.377)	(−0.675)	(2.290)
ROE			0.000	1.026 ***	−0.000	0.791
			(0.127)	(2.957)	(−0.076)	(1.594)
Lev			0.000	−0.556 *	−0.001	−1.119 **
			(0.363)	(−1.945)	(−1.012)	(−2.411)
ROS			−0.000	0.050	−0.000	0.089
			(−0.371)	(0.338)	(−0.230)	(0.554)
Growth			0.000 **	−0.001	0.000	−0.001
			(2.127)	(−0.798)	(0.722)	(−1.016)
NI			0.000	−0.084	0.000	−0.035
			(0.444)	(−0.866)	(0.539)	(−0.394)
Bshare			0.001 *	0.038	0.003	1.315 **
			(1.753)	(0.177)	(1.506)	(2.183)
Bal			0.000	−0.002	0.000	0.193 *
			(1.202)	(−1.212)	(0.406)	(1.925)
Reg_IPO			0.000 ***	−0.356 ***
			(3.079)	(−3.946)
Industry	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
constant	0.004 ***	8.040 ***	0.007 ***	3.837 ***	0.004	1.706
	(9.146)	(5.278)	(5.105)	(3.116)	(0.960)	(0.881)
N	299	299	299	299	110	110
R2	0.204	0.297	0.264	0.390	0.316	0.370
Panel B: Bootstrap test
Models	Type	Value	Ind/Tot	S.D.	Z-value	Confidence
(1)–(2)	Ind eff	−0.085	0.267	0.031	−2.72	[−0.155, −0.032]
	Tot eff	−0.318		0.078	−4.07	[−0.474, −0.167]
(3)–(4)	Ind eff	−0.074	0.253	0.030	−2.52	[−0.144, −0.026]
	Tot eff	−0.292		0.075	−3.89	[−0.439, −0.147]
(5)–(6)	Ind eff	−0.272	0.472	0.085	−3.19	[−0.454, −0.129]
	Tot eff	−0.576		0.143	−4.03	[−0.846, −0.277]

Note: * p < 0.1, ** p < 0.05, *** p < 0.01.

, in the full-sample regressions, the coefficients of VC in columns (1) and (3) were significantly positive, indicating that startup enterprises with VC backing had a higher degree of risk disclosure. In contrast, the coefficients of VC and Risk in columns (2) and (4) were both significantly negative, indicating the existence of a mediated transmission path of risk where startup enterprises with VC backing shortened their time to pass the meeting by increasing the degree of risk disclosure in the prospectus. In the registration regression, the coefficients of VC and Risk in columns (5) and (6) were significant, indicating that the intermediary path of Risk existed in the registration regression. Its intermediary effect value (Ind/Tot = 0.472) was larger than that of the full-sample regression (Ind/Tot = 0.253), indicating that the participation of VCs in companies filing IPOs in the registration board could effectively improve the risk information disclosure in the prospectus and reduced the information asymmetry with the regulator, thereby speeding up the startup enterprises’ passing rate. In addition, to test the reliability of the intermediary regression results in this paper, the bootstrap method with 5,000 replications was implemented and the results are shown in Panel B. The Sobel tests of all models were significant, and the confidence intervals did not include 0, indicating that the intermediary test results were more reliable.

5.3. The Moderating Effect of a VC’s Presence on the Board of Directors and Supervisors and the Speed of Passing

As shown in

Table 6. Moderating effect of the presence of directors and supervisors of VCs on the speed of passing.

	(1)	(2)	(3)	(4)
	Speed1	Speed1	Speed1	Speed1
VC	−0.015	−0.031	−0.027	0.005
	(−0.146)	(−0.318)	(−0.159)	(0.047)
Board	0.577 **	0.591 *	1.147 ***	0.257
	(2.185)	(1.922)	(8.390)	(0.686)
VC*Board	−0.985 ***	−0.940 ***	−1.820 ***	−0.450
	(−3.385)	(−2.880)	(−7.946)	(−1.149)
Size		0.436 ***	0.672 ***	0.362 ***
		(3.987)	(2.711)	(2.985)
ROE		0.841 ***	0.704	1.012 **
		(2.700)	(1.262)	(2.408)
Lev		−0.621 ***	−1.228 **	−0.312
		(−2.619)	(−2.549)	(−1.251)
ROS		0.050	0.175	−0.642 **
		(0.389)	(1.011)	(−1.986)
NI		−0.073	−0.063	0.287
		(−0.921)	(−0.681)	(1.149)
Growth		−0.001	−0.001	−0.629 *
		(−0.911)	(−1.092)	(−1.859)
Bshare		0.071	0.932	0.036
		(0.390)	(1.504)	(0.192)
Bal		−0.003 **	0.178 *	−0.004 ***
		(−2.306)	(1.710)	(−3.240)
Reg_IPO		−0.408 ***
		(−5.460)
Industry	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
Constant	5.989 ***	2.296 **	−0.509	2.871 ***
	(4.645)	(2.430)	(−0.241)	(2.733)
N	299	299	110	189
R2	0.111	0.256	0.336	0.117

Note: * p < 0.1, ** p < 0.05, *** p < 0.01.

, (1) and (2) are full-sample regressions. The cross multipliers VC*Board for VC share-holding (VC) and Board presence (Board) were significantly negatively correlated at the 0.05 level with coefficients of −0.985 and −0.940, respectively, which indicates that there was a significant negative relationship between VC shareholding, board presence, and the speed of passing of the startup enterprise; the presence of VC shareholding and board in the startup enterprise can significantly shorten the company’s speed of passing, supporting the hypothesis of the value-added effect of venture capital proposed by Hellmann and Puri [62]. Meanwhile, the sample was divided into a registration system (column (3)) and an approval system (column (4)). Only under the registration system was the cross-product term VC*Board in column (3) significantly negative. In contrast, the cross-product term VC*Board for companies filing IPOs under the approved system was not significant. This indicates that in startup enterprises filing IPOs in the registration system segment, they have to comply with various disclosure requirements for listing, meaning VCs assigned directors and supervisors have more significant supervisory and governance effects and can increase the speed of passing the meeting more significantly than the companies in the approved system environment.

5.4. Extended Analysis

5.4.1. Timing Capabilities of VC

As shown in

Table 7. Timing capabilities of VC.

	(1)	(2)	(3)	(4)
	Speed2	Speed2	Speed2	Speed2
VC	−0.171 *	−0.299	−0.298 **	−0.124
	(−1.942)	(−0.761)	(−2.249)	(−0.715)
Size	−0.114 **	0.828	−0.520 *	0.171
	(−1.996)	(0.788)	(−1.792)	(0.572)
ROE	−0.354	1.217	−0.047	0.803
	(−0.801)	(0.565)	(−0.058)	(0.884)
Lev	0.559	−0.254	1.080 *	0.341
	(1.531)	(−0.242)	(1.719)	(0.434)
ROS	0.368	−0.069	0.355	−1.749 **
	(1.423)	(−0.091)	(0.855)	(−2.123)
NI	−0.275 *	−0.095	−0.294	0.991 *
	(−1.680)	(−0.255)	(−1.185)	(1.707)
Growth	0.000	−0.066	0.000	−2.294 ***
	(0.114)	(−0.085)	(0.070)	(−3.552)
Bshare	−0.546	0.107	−0.835	−0.864
	(−0.820)	(0.047)	(−0.624)	(−0.550)
Bal	−0.143	−0.049	−0.082	−0.259
	(−1.307)	(−0.233)	(−0.607)	(−1.022)
Reg_IPO	0.003	0.349
	(0.035)	(0.840)
Industry	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
Constant	6.499 ***	−2.852	8.528 ***	4.082
	(4.562)	(−0.332)	(3.468)	(1.508)
N	155	45	74	81
R2	0.098	0.633	0.552	0.607

Note: * p < 0.1, ** p < 0.05, *** p < 0.01.

, column (1) is the IPO hot season and column (2) is the IPO off-season. The regression results show that the coefficient of VC in column (1) was −0.171 and significant at the 90% level, while the coefficient of VC in column (2) was not significant, indicating that VCs have a stronger timing ability during the IPO hot season and could induce startup enterprises to go public during the hot season to obtain higher issuance proceeds. Columns (3) and (4) were regressions of the IPO hot season sample in column (1) on the registration system and the approval system, respectively, to distinguish which listing system had a stronger timing effect on VCs. The results show that only the coefficient of VC in column (3) was −0.298 and significant at the 95% level, indicating that the timing effect of VCs was stronger for firms that go public in the registration system environment.

5.4.2. The Impact of Political Affiliation on the Speed of Passing

Firms with political connections can use their good relationship with the government to help them pass the CSRC review and obtain an IPO status. Startup enterprises can use the political relations of VCs to enhance communication with the regulator, to understand and grasp the policy trends of the regulator in a timely manner, and to be more focused in preparing application materials, thus making it easier for the enterprises to pass the regulator’s review [63]. In this paper, political affiliation (Pol) is defined as 1 when the actual controller of the VC is the government or state-owned capital or when the VC partner has served as a civil servant, an NPC deputy, or a CPPCC member; otherwise, it is defined as 0. As shown in

Table 8. Impact of political affiliation on the speed of passing.

	Full Sample		Registration		Approval
	(1)	(2)	(3)	(4)	(5)	(6)
	Speed1	Speed1	Speed1	Speed1	Speed1	Speed1
	Pol = 1	Pol = 0	Pol = 1	Pol = 0	Pol = 1	Pol = 0
VC	−0.318	−0.130	−0.573	−0.399 **	−0.410 **	−0.008
	(−1.116)	(−1.462)	(−0.607)	(−2.527)	(−2.142)	(−0.078)
Size	0.291	0.384 **	−0.209	0.277	0.136	0.415 **
	(0.643)	(2.419)	(−0.075)	(0.923)	(0.542)	(2.299)
ROE	1.688	0.673 *	2.485	0.748	2.511 *	0.893 *
	(0.755)	(1.790)	(0.481)	(1.164)	(1.820)	(1.916)
Lev	0.381	−0.707 **	−0.554	−1.323 ***	−0.369	−0.127
	(0.298)	(−2.186)	(−0.112)	(−2.650)	(−0.569)	(−0.380)
ROS	0.487	−0.011	4.646	−0.100	−1.126	−0.419
	(0.775)	(−0.078)	(1.043)	(−0.705)	(−0.549)	(−0.887)
Growth	−0.178	−0.001	2.034	0.000	−1.000	−0.555
	(−0.178)	(−1.191)	(0.505)	(0.114)	(−0.500)	(−1.288)
NI	−0.665	−0.044	−3.515	0.053	0.304	0.258
	(−1.267)	(−0.498)	(−0.935)	(0.725)	(0.160)	(0.751)
Bshare	−0.212	−0.160	2.887	−0.111	0.006	0.154
	(−0.132)	(−0.587)	(0.445)	(−0.152)	(0.007)	(0.499)
Bal	−0.066	−0.004 **	0.389	0.039	0.146	−0.002
	(−0.197)	(−1.997)	(0.337)	(0.351)	(0.978)	(−1.194)
Reg_IPO	−0.170	−0.580 ***
	(−0.705)	(−6.021)
Industry	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
Constant	2.782	3.689 **	5.910	3.504	4.679 *	3.187 *
	(0.600)	(2.598)	(0.249)	(1.274)	(1.869)	(1.984)
N	75	224	27	83	48	141
R2	0.483	0.369	0.666	0.203	0.269	0.246

Note: * p < 0.1, ** p < 0.05, *** p <0.01.

, the coefficient of column (5) was −0.410 and significant at the 95% level, indicating that VC support with political affiliation could significantly shorten the IPO process for companies filing IPOs on the approved board. This can substantially shorten the speed of a startup enterprise passing the review process.

5.4.3. IPO Performance

The above has shown that a VC could shorten the IPO process of the invested company. So, will the market performance of VC-backed companies be better after the IPO? This paper uses the market-adjusted underpricing rate (Unp) [64] and the cumulative excess return rate (CAR30) 30 trading days after listing [65]. Taking the institutional investor subscription ratio Inv and the underwriter subscription ratio (Unw) as the dependent variables [66], it is better to study the market performance of VC-backed companies after listing. The regression results are shown in

Table 9. IPO Performance.

	IPO Performance
	Full Sample				Registration				Approval
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)
	Unp	CAR30	Inv	Unw	Unp	CAR30	Inv	Unw	Unp	CAR30	Inv	Unw
VC	−0.435 **	0.039 *	0.045 **	0.041 **	−0.654 **	0.103 **	0.071 **	0.072 *	−0.118	−0.002	0.006	0.027
	(−2.284)	(1.871)	(2.098)	(2.002)	(−2.123)	(2.251)	(2.185)	(1.719)	(−1.108)	(−0.726)	(0.162)	(0.956)
Size	−0.022	0.031	−0.086 **	0.038	0.092	0.077 **	0.007	−0.034 **	−0.137	0.006 *	−0.157 *	0.038
	(−0.073)	(1.435)	(−1.997)	(1.016)	(0.107)	(2.211)	(0.094)	(−2.049)	(−1.101)	(1.883)	(−1.972)	(0.656)
ROE	−2.244 *	−0.135	−0.260 **	0.047	−6.401 **	−0.283	−0.072	−0.067	−0.718	0.000	−0.280	0.107
	(−1.964)	(−1.088)	(−2.245)	(0.500)	(−2.173)	(−0.998)	(−0.378)	(−0.441)	(−1.307)	(0.024)	(−1.558)	(0.765)
Lev	−0.972	−0.018	−0.026	−0.138	−3.817	−0.154	−0.291 **	−0.140	−0.025	0.003	0.113	−0.130
	(−1.026)	(−0.375)	(−0.361)	(−1.492)	(−1.347)	(−0.773)	(−2.045)	(−0.801)	(−0.089)	(0.395)	(0.698)	(−0.950)
ROS	0.033	−0.071	−0.143 **	−0.047	1.753	−0.065 **	0.063	−0.098	−0.457	0.004	−0.270	−0.112
	(0.089)	(−1.593)	(−2.369)	(−1.127)	(1.595)	(−2.901)	(0.671)	(−1.401)	(−1.457)	(0.559)	(−1.074)	(−1.048)
Growth	−0.003	0.000 *	0.000	0.000	−0.013 *	0.000	−0.000	0.000	−0.274	0.007	−0.040	−0.131
	(−1.121)	(1.857)	(1.556)	(0.848)	(−1.756)	(0.262)	(−0.364)	(1.121)	(−0.959)	(1.025)	(−0.165)	(−0.956)
NI	−0.267	0.039	0.085 **	0.026	−1.086 *	0.053	−0.018	0.050 **	0.348	−0.005	0.210	0.097
	(−1.187)	(1.441)	(2.252)	(1.188)	(−1.978)	(0.657)	(−0.377)	(2.017)	(1.409)	(−1.097)	(1.008)	(1.109)
Bshare	−0.145	0.194 *	0.052	−0.206 *	3.685	0.647	−0.076	−0.136	−0.263	−0.005	0.070	−0.165
	(−0.175)	(1.852)	(0.594)	(−1.878)	(1.205)	(1.586)	(−0.284)	(−0.496)	(−0.631)	(−0.458)	(0.536)	(−1.246)
Bal	−0.021	0.027	0.012	−0.012	0.430	0.083	−0.002	−0.015	−0.094	−0.000	0.015	−0.009
	(−0.177)	(1.602)	(0.642)	(−0.907)	(1.262)	(1.265)	(−0.062)	(−0.493)	(−1.628)	(−0.178)	(0.640)	(−0.488)
Reg_IPO	1.367 ***	−0.001	0.393 ***	0.039 *
	(6.872)	(−0.040)	(17.771)	(1.843)
Industry	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Constant	2.118	−0.379 *	0.914 **	0.571 *	3.017	−1.053	0.501	1.376 ***	2.390 **	−0.068 **	1.424 **	0.545
	(0.816)	(−1.755)	(2.322)	(1.751)	(0.469)	(−1.264)	(0.709)	(2.699)	(2.009)	(−2.231)	(2.085)	(1.118)
N	219	219	219	219	84	84	84	84	135	135	135	135
R2	0.488	0.244	0.736	0.384	0.518	0.376	0.427	0.421	0.545	0.450	0.401	0.396

Note: * p < 0.1, ** p < 0.05, *** p < 0.01.

. The results show that the coefficients of VC in both the full sample and the registered sample were significant. The coefficient of VC in the registered system was greater than that of the full sample VC, indicating that after the trial implementation of the registration system, startup enterprises with VC support were better recognized by the market, which was reflected in the lower underpricing rate of issuance, higher excess return within 30 days, higher subscription ratio of institutional investors, and higher subscription ratio of underwriters.

6. Robustness Test

6.1. Propensity Score Matching (PSM)

Considering that it is not necessarily venture capital holdings that shorten the time required for firms to pass the IPO review process and be listed, there may be some differences in firm characteristics between firms that receive VC holdings and those that do not receive VC holdings. These differences may also lead to variations in the speed at which startup enterprises pass the IPO screening and are listed. In this paper, we adopt the PSM method to eliminate the effect of such differences. First, with the existing command “pstest” in Stata 15.1, the matching feature variables are filtered through iterations. The pre- and post-matching effects in Figure 1 and Figure 2, respectively, show significant differences in the control and control groups before matching. At the same time, there were no significant differences after matching. The results in Panel A of

Table 10. Robustness tests.

Panel A: PSM
	Full Sample		Registration		Approval
	(1)	(2)	(3)	(4)	(5)	(6)
	lnspeed1	lnspeed2	lnspeed1	lnspeed2	lnspeed1	lnspeed2
VC	−0.292 ***	−0.227 *	−0.665 ***	−0.485 ***	−0.134	−0.109
	(−3.713)	(−1.760)	(−3.542)	(−3.787)	(−1.650)	(−0.764)
Control	Yes	Yes	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
Constant	3.285 ***	6.848 ***	−0.271	8.340 ***	3.691 ***	4.607 *
	(2.814)	(2.812)	(−0.080)	(3.094)	(3.035)	(1.982)
N	208	129	71	48	137	81
R2	0.288	0.320	0.292	0.406	0.158	0.147
Panel B: Heckman two-stage regression
	Full Sample		Registration		Approval
	(1)	(2)	(3)	(4)	(5)	(6)
	Speed1	Speed2	Speed1	Speed2	Speed1	Speed2
VC	−0.355 ***	−0.282 ***	−0.676 ***	−0.435 ***	−0.149	−0.200
	(−3.995)	(−3.421)	(−3.643)	(−3.601)	(−1.493)	(−1.652)
Imr	−0.168	−0.145	−0.340 *	−0.208	−0.004	−0.106
	(−1.545)	(−1.348)	(−1.883)	(−1.223)	(−0.029)	(−0.699)
Control	Yes	Yes	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
Constant	2.803 **	7.024 ***	4.535	10.900 ***	2.562 *	5.126 ***
	(2.253)	(5.253)	(1.209)	(5.979)	(1.855)	(3.263)
N	299	200	110	80	189	120
R2	0.338	0.113	0.360	0.299	0.332	0.085
Panel C: Regression results after excluding abnormal samples
	Full Sample		Registration		Approval
	(1)	(2)	(3)	(4)	(5)	(6)
	Speed1	Speed2	Speed1	Speed2	Speed1	Speed2
VC	−0.182 **	−0.207 **	−0.275 **	−0.314 **	−0.136	−0.203
	(−2.545)	(−2.556)	(−2.157)	(−2.333)	(−1.628)	(−1.535)
Control	Yes	Yes	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
Constant	3.385 ***	5.419 ***	6.857 ***	8.787 **	2.840 ***	5.982 **
	(2.837)	(3.755)	(2.778)	(2.600)	(2.690)	(2.382)
N	257	152	72	40	185	112
R2	0.511	0.173	0.552	0.297	0.101	0.364
Panel D: Bootstrap Test
	Full Sample		Registration		Approval
	(1)	(2)	(3)	(4)	(5)	(6)
	Speed1	Speed2	Speed1	Speed2	Speed1	Speed2
VC	−0.292 ***	−0.242 ***	−0.576 ***	−0.383 ***	−0.132	−0.157
	(−3.89)	(−3.26)	(−4.03)	(−3.49)	(−1.59)	(−1.60)
BC Confidence	(−0.433, −0.143)	(−0.400, −0.102)	(−0.846, −0.277)	(−0.611, −0.183)	(−0.306, −0.023)	(−0.347, 0.040)

Note: * p < 0.1, ** p < 0.05, *** p < 0.01.

show that the conclusion still holds after the propensity score matching screening of the samples.

6.2. IV Test

In this paper, we used Heckman and instrumental variables to address possible sample selection problems. Referring to Baik et al. [67] and Shen et al. [17], the inverse of the geographic distance between the main VC and the startup enterprise (Distance) was used as the instrumental variable and Distance was taken as 0 for the startup enterprise without VC support; the closer the geographic distance, the lower the supervision cost, and the higher the value of Distance. The calculation formula is as follows:

$$\mathrm{D}\left(\mathsf{\rho },\mathsf{\sigma }\right)=\arccos \left[\cos \left({\rho }_1\right)\cos \left({\rho }_2\right)\cos \left({\sigma }_1\right)\cos \left({\sigma }_2\right)+\cos \left({\rho }_1\right)\sin \left({\rho }_2\right)\cos \left({\sigma }_1\right)\sin \left({\sigma }_2\right)+\sin \left({\rho }_1\right)\sin \left({\sigma }_1\right)\right]\times \mathrm{R}$$

(4)

$$\mathrm{D}\mathrm{i}\mathrm{s}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{c}\mathrm{e}=\frac{1}{\mathrm{D}\left(\mathsf{\rho },\mathsf{\sigma }\right)}$$

(5)

where $\mathrm{D}\left(\mathsf{\rho },\mathsf{\sigma }\right)$ denotes the geographic distance between the master VC and the startup enterprise (in kilometers); ${\rho }_1$ and ${\sigma }_1$ (${\rho }_2$ and ${\sigma }_2$) denote the registered addresses of the leading VC and startup enterprise, respectively (accurate to prefecture level); latitude and longitude are in radians; and R is the Earth’s radius.

Since risk support (VC) is a dummy variable, this paper uses the Heckman two-stage approach for regression analysis. The first stage uses the Probit model to estimate the data of the passing sample and the listed sample separately to obtain the inverse Mills ratio Imr. Then, the Imr waited in the first stage is added to the regression model for estimation; the results are shown in Panel B of Table 10, and the conclusions remain robust.

6.3. Excluding the Stock Market Shutdown and Startup Enterprises’ Secondary Filing Sample

Zeng et al. [35] argue that when a startup enterprise declares an IPO again, the passing time of the startup enterprise is influenced due to its relatively better understanding of the whole listing process and the CSRC pre-screeners of the startup enterprise. So the relevant sample is excluded from this paper. Since A-shares suffered a stock market crash in June 2015, the CSRC announced a moratorium on the listing and issuance of shares by enterprises from 4 July 2015 until the resumption of issuance on 6 November 2015, so this paper excludes the samples declared during this period. The regression results after exclusion are shown in Panel C of Table 10. The regression results are consistent with the baseline regression findings, and the conclusions of this paper remain robust.

6.4. Bootstrap Test

The previous paper constructs a linear model with OLS for regression analysis; however, the small sample size of this paper due to the small number of companies listed on the NEEQ transfer board easily leads to the problem of model overfitting. Accordingly, this paper further considers using the Bootstrap method to resample the sample to obtain confidence intervals of the parameters to judge the significance of the regression results. Specifically, the baseline regression was tested again in this paper. Each test process was repeated 5000 times with put-back sampling to return the 95% BCa confidence intervals corresponding to the regression coefficients of the explanatory variables. It was decided that the original hypothesis would be accepted or rejected on the basis of whether the confidence interval contained 0. The results of the critical variables are shown in Panel D of Table 10. The regression coefficients and significance obtained using the bootstrap method are generally consistent with those obtained using the previous test, demonstrating the robustness of the benchmark regression.

7. Discussion

First, from the perspective of IPO allocator efficiency, this paper evaluates the initial effect of China’s capital market registration system reform to answer the theoretical dispute about the advantages and disadvantages of the registration and approval systems. The question of which is better between the registration system and the approval system has not been settled, and to solve this dispute, we need to study and judge the actual effect of the registration system reform. The results of this paper show that the IPO allocation efficiency has improved after the registration system reform, which is manifested as a significant acceleration of the meeting speed of startups. It shows that the implementation of the registration system broadens the financing channels of startups, reduces the cost of listing, and promotes the efficiency of the financial market.

Second, as China’s capital market has developed and matured, it is crucial to standardize the governance system of listed companies and improve the quality of listed companies to promote high-quality economic development. Under this trend, giving full play to the subjective initiative of institutional investors in the governance system of listed companies is increasingly important to improve the governance level of listed companies. However, venture capitalism started relatively late in China and did not develop substantially until the late 1990s. As per previous studies, the role of venture capital in startups in emerging markets such as China may differ from that of venture capital in mature companies in developed countries. Chinese venture capitalists are keen to invest in enterprises at the pre-IPO stage or “invest and leave”, and playing the role of an “incubator” is difficult. This study finds that after the implementation of the registration system, the fairer listing review procedure enables startup companies with venture capital to actively participate in corporate governance and pass faster, which is manifested as higher disclosure levels, and the startup companies with venture capital appointing directors pass faster. Finally, as the largest economy among emerging market countries, China’s brave attempt to reform the issuance system is worthy of reference by other emerging market countries. China has been adhering to the tentative and phased reform steps; that is, the registration system is now tried out on the Science and Technology Innovation Board and the Growth Enterprise Board, and the comprehensive registration system is being gradually implemented. Therefore, this special reform period, in which the registration system and the approval system coexist, provides a suitable empirical sample for this paper. A comparative study of the impact of venture capital under the registration system and the approval system on the process of startup enterprises finds that after the implementation of the registration system, the role of venture capital is more obvious. It is of great significance to give full play to the role of the “gatekeeper” of venture capital, promoting the smooth transition of registration system reform, and maintaining the sustainable development of the financial market.

8. Conclusions

The reform of the registration system is an essential institutional change in the development history of China’s capital market. From the perspective of resource allocation efficiency, this paper explored the microeconomic impact of this reform and answered the theoretical controversy about which is better, the registration system or the approval system. Taking 299 NEEQ enterprises as research samples, this paper analyzed the influence of VC shareholding on the listing process of startup enterprises under the special issuance system environment where the registration system and the approval system coexist. The hypothesis of certification and the value-added effect of venture capital in the process of issuance system reform is verified [7,62]. The empirical results show that: (1) after the trial registration system, the VC-owned enterprises have a higher probability of passing and a faster speed of passing and listing, which indicates that the certification role of VC reduces the degree of information asymmetry between VC-owned enterprises and regulators after the registration system reform. (2) After the trial registration system, governance measures such as sending directors to startup enterprises and improving the risk disclosure level in the prospectus can accelerate the speed of passing of startup enterprises, which shows that the supervision effect of a VC after the reform of the registration system can accelerate the speed of passing of startup enterprises. (3) Further research shows that after the trial registration system, the short-term performance indicators, such as underpricing rate, stock price volatility, and the excess return rate of startups, of enterprises with VC support are better than those of enterprises without VC support, indicating that VCs can better push high-quality enterprises to the market after the registration system reform. Under the registration system, VC-owned startups tend to go public in the hot IPO season, which shows that VCs can choose the right time to go public

Policy suggestions and enlightenment: first, improve the information disclosure system and strengthen research and development information disclosure. In the prospectus, mandatory disclosure provisions should be made for specific information, such as research and development expenditures, the number of patents, and the income from new products. For highly uncertain information and information that is difficult to describe quantitatively, predictive information disclosure criteria should be formulated to improve the quality of R&D information disclosure. Full disclosure through various channels increases the transparency of IPO pricing information; reduces issuers’ financing costs, investor risks, and information asymmetry; and facilitates investors to properly predict enterprise value and industrialization capacity. Second, strengthen the venture capital industry’s supervision and enhance the venture capital’s certification and supervision role. Guide the venture capital industry association to formulate industry rules and conventions and form a self-discipline mechanism. Improve the measures for the establishment, examination, approval, and administration of venture capital companies and venture capital funds. Guide and supervise venture capitalists to play the role of professional investment managers, improve their professional ability to screen IPO companies with potential and value, and improve the growth speed and quality of innovative enterprises. By giving play to the function of certification supervision, we can help enterprises closely combine market expectations with reasonable pricing to reduce the problem of high IPO underpricing. Finally, as a member of emerging market countries, China’s experience in market reform can provide reliable empirical evidence for the market-oriented process of emerging market countries and has guiding significance for the sustainable development of financial markets in emerging market countries.

9. Limitations and Prospects

Subject to the progress of China’s registration system and the difficulty of data acquisition, this paper can promote the research design from three aspects in the future. First, this paper studies the impact of venture capital support on the process. However, multiple venture capitalists often invest in the same startup in the form of joint investment to improve portfolio richness, reduce investment risk, and improve the exit rate of investment. Therefore, future studies can compare the impact of venture capital joint investment on the listing process of startups under the registration system and the approval system, for example, the different combinations of joint investment partners, individual VC and corporate VC joint, and individual VC and government VC joint, to study the impact of different types of venture capital skills and strategic goals on the listing process of startups. Second, this paper only studies the changes in the role of venture capital in the trial period of China’s registration system. However, horizontal comparisons within a single economy cannot fully illustrate the significance of registration reform for financial development and stability in emerging markets. Therefore, future research could be expanded to study economic system differences and cultural differences. For example, a comparative study could be carried out examining the differences in registration system audit between mainland China and Hong Kong and the impact of economic system differences on registration system audit under the premise of controlling cultural differences. The data from Germany and mainland China (the coexistence of a registration system and an approval system) can also be compared to study the impact of cultural differences on the role of venture capital.