Abstract
CHINA JOURNAL OF ACCOUNTING STUDIES 2020, VOL. 8, NO. 2, 272–297 https://doi.org/10.1080/21697213.2020.1822024 ARTICLE Transportation infrastructure and resource allocation of capital market: evidence from high-speed rail opening and company going public a a b Zhi Jin , Liguang Zhang and Qingquan Xin a b School of Accountancy, Southwestern University of Finance and Economics, Chengdu, China; School of Economics and Business Administration, Chongqing University, Chongqing, China ABSTRACT KEYWORDS Transportation Using the high-speed rail opening of each city in China as a natural infrastructure; company experiment, we apply the difference-in-differences model to investi- going public; high-speed rail gate how the transportation infrastructure in a region affects the opening; information behaviour of local company going public. We find that after high- asymmetry speed rail runs through a city, the number of local company going public increases significantly, and the approval probability of going public is improved significantly. Further mechanism analysis shows that the high-speed rail opening reduces the cost of obtaining pri- vate information about local companies, making it easier for them to absorb venture capital and hire high-quality intermediary institu- tions. In addition, the decline of information cost both enables external stakeholders to select better companies and lowers the financing cost of these companies. This paper shows that the improvement of transportation infrastructure can improve corporate financing efficiency and optimise the efficiency of resource allocation in capital markets. 1. Introduction China covers a vast territory with 9.6 million square kilometres, but the economic develop- ment between regions is very uneven. Listed companies are very important to local economic development, especially in areas with relatively backward economies, because they not only can bring much capital to the region but also increase employment and tax revenue. In terms of the number of listed companies, the land area of the four first-tier cities such as Beijing, Shanghai, Guangzhou and Shenzhen accounts for only 0.33% of China’s land area, but 27.47% of the listed companies concentrate in these cities. For three major city groups such as the Pearl River Delta, Yangtze River Delta and Beijing-Tianjin-Hebei, their land area accounts for only 2.31% of China’s land area, but 54% of listed companies are located here. This shows that the geographical distribution of listed companies in China is extremely unbalanced. If this situation continues to exacerbate, it will further widen the gap CONTACT Liguang Zhang zlg1990zlg@163.com School of Accountancy, Southwestern University of Finance and Economics, Chengdu, China Paper accepted by Kangtao Ye It is the distribution of A-share listed companies in China by the end of 2017. © 2020 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/ licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. CHINA JOURNAL OF ACCOUNTING STUDIES 273 between the rich and the poor regions, thus affecting the sustainable development of China’s economy. The reasons why listed companies in China are so concentrated in those places are as follows: Since the approval system for the public offering of stocks is launched in China’s capital market, filing an application package with the Chinese Securities Regulatory Commission (CSRC) for approval is necessary for listing; Theretofore, companies applying for listing need not only hire a broker for pre-listing counselling and stock underwriting after CSRC approval but also hire an audit firm and get support from institutional investors and analysts; More crucially, the CSRC staff also need to go to the company to investigate and collect evidence, and communicate with the company face to face; Only those approved by the CSRC can engage in an initial public offering. In other words, regulators, intermediaries (brokers, audit firms and law firms) and institutional investors need to conduct corporate site visits before CSRC approval, so the accessibility of the company’s location becomes very important. China’s two major stock exchanges are located in Shenzhen and Shanghai, the CSRC is located in Beijing, and audit firms, brokers and other intermediaries are mainly concentrated in the four first-tier cities such as Beijing, Shanghai, Guangzhou and Shenzhen. These four cities as well as other cities around them not only have the geographical advantage of being close to the regulators and inter- mediaries but also have the advantage of developed transportation infrastructure, mak- ing it easier for the local companies to reach them and agree with them, finally reducing the listing cost of the company and increasing the approval probability of listing. On the contrary, if companies are far from intermediaries, regulators and institutional investors, or transportation is inconvenient, communication costs will be more expensive. Then, it will be difficult for the company to obtain the support and understanding from them, thereby increasing the cost of listing and the risk of being denied. As a developing country in transition, China’s transportation infrastructure construction fell far behind most developed countries until the end of the last century. Since the 1990s, China has accelerated the pace of transportation infrastructure construction, of which the construction of high-speed railway (HSR) is the most remarkable. In 2008, China officially opened the Beijing–Tianjin intercity high-speed railway. By the end of 2017, China has built the world’s largest ‘four verticals and four horizontals’ high-speed rail network and con- tinues to build the ‘eight vertical and eight horizontal’ high-speed rail network with comprehensive coverage for the central and western regions. China’s high-speed rail construction starts from the four first-tier cities as well as other large cities, and gradually extends to other small and medium-sized cities. In other words, with the construction of high-speed rail, the improvement of transportation infrastructure makes the accessibility between local companies and regulators, intermediaries and institutional investors better, helping to reduce the cost of information access between each other. This effect is particularly pronounced in cities where transport infrastructure is relatively poor. Exploiting the opening of China’s high-speed rail as the natural experiment and using the 2004–2017 company listing application as the experimental scenario, we apply the difference-in-differences (DID) model to investigate the impact of the opening of high- The reasons for the unbalanced distribution of listed companies in China are quite complex and cannot be determined solely by geographical factors, but the geographical factors described here are at least an important reason that cannot be ignored. 274 Z. JIN, ET AL. speed rail on the number of local companies applying for listing, company’s approval probability of listing and the quality of the application company. We find that more local companies apply for listing after the opening of high-speed rail in one city compared with other cities that have not yet opened the high-speed rail, which is more pronounced in cities beyond the three major city groups (Pearl River Delta, Yangtze River Delta and Beijing-Tianjin-Hebei), cities whose HSR can go directly to Beijing, Shanghai, Guangzhou and Shenzhen, and cities without an airport. Furthermore, we find that the opening of high-speed rail reduces the cost for intermediaries and venture capitalists to have access to company information, making it easier for companies to obtain venture capital parti- cipation and hire high-quality intermediaries, thus enhancing the motivation and increase the approval probability of company going public. Our results also show that the opening of the high-speed rail reduces the cost for regulators, investors and intermediaries to obtain information from the applicants, allowing them to select higher-quality applicants and reducing the cost of financing when they go public. Our research indicates that the improvement of transportation infrastructure increases the motivation and the approval probability of IPO application for local companies. The main contributions of this paper are as follows. First, the research on the listing of the company mainly focuses on the Post-IPO market performance, such as IPO under- pricing (e.g. Allen & Faulhaber, 1989; Huang et al., 2016; Ljungqvist, 2007, etc.). These papers mainly examine how information asymmetry of investors in the secondary market affects resource allocation. The literature on the factors of approval probability of IPO application mainly examines how company’s political connection (Cai et al., 2013;Zhang et al., 2012), social relationship of intermediaries (Chen et al., 2014; Dai et al., 2014; Du et al., 2013; Li & Liu, 2012), and venture capital participation (Shen et al., 2013; Zeng et al., 2016) help companies to go public successfully, aimed at investigating the impact of software environment. We examine how the improvement of hardware environment, such as transportation infrastructure, can help local companies access capital market resources to uncover the impact of the hardware environment on capital market resource allocation. Based on the primary market participants, this paper enriches the existing literature on the company’s going public from the perspective of hardware environment. Second, some literature applies the relevant theory of new economic geography to investigate the influence of distance on the behaviour of companies. Initially, these studies focus on how the actual distance between the company and the relevant entities affects the behaviour of stakeholders. For example, Coval and Moskowitz (2001), Malloy (2005), Kedia and Rajgopal (2011), and Choi et al. (2012), respectively, look at how distance affects the behaviour of investors, analysts, regulators, and auditors. However, scholars gradually find that the study of the impact of geographical distance on the company’s economic behaviour faces some endogeneity problems. The airline and the opening of high-speed rail compress the space-time distance, which can be regarded as a substitute for distance and an exogenous shock of distance change, finally effectively making up for the defects of geographical distance. Some literature has examined the role of geogra- phical distance in the company’s economic behaviour from the perspective of airlines or The market environment contains two aspects: software environment and hardware environment. Software environ- ment refers to culture, system and social capital, etc., while the hardware environment is based on the actual conditions of transportation infrastructure, industrial plants and commercial entities. CHINA JOURNAL OF ACCOUNTING STUDIES 275 high-speed rail. For example, Giroud (2013), Chemmanur et al. (2014) and Bernstein et al. (2016) use opening airline as an external variable, respectively, to examine its impact on parent-subsidiary investment, corporate mergers and acquisitions, and VC investment; Huang et al. (2016), Long et al. (2017), and Zhao et al. (2018), respectively, examine the impact of the opening of high-speed rail on the company’s IPO underpricing, VC invest- ment and stock price crash. This paper takes the opening of China’s high-speed rail as the natural experiment, and examines how the improvement of transportation infrastructure affects the company’s financing behaviour to reveal how the company’s hardware envir- onment affects its financing capacity in the capital market, enriching the literature on new economic geography. Third, existing research shows that improvements in transportation infrastructure can boost economic growth (Liu & Li, 2017; Zhang, 2017; Zhou & Zheng, 2012), while economic growth depends on the economic benefits of micro-enterprises. By verifying that the opening of high-speed rail can improve the financing efficiency of local companies in the capital market, this paper reveals an important mechanism for transportation infrastructure to promote economic growth from the micro-level of the company. The rest of the paper is organised as follows: We introduce the institutional back- ground and develop hypothesis in the second part. The third part is the research design. The fourth part examines how the launch of HSR affects the behaviour and outcomes of company’s going public. Finally, the fifth part concludes. 2. Institutional background and theory analysis 2.1. Institutional background 2.1.1. Institutional background of HSR China’s high-speed rail construction is mainly to meet economic development and the movement of people. On the whole, the construction of China’s high-speed railway follows the order from big cities to medium and small cities, from the economically developed eastern coastal areas to the inland areas of the central and western regions. The Beijing–Tianjin intercity high-speed railway, which opened in 2008, is 120 km long and designed to travel at 350 km per hour, reducing the travel time between Beijing and Tianjin from 2 hours to about 30 minutes. In 2011, the Beijing–Shanghai high-speed railway reduced the travel time between Beijing and Shanghai from the original 10 hours to less than 5 hours, greatly compressing the space-time distance and improving the mobility of city people. Regionally, by the end of 2011, 43 cities had opened high- speed rail in the eastern region, while the number of HSR cities is only 26 in the central and western regions, lagging far behind the eastern region. By the end of 2018, 222 cities across the country had opened high-speed rail, including 89 cities in the eastern region, 75 in the central region and 58 in the western region. China’s high-speed railway has a business mileage of 29,000 km, covering 23 provinces. China has the world’s longest high-speed rail mileage and the highest transport density, and the total mileage of high- speed rail accounts for two-thirds of the world’s total high-speed rail mileage. Data source: https://www.sohu.com/a/285874760_656927. 276 Z. JIN, ET AL. 2.1.2. Institutional background of IPO On 1 February 2004, the CSRC issued the Interim Measures for the Sponsorship System for the Listing of Securities, which formally stipulated that the sponsor system of new shares issue began to be implemented, and the sponsor shall be responsible for recommending qualified companies. Specifically, the sponsor institution and its sponsor representative are responsible for recommending the proposed listed company to the CSRC and provid- ing pre-listing counselling to the company. To this end, the sponsor needs to do due diligence on the company applying for listing, make its corporate governance structure meet the requirements of the listing specification, and pre-verify the authenticity, accu- racy and completeness of the application materials. After the successful listing of the company, the sponsor and its sponsor representative still have the responsibility to conduct continuous supervision of the company it recommends for a period of time, and bear legal responsibility for the non-standard behaviour of the recommended com- pany throughout the period of coaching and supervision. Finally, the CSRC reviews the listing information of the applicant company endorsed by the sponsor to determine whether the company has met the listing conditions. Under the sponsor system, in order to guide the company to meet the requirements of the listing norms, the sponsor needs to conduct company site visits many times and, if necessary, stays in the proposed company for counselling and research for a long time. The recommendation system requires intermediaries such as brokers and auditors to play a greater role in stock issuance, and also increases their corresponding legal risks. Therefore, they also need to go to the company for information acquisition many times. As the final auditor, the CSRC will conduct many company site visits in the final stage to obtain more adequate or accurate information to determine whether the company is eligible for listing financing. Therefore, whether it is regulators, brokers, auditors, or primary market investors (private equity institutions and venture capital institutions) have the incentive to obtain more adequate and accurate information of the company. 2.2. Theory analysis Based on the above-mentioned institutional background, both the motivation of the company to apply for listing and the approval probability of company application are closely related to the intermediary, regulator and investors’ understanding of the company. Less information asymmetry between companies and regulators can help regulators judge whether a company can go public more accurately, reducing the risk of accountability on the regulators themselves. Therefore, if the regulator has a higher cost of obtaining information about a company, they will have a higher risk perception of the company and then deny the IPO approval. From the broker’s point of view, the more adequate and accurate information the broker has obtained from the company, the better they can guide the company to meet the requirements of listing norms, so are other intermediaries. From the company’s perspective, the lower the cost of obtaining its information is, the lower the cost of applying for listing, and the higher the approval probability of application is, prompting the company to be more willing to submit a listing application. Therefore, if the first problem for the company to go public to finance is the information asymmetry between the company and the external stakeholders, which can be solved by reducing the cost of information users to obtain information. CHINA JOURNAL OF ACCOUNTING STUDIES 277 From the information provider’s perspective, companies can reduce information asymmetry by increasing public disclosure. From the point of view of information needers, external stakeholders can obtain private information from the company through information search and site visits. According to research by Ball et al. (2003), Fan and Wong (2002), and Wong (2016), in emerging markets in East Asia, the quality of public information disclosure by listed companies is generally low, and external stake- holders usually obtain company information through private communication. Meanwhile, companies that apply for listing disclose far less public information than listed companies. Moreover, the applicant company has a stronger incentive to manip- ulate earnings to meet listing requirements (Chen & Yuan, 2004; Lin & Wei, 2000), so it is more necessary to obtain information through private communication. Private commu- nication can be made by voice or video through communication tools, as well as by face- to-face site visits. Cheng et al. (2019) and Han et al. (2018) show that site visits are more effective than public disclosure channels and telephone or video communication in obtaining private information, because ‘hard information’ such as financial indicators can be obtained through public information disclosure and telephone communication. But it is difficult to obtain ‘soft information’ such as corporate culture, human capital and business environment (Petersen & Rajan, 2002). The judgement of company value depends on both hard information and soft information. Site visits are the primary way to obtain soft information about the company (Long et al., 2017; Stein, 2002), and by which the authenticity of hard informa- tion can be confirmed. Chen and Yuan (2004), Lin and Wei (2000) find that the fraud of hard information such as financial data in the process of applying for listing by Chinese companies makes it particularly important for external stakeholders to obtain soft information through site visits. The efficiency of site visits depends on the facilitation of transport infrastructure. The opening of high-speed rail in the city where a company is located, by compressing the space-time distance, can improve the efficiency of company site visits by external stakeholders, reduce the cost of company information acquisition, and finally reduce the information asymmetry between the company and external stakeholders. The convenience brought about by the opening of the high- speed rail can not only enable a wider range of intermediaries such as sponsors, audit firms and law firms to provide services to the company, but also make regulators more willing to understand the company, which enhance the company’s motivation to submit a listing application. Although understanding does not mean the company can neces- sarily be recognised, understanding is a precondition. Because the opening of high- speed rail makes it easier for companies to get pre-IPO guidance from intermediaries, and better understanding by regulators, caeteris paribus, the opening of high-speed rail can increase the approval probability of company application. The above analysis leads to the following hypothesis: Hypothesis 1. Caeteris paribus, after a city launches a HSR route, the number of compa- nies applying for listing will increase significantly in the city and the approval probability of company application will also increase significantly. Soft information refers to the unique information that is difficult to be recorded, stored or transmitted. It is a kind of non- standard information and ‘oral information’ (Liu & Zhu, 2015; Petersen & Rajan, 2002). 278 Z. JIN, ET AL. 3. Research design 3.1. Empirical model According to the Medium- and Long-Term Railway Network Plan, China’s railway autho- rities have planned a high-speed rail network across the country. China’s high-speed rail construction presents a regional look: Not every region has opened high-speed rail, and the opening time of high-speed rail in each region is not the same, which provides us an opportunity to rely on the launch of high-speed railway (HSR) service in China as a natural experiment to examine how geographical distance affects firms’ IPO activity. Following Atanassov (2013), Bertrand et al. (2004) and Imbens and Wooldridge (2009), we estimate a differences-in-differences regression model: IPONUM ¼ β þ β POST þ β GDP þ β GDP PER þ β PEOPLE i;tþ1 i;t i;t i;t i;t 0 1 2 3 4 X X (1) þ β AIRPORT þ β GENTI þ CITY þ YEARþ ε i;t i;t i;tþ1 5 6 where i indexes cities, t indexes years, the dependent variable IPONUM equals the i,t+1 number of firms applying for IPOs in year t + 1 in city i, Post is a dummy variable that i,t equals one if the HSR service is available by year t in city i, control variables include GDP, GDP_PER, People and so on, and more detailed variable definitions are in Table 1, and ε i,t+1 is an error term. This methodology can better control for the fixed differences between the treated group and the control group via the city-fixed effects. The year dummies control for aggregate fluctuations. Next, this paper will examine how the opening of high-speed rail affects the company’s listing application. Following Chan et al. (2012), Chu and Fang (2016), we estimate the following differences-in-differences Logit regression model: PASS ¼ β þ β TRAIN þ β POST þ β SIZE þ β LEV þ β WXRATIO i;tþ1 i t i;t i;t i;t 0 1 2 3 4 5 þ β TATR þ β M GROWTH þ β MEANROE þ β MEANEI i;t i;t i;t i;t 6 7 8 9 (2) þ β MEANCASH þ β BIG4 þ β RANK þ β OTHER þ β SOE i;t i;t i;t i;t i;t 10 11 12 13 14 X X þ β AIRPORT þ β G GROWTH þ INDUSTRY þ YEARþ ε i;t i;t i;tþ1 15 16 where i indexes firms, t indexes years, the dependent variable PASS is a dummy variable i,t that equals one if IPO application of firm i is approved by CSRC in year t + 1, POST is i,t a dummy variable that equals one for firm i in year t after the launch of an HSR route to a city where firm i is located, TRAIN is a dummy variable that equals one for firm i in any year if an HSR route is launched in a city where firm i is located, and ε is an error i,t+1 term. Following prior literature (Dai et al., 2014; Li & Liu, 2012; Zeng et al., 2016), we control some variables of firm characteristics, such as firm size (SIZE), financial leverage (LEV), intangible assets ratio (WXRATIO) and so on, and more detailed variable definitions are in Table 1. Industry and year fixed effects are also controlled in the regression model (2) by industry and year dummies. 3.2. Data and sample We obtain data on HSR sites from China Railway website (www.china-railway.com.cn), and manually collect the earliest date when an HSR route was launched in each city. CHINA JOURNAL OF ACCOUNTING STUDIES 279 Table 1. Variable definitions. Variables Definition IPONUM The number of firms applying for IPOs. LNIPONUM The natural logarithm of one plus IPONUM. PASS A dummy variable that equals one if IPO application of a firm is approved by CSRC, and 0 otherwise. TRAIN A dummy variable that equals one for a firm in any year if a HSR route is launched in the city where it is located, and 0 otherwise. POST A dummy variable that equals one after a HSR route is launched in a city, and 0 otherwise. POST2 A dummy variable that equals one for a firm after the HSR opening of the second year in the city where it is located, and 0 otherwise. POST3 A dummy variable that equals one for a firm after the HSR opening of the third year in the city where it is located, and 0 otherwise. ROA Net income divided by total assets at the end of year. GROWTH One-year percentage growth in sales. UNDERPRICING (first-day closing price of post-IPO – IPO offer price)/IPO offer price. VOLATILITY Standard deviation on the daily return of stocks in the first year following firm’s IPO. VC A dummy variable that equals one for a firm if at least a venture capital (VC) is one of the top ten shareholders when the firm applies for listing and 0 otherwise. RELAT A dummy variable that equals one for a firm if at least one employee of underwriters or audit firms have ever served as a member of the issuance examination committee, and 0 otherwise. 0 1 DROA /DROA ROA in the year prior to the IPO minus ROA in the IPO year or in the first year after the IPO and a larger value means a larger decline in profitability. BH12 Market-adjusted 12 months stock return: � ðmonthly adjusted return þ 1Þ 1, where the i¼1 i monthly adjusted return equals the raw return minus the monthly market return both in month i after the IPO month. BH24 Market-adjusted 24 months stock return: � ðmonthly adjusted return þ 1Þ 1, where the i¼1 i monthly adjusted return equals the raw return minus the monthly market return both in month i after the IPO month. GDP The natural logarithm of GDP in the city. GDP_PER The natural logarithm of GDP per capita in the city. G_GROWTH One-year percentage growth in GDP of the city. PEOPLE The natural logarithm of population in the city. GENGTI A dummy variable that equals one if a city changes its Secretary of CPC Municipal Committee or mayor in the next year, and 0 otherwise. SIZE Natural logarithm of total assets at the end of year. LEV Total liabilities divided by total assets at the end of year. OTHER Other receivables divided by total assets at the end of year. CR Current liabilities divided by current assets at the end of year. WXRATIO Intangible assets divided by total assets at the end of year. TATR Operating revenue divided by total assets at the end of year. EI Non-recurring profit and loss divided by net income. MFEE The administrative expenses divided by annual operating revenue. CASH Net cash flow from operations divided by total assets. HDINDEX The sum of market share squared in the industry in which the firm operates (Herfindahl index). MEANEI Three-year average EI in three years prior to IPO, where EI is defined as non-recurring profit and loss divided by net income. MEANCASH Three-year average CASH in three years prior to IPO, where CASH is defined as net cash flow from operations divided by total assets. MEANROE Three-year average ROE in three years prior to IPO, where ROE is defined as net income divided by net asset. M_GROWTH Three-year average GROWTH in three years prior to IPO. BIG4 A dummy variable that equals one if the company hires a ‘big four’ audit firm when it applies for listing, and 0 otherwise. RANK A dummy variable that equals one if the company’s underwriter ranks in the top 10 among brokers when it applies for listing, and 0 otherwise. The rank is calculated based on the sum of the underwriting amount of the lead underwriters in the listing year and the last year. SOE A dummy variable that equals one if a firm is state-owned, and 0 otherwise. EST_AGE The natural logarithm of one plus difference between the current year and the year the company is founded. AIRPORT A dummy variable that equals one after the introduction of an airline route, and 0 otherwise. CDRETWDOS Market Return on the first day of company listing (weighted by current market value) BM The book value of total assets divided by the market value of total assets. 280 Z. JIN, ET AL. We manually collect the data on listing declaration time, property and VC participation from the prospectus of the company. We also manually judge whether the intermedi- ary has ever been in the position of the Commission by checking the list of past issuing committee members published on the CSRC website. We obtain company’s listing application data and financial data from the Wind database, the official turnover data from the CNRDS database, city characteristics data from the China Economic And Social Statistics database. The sample period is from 2004 to 2017. For Model (1), we add up the number of companies applying for listing of each city in each year and finally obtain city-year observations of 2723, covering 205 cities, of which 145 cities have at least one HSR route. We remove the observations of the current year when an HSR route is launched in each city. For Model (2), we delete listed firms in the financial industry and firms with missing data. We also remove the observations in the year when an HSR route is launched in each city. Finally, we obtain 1905 firm observations, of which mainboard, small and medium board, and GEM account for 27.72%, 37.64% and 34.65%, respectively. Table 2, Panel A presents the yearly observation distribution of Model (1). Panel B of Table 2 presents the province observation distribution of Model (1). Table 2, Panel C presents the yearly observation distribution of Model (2). Panel D of Table 2 presents the industry observation distribution of Model (2). 3.3. Descriptive statistics Panel A of Table 3 shows descriptive statistics of the main variables in model (1) above. The average value of IPONUM is 0.866, the maximum value is 46 and the minimum value is 0, indicating that the number of listing applications varies greatly among cities. The average and median of TRAIN are 0.692 and 1, respectively, indicating that more than half of the cities have at least an HSR route, and China’s high-speed rail develops rapidly. The mean and median of POST is 0.256 and 0, respectively, indicating that 25.6% of the samples are in the period following the opening of the high-speed rail. The mean and median value of GDP_PER is 10.571 and 10.6, it is 38987 yuan and 40,134 yuan, respectively. The average and median of AIRPORT are 0.45 and 0, indicating less than half of the cities have an airport. The average and median of GENTI are 0.54 and 1, respectively, indicating that 54% of the cities in the sample period experience the turnover of city party secretaries or mayors. Panel B of Table 3 shows descriptive statistics of the main variables in the model (2) above. The mean and median of PASS is 0.909 and 1, respectively, indicating the approval probability of listing application reaches 90.9%. The average value of POST is 0.674, indicating that 67.4% of IPOs are distributed in the period following the opening of the high-speed rail. 3.4. Correlation analysis Table 4, Panel A is an analysis of the correlations of the main variables of the model (1). Most of the correlation coefficients between the variables are below 0.4, indicating If no company applies for listing in a certain year in a city, the value of IPONUM is 0. CHINA JOURNAL OF ACCOUNTING STUDIES 281 Table 2. Sample distribution. City-level Data for Model (1) Panel A: Distribution by year 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Total Freq. 204 205 205 205 196 188 186 190 188 189 179 190 198 200 2723 Per. (%) 7.49 7.53 7.53 7.53 7.20 6.90 6.83 6.98 6.90 6.94 6.57 6.98 7.27 7.34 100 Panel B: Distribution by province Freq.Per. (%) Shandong 228 8.37 Guangdong 211 7.75 Jiangsu 173 6.35 Zhejiang 144 5.29 Sichuan 176 6.46 Anhui 169 6.21 Henan 161 5.91 Hubei 147 5.40 Hunan 144 5.29 Hebei 131 4.81 Fujian 118 4.33 Liaoning 106 3.89 Jiangxi 104 3.82 Shanxi 82 3.01 Inner Mongolia 69 2.53 Jilin 68 2.50 Heilongjiang 67 2.46 Gansu 67 2.46 Shaanxi 66 2.42 Yunnan 54 1.98 Guangxi 52 1.91 Guizhou 40 1.47 Ningxia 28 1.03 Xinjiang 27 0.99 Hainan 26 0.95 Shanghai 13 0.48 Beijing 13 0.48 Tianjin 13 0.48 Chongqing 13 0.48 Qinghai 13 0.48 Total 2,723 100.00 Firm-level Data for Model (2) Panel C: Distribution by year 2004 2006 2007 2008 2009 2010 2011 2012 2014 2015 2016 2017 Total Freq. 24 47 96 77 130 224 206 145 95 232 232 397 1905 Per. (%) 1.26 2.47 5.04 4.04 6.82 11.76 10.81 7.61 4.99 12.18 12.18 20.84 100.00 Panel D: Distribution by industry Freq.Per. (%) Mining 29 1.52 Production and supply of 24 1.26 electricity, steam, and tap water Real estate 7 0.37 Construction 56 2.94 Transportation and warehousing 37 1.94 Agriculture, forestry, husbandry, 20 1.05 and fishery Wholesale and retail 61 3.2 Social services 108 5.67 Culture, sports, and entertainment 32 1.68 (Continued) 282 Z. JIN, ET AL. Table 2. (Continued). Information technology 180 9.45 Electronics 216 11.34 Textile, garment manufacturing, 53 2.78 and products of leather and fur Machinery, equipment, and 494 25.93 instrument manufacturing Metal and non-metal 127 6.67 Wood and furniture 22 1.15 Petroleum, chemical, plastics, and 199 10.45 rubber products Food and beverage 62 3.25 Medicine and biological products 128 6.72 manufacturing Papermaking and printing 35 1.84 Other manufacturing 15 0.79 Total 1905 100.00 Table 3. Descriptive statistics. VARIABLES N MEAN SD MIN P25 P50 P75 MAX Panel A: City-level Data for Model (1) IPONUM 2723 0.866 2.978 0.000 0.000 0.000 1.000 46.000 TRAIN 2723 0.692 0.462 0.000 0.000 1.000 1.000 1.000 POST 2723 0.256 0.436 0.000 0.000 0.000 1.000 1.000 GDP 2723 24.495 1.259 21.800 23.600 24.400 25.300 27.400 GDP_PER 2723 10.571 0.812 8.550 10.000 10.600 11.100 12.200 PEOPLE 2723 13.924 0.754 12.400 13.400 13.900 14.400 16.000 AIRPORT 2723 0.450 0.498 0.000 0.000 0.000 1.000 1.000 GENGTI 2723 0.540 0.498 0.000 0.000 1.000 1.000 1.000 Panel B: Firm-level Data for Model (2) PASS 1905 0.909 0.287 0.000 1.000 1.000 1.000 1.000 POST 1905 0.674 0.469 0.000 0.000 1.000 1.000 1.000 SIZE 1905 20.327 1.032 18.600 19.600 20.100 20.800 24.500 LEV 1905 0.424 0.183 0.009 0.299 0.433 0.557 0.825 WXRATIO 1905 0.945 0.452 0.208 0.655 0.853 1.120 2.910 TATR 1905 0.236 0.096 0.066 0.170 0.220 0.285 0.566 M_GROWTH 1905 0.091 0.109 −0.059 0.020 0.056 0.124 0.538 MEANROE 1905 17.830 1.135 15.400 17.100 17.700 18.400 21.800 MEANEI 1905 0.171 0.122 −0.087 0.084 0.160 0.245 0.512 MEANCASH 1905 0.121 0.077 −0.028 0.068 0.113 0.163 0.371 BIG4 1905 0.035 0.184 0.000 0.000 0.000 0.000 1.000 RANK 1905 0.420 0.494 0.000 0.000 0.000 1.000 1.000 OTHER 1905 0.015 0.017 0.000 0.004 0.009 0.018 0.103 SOE 1905 0.137 0.344 0.000 0.000 0.000 0.000 1.000 AIRPORT 1905 0.783 0.412 0.000 1.000 1.000 1.000 1.000 G_GROWTH 1905 0.113 0.053 −0.017 0.082 0.107 0.143 0.309 that model (1) does not have a serious problem of multiple collinearity. TRAIN and POST are both significantly and positively correlated with IPONUM, preliminary indi- cating that the opening of high-speed rail has a positive impact on the number of IPO applications. Table 4, Panel B is an analysis of the correlations of the main variables of the model (2). Most of the correlation coefficients between the variables are below 0.3, indicating that the model (2) does not have a serious problem of multiple collinea- rities. PASS is not significantly and positively correlated with POST, which may be due CHINA JOURNAL OF ACCOUNTING STUDIES 283 Table 4. Correlation coefficient matrix. Panel A: City-level variables for Model (1) VARIABLES 1. 2. 3. 4. 5. 6. 7. 8. 1.IPONUM 1.000 2.TRAIN 0.159*** 1.000 3.POST 0.261*** 0.391*** 1.000 4.GDP 0.417*** 0.285*** 0.505*** 1.000 5.GDP_PER 0.316*** 0.172*** 0.490*** 0.821*** 1.000 6.PEOPLE 0.376*** 0.294*** 0.326*** 0.785*** 0.296*** 1.000 7.AIRPORT 0.187*** 0.075*** 0.151*** 0.433*** 0.351*** 0.346*** 1.000 8.GENGTI −0.014 −0.023 0.126*** 0.030 0.095*** −0.054*** 0.031 1.000 Panel B: Firm-level variables for Model (2) VARIABLES 1. 2. 3. 4. 5. 6. 7. 8. 9. 1.PASS 1 2.TRAIN 0.043* 1 3.POST 0.006 0.349*** 1 4.SIZE 0.090*** −0.042* 0.077*** 1 5.LEV 0.308*** −0.006 −0.186*** 0.437*** 1 6.TATR 0.035 0.020 −0.097*** −0.024 0.151*** 1 7.MEANROE 0.180*** 0.024 −0.115*** −0.230*** −0.004 0.236*** 1 8.MEANEI −0.048** 0.026 0.172*** −0.027 −0.047** −0.164*** −0.133*** 1 9.MEANCASH 0.026 0.006 0.031 −0.183*** −0.352*** 0.071*** 0.503*** −0.067*** 1 10.M_GROWTH 0.032 0.004 −0.246*** −0.157*** 0.100*** 0.124*** 0.395*** −0.042* 0.051** 11.WXRATIO 0.032 −0.043* −0.003 0.021 −0.050** −0.077*** −0.042* 0.001 0.096*** 12.BIG4 −0.009 0.021 −0.019 0.302*** 0.049** −0.015 −0.079*** −0.039* −0.040* 13.RANK 0.032 0.040* 0.027 0.079*** −0.001 0.020 0.042* 0.006 0.017 14.OTHER −0.010 0.042* −0.047** 0.005 0.118*** 0.113*** 0.014 0.052** −0.084*** 15.SOE 0.030 −0.010 −0.130*** 0.341*** 0.125*** −0.106*** −0.132*** −0.019 −0.020 16.AIRPORT −0.011 0.245*** 0.246*** 0.015 −0.064*** −0.057** 0.020 0.101*** 0.058** 17.G_GROWTH −0.008 −0.015 −0.273*** −0.043* 0.108*** 0.062*** 0.120*** −0.065*** −0.012 VARIABLES 10. 11. 12. 13. 14. 15. 16. 17. 10.M_GROWTH 1 11.WXRATIO −0.099*** 1 12.BIG4 −0.024 −0.055** 1 13.RANK 0.086*** −0.034 0.080*** 1 14.OTHER 0.088*** −0.076*** 0.047** −0.003 1 15.SOE −0.040* 0.033 0.156*** −0.033 0.048** 1 16.AIRPORT 0.014 −0.059** 0.011 0.056** 0.091*** 0.061*** 1 17.G_GROWTH 0.266*** −0.027 0.060*** −0.020 0.082*** 0.107*** 0.008 1 ***, *denote statistically significant at 1% and 10% levels, respectively. 284 Z. JIN, ET AL. to not taking other factors into consideration. We will conduct further multiple regression analysis. 4. Empirical results 4.1. The impact of HSR on firm’s IPO application Table 5 reports how the launch of an HSR route affects firm’s IPO application. Since viable IPONUM is a discrete variable of non-negative integers, we estimate a Poisson regression of Model (1) in Column 1 of Table 5, and we find the coefficient on POST is 0.265 and significant at the 5% level. We also specify LNIPONUM as the dependent variable and estimate an OLS regression of Model (1) in Column 2 of Table 5 and the coefficient on POST is also positive and significant. These two regression results indicate that HSR can enhance the motivation of companies applying for listing. Following DeFond et al. (2015), Zhong and Lu (2018), we use the propensity score matching (PSM) approach to construct the paired sample and then estimate a Poisson regression in Column 3 of Table 5, finally obtaining the same result. Considering that it may take some time for HSR to come into play, we also specify IPONUM in year t + 2(IPONUM ) as the dependent variable in Column 4 and 6 of Table 5 and specify t+2 LNIPONUM in year t + 2(LNIPONUM ) as the dependent variable in Column 5 of Table 5, t+2 and the empirical results are consistent with our hypothesis. Table 5. The impact of HSR on firm’s IPO application. (1) Poisson (2) OLS (3) PSM (4) Poisson (5) OLS (6) PSM Variables IPONUM LNIPONUM IPONUM IPONUM LNIPONUM IPONUM t+1 t+1 t+1 t+2 t+2 t+2 CONSTANT 1.295 −0.150 (0.84) (−0.09) POST 0.265** 0.186*** 0.590** 0.159* 0.144*** 0.543** (2.38) (6.04) (2.47) (1.83) (4.66) (2.23) GDP −0.027 −0.484 0.229 −0.098 −0.562* −0.882 (−0.12) (−1.50) (0.07) (−0.42) (−1.83) (−0.31) GDP_PER 0.204 0.460 −0.368 0.405 0.568* 0.886 (0.68) (1.33) (−0.12) (1.43) (1.75) (0.33) PEOPLE −0.061 0.442 −0.208 0.175 0.599* 1.249 (−0.20) (1.30) (−0.07) (0.54) (1.83) (0.44) AIRPORT 0.097 −0.011 −0.185 −0.158 0.116*** 0.082 (0.55) (−0.39) (−0.61) (−0.89) (2.97) (0.21) GENGTI 0.055 0.012 0.013 −0.008 −0.008 −0.172 (1.16) (0.78) (0.08) (−0.15) (−0.50) (−1.05) CITY Yes Yes Yes Yes Yes Yes YEAR Yes Yes Yes Yes Yes Yes N 2723 2723 1536 2523 2523 1424 Adj. R /Pseudo-like −1524.117 0.19 −455.192 −1360.099 0.20 −411.410 The numbers in parentheses are the t(z)-statistics. *, **, and *** represent significance at the 0.1, 0.05, and 0.01 levels, respectively. The standard errors are robust to heteroscedasticity and are clustered by city. Pseudo-like is the abbreviation of Log Pseudo likelihood. In the first stage regression, we specify POST as the dependent variable and all control variables in Model (1) as the covariates and estimate a logit regression. We also estimate a Zero-Inflated Poisson regression of Model (1), use the sample from three years before to three years after the opening of HSR to repeat the result of Column 1 of Table 5, and use Tobit model to repeat the result of Column 2 of Table 5, and our conclusion remains unchanged. These regression results are untabulated. CHINA JOURNAL OF ACCOUNTING STUDIES 285 4.2. The impact of HSR on firm’s IPO application: the validity of the parallel trends assumption In reality, the high-speed railway sites are almost always based on the original ordinary railway planning route design, so we believe that the high-speed rail station route planning is mainly affected by the geographical environment, rather than the degree of local economic development. However, in order to lend more support for our primary hypothesis, we continue to conduct parallel trend tests to further eliminate the possibi- lity that a better local economy leads to more IPO applications and the opening of a local high-speed rail. To rule out this possibility, following Bertrand and Mullainathan (2003), we further examine the dynamic impact of the opening of high-speed rail on IPO applications of companies. We estimate a Poisson regression of model (3): IPONUM ¼ β þ β BEFORE1 þ β AFTER1 þ β AFTER2 þ β GDP þ β GDP PER i;tþ1 t i;t i;t i;t i;t 0 1 2 3 4 5 X X þ β PEOPLE þ β AIRPORT þ β GENTI þ CITY þ YEARþ ε i;t i;t i;t i;tþ1 6 7 8 (3) where BEFORE1 is a dummy variable that equals one for a city that will open an HSR route a year later, AFTER1 is a dummy variable that equals one for a city that opened an HSR route last year, and AFTER2_ is a dummy variable that equals one for a city that opened an HSR route at least two years ago. A positive coefficient on BEFORE1 indicates that the local economic development leads to an increase in IPO applications. The result is in Column 1 of Table 6, and we find that the estimated coefficient on BEFORE1 is statistically insignif- icant. Consistent with a causal interpretation of our basic result, we find that the estimated coefficient on the AFTER1 dummy is economically smaller than the coefficient on the AFTER2_ dummy. In order to examine the dynamic impact of the opening of HSR over a longer period of time, we add variables BEFORE3 and BEFORE2 to model (3) and replace variable AFTER2_ with variables AFTER2, AFTER3_ for further analysis in column 2 of Table 6. We find only coefficients on AFTER1, AFTER2 and AFTER3_ are significantly positive, while coefficients on Before3, Before2 and Before1 are not significant. Table 6 shows that the treated group and the control group in Model (1) meet the parallel trend assumption. Meanwhile, these results also show that our conclusion is not driven by the level of local economic. In addition, we use instrumental variable method to conduct further analysis. Following Liu and Li (2017), this paper uses the total passenger volume of a city in a particular year (1989) to construct an instrumental variable of whether a city has opened a high-speed rail. Because the total number of passengers can reflect the local city’s passenger demand, which is an important consideration in the opening of the high-speed rail in a region. Therefore, the total number of passengers in 1989 is related to whether a region has opened a high-speed rail, but the number of present IPO applications of companies in one region will not affect the region’s total passenger volume in 1989. Therefore, the total passenger A very important premise of the differences-in-differences model is the parallel trend assumption, which requires that the treated group and the control group have the same characteristics and development trend in the case that the event does not occur. BEFORE3 is a dummy variable that equals one for a city that will open a HSR route three years later, BEFORE2 is a dummy variable that equals one for a city that will open a HSR route two years later, AFTER2 is a dummy variable that equals one for a city that opened a HSR route two years ago, and AFTER3_ is a dummy variable that equals one for a city that opened a HSR route at least three years ago. 286 Z. JIN, ET AL. Table 6. The impact of HSR on firm’s IPO application: the validity of the parallel trends assumption. (1) (2) VARIABLES IPONUM IPONUM t+1 t+1 BEFORE3 −0.075 (−0.65) BEFORE2 0.141 (1.26) BEFORE1 0.029 0.090 (0.29) (0.75) POST AFTER1 0.262* 0.358** (1.88) (2.13) AFTER2 0.294* (1.73) AFTER2_ 0.307** (2.06) AFTER3_ 0.533*** (2.66) GDP −0.027 −0.010 (−0.12) (−0.04) GDP_PER 0.205 0.223 (0.68) (0.75) PEOPLE −0.064 −0.070 (−0.20) (−0.23) AIRPORT 0.096 0.097 (0.54) (0.54) GENGTI 0.054 0.047 (1.12) (1.02) CITY Yes Yes YEAR Yes Yes N 2723 2723 Pseudo-like −1523.950 −1519.849 The numbers in parentheses are the t(z)-statistics. *, **, and *** represent sig- nificance at the 0.1, 0.05, and 0.01 levels, respectively. The standard errors are robust to heteroscedasticity and are clustered by city. Pseudo-like is the abbre- viation of Log Pseudo likelihood. volume in 1989 meets the two basic requirements of the instrumental variable: relevance and exogeneity. Since the total passenger volume in 1989 does not change over time, we add the interaction terms between the year dummies and passenger volume to the regression. In the untabulated regressions, we find that the coefficient on high-speed rail opening variable (POST) is still significantly positive, consistent with the conclusions above. 4.3. The impact of HSR on firm’s IPO application: first-tier cities and non-first-tier cities China is vast and its economic development is uneven. The stock exchange is located in Shanghai and Shenzhen, respectively, and the central government and regulatory agen- cies are located in Beijing. Brokers, audit firms and other intermediaries are mainly concentrated in Beijing, Shanghai, Guangzhou and Shenzhen. The three major city groups (Beijing-Tianjin-Hebei, Yangtze River Delta and Pearl River Delta) are the exporting Beijing-Tianjin-Hebei city group contains Beijing city, Tianjin city and Hebei province; Yangtze River Delta city group contains Shanghai city, Zhejiang province and Jiangsu province; Pearl River Delta contains Guangdong province. CHINA JOURNAL OF ACCOUNTING STUDIES 287 parties of listed resources, because the three major city groups are economically devel- oped, accounting for a larger proportion of China’s total economic output, and have a large number of brokers, audit firms and other intermediary institutions. Therefore, we expect the opening of high-speed rail to have a smaller impact on company’s IPO applications in these cities than in other cities. Column 1 and 2 of Table 7 report the results. We find that the coefficient on POST is significantly positive only in Column 2, indicating that the impact of the high-speed rail opening is more pronounced in cities except for the three large city groups above. Moreover, Beijing, Shanghai, Guangzhou, Shenzhen (big four cities) is the China’s four most developed financial cities, half of the brokers are headquartered here. Therefore, compared with companies in cities that do not have an HSR route directly to the four big cities, companies in cities that have a direct high-speed train to the above four big cities are more likely to hire well-known intermediaries, and IPO activities are more affected. BIGFOURPOST is a dummy variable that equals one if a city’s high-speed rail goes straight to the above big four cities, and 0 otherwise; NOBIGFOURPOST is a dummy variable that equals one if a city’s high-speed rail does not go straight to the above big four cities, and 0 otherwise. From the results in Column 3–5, we find that the coefficient on BIGFOURPOST is significantly positive, and the difference between BIGFOURPOST and NOBIGFOURPOST coefficient is significant, indicating that the number of IPO applications in cities that can go straight to the above big four cities increases more. Table 7. The impact of HSR on firm’s IPO application: first-tier cities and non-first-tier cities. (1) Three large city groups (2) Other cities (3) (4) (5) VARIABLES IPONUM IPONUM IPONUM IPONUM IPONUM t+1 t+1 t+1 t+1 t+1 POST 0.171 0.377*** (1.06) (3.11) BIGFOURPOST 0.397*** 0.409*** (3.25) (3.15) NOBIGFOURPOST −0.176 0.031 (−1.33) (0.22) GDP 0.242 0.136 0.008 0.120 0.006 (1.05) (0.22) (0.02) (0.22) (0.01) GDP_PER −0.086 0.711 0.408 0.288 0.411 (−0.36) (1.12) (0.78) (0.53) (0.79) PEOPLE −0.412 −0.106 −0.200 −0.244 −0.196 (−1.28) (−0.16) (−0.35) (−0.41) (−0.35) AIRPORT 0.037 0.216 −0.015 −0.003 −0.016 (0.14) (0.83) (−0.09) (−0.02) (−0.10) GENGTI 0.036 0.124 0.076 0.067 0.076 (0.64) (1.52) (1.34) (1.20) (1.33) CITY Yes Yes Yes Yes Yes YEAR Yes Yes Yes Yes Yes N 698 2025 2671 2671 2671 Pseudo-like −612.068 −893.827 −1403.797 −1410.185 −1403.769 The numbers in parentheses are the t(z)-statistics. *, **, and *** represent significance at the 0.1, 0.05, and 0.01 levels, respectively. The standard errors are robust to heteroscedasticity and are clustered by city. Pseudo-like is the abbreviation of Log Pseudo likelihood. From 2006 to 2015, the Yangtze River Delta, the Pearl River Delta and the Beijing-Tianjin-Hebei three major city groups account for more than 40% of GDP in China. Data source: http://www.xinhuanet.com/fortune/2019-03/20/c_ 1124256239.htm 288 Z. JIN, ET AL. 4.4. The impact of HSR on firm’s IPO application: the alternative of airline The airline is an alternative to high-speed rail. Compared with companies in cities that have both high-speed rail and the airport, companies in cities that do not have an airport are more affected. Column 1 and 2 of Table 8 report the results. We find the coefficient on POST is significantly positive only in Column 2 but not significant in Column 1, indicating that the impact of the opening of the high-speed rail exists mainly in the city without airports. 4.5. The impact of HSR on the approval probability of firm’s IPO application We have confirmed that the opening of high-speed rail has contributed to a significant increase in the number of IPO applications. Next, we examine how the opening of high- speed rail affects the approval probability of company’s listing application, and Table 9 reports the results. PASS is a dummy variable that equals one if the company’s listing application is approved, and 0 otherwise. We specify PASS as the dependent variable and estimate a logit regression. We find that the coefficient on POST is significantly positive. At the same time, this paper uses POST2 and PSOT3 to replace POST for the robustness test, and the results remain the same. The results in Table 9 consistently show that the opening of high-speed rail significantly increases the approval probability of company’s listing application. 4.6. The impact of HSR on the approval probability of firm’s IPO application: mechanism analysis This paper analyzes how the opening of high-speed rail affects the approval probability of a company’s listing application from three perspectives: the social capital of intermediaries, Table 8. The impact of HSR on firm’s IPO application: the alternative of airline. (1) With airports (2) Without airports VARIABLES IPONUM IPONUM t+1 t+1 POST 0.170 0.386** (1.38) (2.45) GDP −0.047 1.979** (−0.16) (2.33) GDP_PER 0.437 −1.751** (1.02) (−2.57) PEOPLE −0.061 −1.829** (−0.12) (−2.21) GENGTI 0.131** 0.008 (2.31) (0.10) CITY Yes Yes YEAR Yes Yes N 1225 1498 Pseudo-like −868.735 −528.443 The numbers in parentheses are the t(z)-statistics. *, **, and *** represent sig- nificance at the 0.1, 0.05, and 0.01 levels, respectively. The standard errors are robust to heteroscedasticity and are clustered by city. Pseudo-like is the abbre- viation of Log Pseudo likelihood. CHINA JOURNAL OF ACCOUNTING STUDIES 289 Table 9. The impact of HSR on the approval probability of firm’s IPO application. (1) (2) (3) VARIABLES PASS PASS PASS t+1 t+1 t+1 INTERCEPT −8.919*** −8.632*** −8.805*** (−2.83) (−2.73) (−2.77) POST 0.787** (2.32) POST2 0.721** (2.48) POST3 0.756*** (2.92) TRAIN 0.098 0.198 0.277 (0.20) (0.43) (0.62) SIZE 0.272* 0.252 0.259* (1.76) (1.61) (1.65) LEV 6.412*** 6.451*** 6.419*** (7.23) (7.23) (7.20) TATR −0.402 −0.401 −0.360 (−1.22) (−1.20) (−1.07) MEANROE 12.159*** 12.212*** 12.101*** (3.76) (3.78) (3.72) MEANEI −1.004 −1.105* −1.089* (−1.59) (−1.79) (−1.77) MEANCASH −1.267 −1.089 −1.011 (−0.57) (−0.49) (−0.45) M_GROWTH −1.984** −1.954** −1.878* (−1.99) (−1.97) (−1.82) WXRATIO 4.693** 4.654** 4.848** (2.04) (2.06) (2.12) BIG4 −0.164 −0.121 −0.109 (−0.29) (−0.22) (−0.19) RANK 0.349* 0.345* 0.314 (1.76) (1.72) (1.52) OTHER −1.954 −1.107 −1.056 (−0.37) (−0.21) (−0.20) SOE 0.097 0.082 0.077 (0.34) (0.29) (0.26) AIRPORT −0.154 −0.113 −0.121 (−0.57) (−0.41) (−0.44) G_GROWTH 0.838 0.548 0.605 (0.37) (0.24) (0.27) IND Yes Yes Yes YEAR Yes Yes Yes N 1905 1905 1905 Pseudo R 0.30 0.30 0.30 The numbers in parentheses are the t(z)-statistics. *, **, and *** represent sig- nificance at the 0.1, 0.05, and 0.01 levels, respectively. The standard errors are robust to heteroscedasticity and are clustered by firm. venture capital participation and the pre-IPO performance. Prior literature shows that the company hiring ‘a related intermediary’ that has ever served as a member of issuance examination committee can improve the approval probability of IPO application (Chen et al., 2014; Du et al., 2013; Li & Liu, 2012). Prior research has also confirmed that venture capital participation can improve the approval probability of listing application (Cai et al., 2013; Shen et al., 2013; Zeng et al., 2016). First, we examine how the opening of high-speed rail affects the decision of hiring a related intermediary, and Column 1 of Table 10 reports the result. We find the coefficient of POST is significantly positive, indicating that after the opening of high-speed rail, the company is more likely to hire a related intermediary when it 290 Z. JIN, ET AL. Table 10. The impact of HSR on the approval probability of firm’s IPO application: related intermediary and venture capital. (1) (2) VARIABLES RELAT VC INTERCEPT 1.729 3.712** (1.25) (2.13) POST 0.353* 0.329** (1.76) (2.09) TRAIN −0.172 −0.157 (−0.58) (−0.62) SIZE −0.044 −0.162* (−0.64) (−1.86) LEV 0.208 0.960*** (0.50) (2.76) WXRATIO −0.013 −0.355*** (−0.12) (−2.84) TATR −0.272 −1.358* (−0.34) (−1.66) M_GROWTH 0.782 −0.314 (1.58) (−0.68) MEANROE 0.818 −1.173 (1.08) (−1.48) MEANEI −0.719 3.240*** (−1.45) (6.38) MEANCASH 0.927 −0.715 (0.60) (−0.51) BIG4 1.345*** 0.116 (4.11) (0.38) RANK 0.624*** 0.027 (4.29) (0.25) OTHER 1.226 −1.028 (0.41) (−0.41) SOE 0.194 −0.459** (1.22) (−2.31) AIRPORT 0.002 −0.214 (0.01) (−1.40) G_GROWTH 0.186 0.996 (0.16) (0.88) IND Yes Yes YEAR Yes Yes N 1905 1905 Pseudo R 0.10 0.10 The numbers in parentheses are the t(z)-statistics. *, **, and *** represent sig- nificance at the 0.1, 0.05, and 0.01 levels, respectively. The standard errors are robust to heteroscedasticity and are clustered by firm. applies for listing. Next, we also examine whether a company is more likely to obtain venture capital investment after the opening of high-speed rail in Column 2 of Table 10, and the result shows the coefficient on POST is significantly positive, which indicates that the opening of high-speed rail significantly increases the probability of the company applying for listing to obtain a venture capital equity. There is also a possibility that regulators, intermediaries and investors will have lower costs of company information acquisition after the opening of high-speed rail, allowing them to pick out better companies. Therefore, we examine whether the pre-IPO perfor- mance of company applying for listing is better after the opening of high-speed rail. We specify ROA and GROWTH as the dependent variable in Column 1 and 2 of Table 11, respectively, and Table 11 reports the regression results. We find that the coefficients on CHINA JOURNAL OF ACCOUNTING STUDIES 291 Table 11. The impact of HSR on the approval probability of firm’s IPO application: the pre-IPO performance. (1) (2) VARIABLES ROA GROWTH INTERCEPT 0.338*** 0.544*** (10.77) (6.16) POST 0.006** 0.017* (2.51) (1.78) TRAIN −0.001 −0.007 (−0.31) (−0.56) SIZE −0.010*** −0.022*** (−7.18) (−5.84) LEV −0.132*** 0.142*** (−21.54) (8.83) WXRATIO −0.067*** −0.234*** (−3.23) (−3.38) TATR 0.018*** 0.008 (8.10) (1.18) EI −0.033*** −0.078*** (−4.37) (−3.38) CASH 0.328*** 0.202*** (34.80) (6.01) OTHER −0.007 0.263*** (−0.23) (3.07) SOE −0.005** −0.010 (−2.31) (−1.56) AIRPORT 0.004* 0.008 (1.73) (0.95) G_GROWTH 0.006 0.054 (0.35) (0.95) IND Yes Yes YEAR Yes Yes N 5686 4955 Adj. R 0.57 0.20 The numbers in parentheses are the t(z)-statistics. *, **, and *** represent sig- nificance at the 0.1, 0.05, and 0.01 levels, respectively. The standard errors are robust to heteroscedasticity and are clustered by firm. Observations in Table 11 includes firms’ three years of pre-IPO data. POST are significantly positive, indicating that the pre-IPO performance of company is better after the opening of the high-speed rail, thus making it easier for companies to pass through the review of IPO applications. 4.7. The impact of HSR on pricing efficiency of capital markets The empirical tests above find that the opening of high-speed rail has increased the motivation for companies to go public and increased the approval probability of compa- nies’ IPO applications. Besides, our empirical analysis also shows that the opening of high- speed rail reduces the cost of information acquisition of regulators, intermediaries and investors, allowing them to pick out better-quality applicants. Further, we examine whether the selected companies have better market performance. We specify variable UNDERPRICING and VOLATILITY as the dependent variable, respectively, in Table 12. The issuance pricing in China is regulated by the government, and the policy varies widely from time to time. Following Zhu and Qian (2010), this paper divides the sample into the following five parts according to the different stock issuance pricing policies, and sets four dummy variables: 2001–2004 (GUIDEYEAR0), the stock issue is priced at a fixed P/E ratio; 292 Z. JIN, ET AL. Table 12. The impact of HSR on the IPO underpricing and volatility of stock returns. (1) (2) VARIABLES UNDERPRICING VOLATILITY INTERCEPT 2.013*** 0.060*** (8.04) (7.76) POST −0.070* −0.002*** (−1.89) (−3.05) TRAIN 0.014 −0.001 (0.22) (−0.84) SIZE −0.076*** −0.001*** (−5.86) (−2.81) LEV 0.020 −0.000 (0.29) (−0.07) TATR 0.010 −0.000 (0.41) (−0.48) MEANROE −0.589*** −0.005 (−4.47) (−1.50) MEANEI 0.009 0.003 (0.12) (1.62) MEANCASH −0.088 −0.009*** (−0.44) (−2.99) M_GROWTH −0.032 0.003 (−0.35) (1.33) WXRATIO −0.002 0.010** (−0.01) (2.07) BIG4 −0.040 0.001 (−0.81) (0.49) RANK −0.001 −0.001** (−0.06) (−2.43) OTHER 0.841 0.022 (1.06) (1.57) CDRETWDOS 1.630* 0.011 (1.86) (0.65) SOE 0.037 0.000 (1.04) (0.54) AIRPORT 0.047** 0.000 (2.27) (0.61) G_GROWTH −0.149 0.002 (−0.48) (0.31) IND Yes Yes GUIDEYEAR Yes Yes N 1727 1708 Adj. R 0.45 0.38 The numbers in parentheses are the t(z)-statistics. *, **, and *** represent sig- nificance at the 0.1, 0.05, and 0.01 levels, respectively. The standard errors are robust to heteroscedasticity and are clustered by firm. January 2005 – June 2009 (GUIDEYEAR1), stock issue price implements the floating price- earnings ratio method within a certain range; July 2009–April 2012, the pricing of stock issues are determined by the market; May 2012-January 2014 (GUIDEYEAR2), the price of the issue of shares refers to the same industry price-earnings ratio, but when it exceeds the same industry price-earnings ratio of 25%, the company must advertise or re-inquire; February 2014–December 2017 (GUIDEYEAR3), the price of the stock issue refers to the same industry, but the price-earnings ratio limit is 23 times actually. We control the four dummy variables in the regression and find that the coefficients on POST are significantly negative, indicating that the opening of high-speed rail reduces the IPO underpricing and CHINA JOURNAL OF ACCOUNTING STUDIES 293 volatility of stock returns. These results show that the opening of high-speed rail reduces information asymmetry, reduces the financing cost of companies applying for listing, and improves the pricing efficiency of capital markets. 4.8. The impact of HSR on the post-IPO performance If the opening of high-speed rail optimises the allocation of resources, then the post-IPO performance of company listing successfully after the opening of the high-speed rail should be better. We examine this issue from the perspective of the company’s post-IPO accounting performance and market returns. Following Yang (2013), we specify DROA0 and DROA1 as the dependent variables, respectively, and examine the impact of the opening of high-speed rail on post-IPO accounting performance. Smaller values of DROA0 and DROA1 mean better company’s post-IPO accounting performance. The results are shown in Column 1 and 2 of Table 13. In Column 3 and 4 of Table 13, we also specify Table 13. The impact of HSR on the post-IPO performance: accounting performance and market performance. (1) Full Sample (2) Full Sample (3) Full Sample (4) Full Sample 0 1 VARIABLES DROA DROA BH12 BH24 t+1 t+1 t+1 t+1 INTERCEPT −0.104*** −0.063 2.358** 2.389** (−2.72) (−1.14) (2.05) (2.29) POST −0.010*** −0.010** 0.106** 0.123** (−2.68) (−2.12) (2.13) (2.42) TRAIN 0.003 0.005 −0.102 −0.070 (0.45) (0.56) (−1.22) (−0.71) SIZE 0.005*** 0.006*** −0.111*** −0.117*** (2.89) (2.70) (−2.81) (−3.81) LEV −0.067*** −0.069*** 0.987*** 1.063*** (−8.30) (−6.03) (6.86) (5.82) CR 0.003*** 0.003*** 0.003 −0.000 (9.04) (5.99) (1.01) (−0.04) TATR −0.017*** −0.031*** −0.023 −0.010 (−4.44) (−6.83) (−0.46) (−0.18) WXRATIO −0.066** −0.009 0.188 0.371 (−2.35) (−0.21) (0.28) (0.49) CASH −0.083*** −0.064*** 0.647** 0.543* (−4.52) (−3.33) (2.25) (1.79) MFEE −0.021 0.092*** −0.027 0.299 (−0.89) (2.87) (−0.08) (0.72) SOE −0.007*** −0.019*** −0.009 −0.040 (−2.64) (−4.16) (−0.13) (−0.61) EST_AGE −0.006** −0.014*** −0.017 −0.019 (−2.56) (−3.67) (−0.43) (−0.60) HDINDEX −0.020 −0.053*** 0.263 0.667*** (−1.27) (−2.71) (1.12) (2.94) AIRPORT −0.002 −0.005 −0.046 −0.055 (−0.96) (−1.26) (−1.09) (−1.19) GDP 0.003*** 0.004*** 0.024 0.027 (2.74) (2.79) (1.44) (1.44) BM −0.114 −0.360*** (−1.20) (−3.60) IND Yes Yes Yes Yes YEAR Yes Yes Yes Yes N 1621 1442 1356 1356 2 2 Adj. R /Pseudo R 0.43 0.37 0.34 0.24 The numbers in parentheses are the t(z)-statistics. *, **, and *** represent significance at the 0.1, 0.05, and 0.01 levels, respectively. The standard errors are robust to heteroscedasticity and are clustered by firm. 294 Z. JIN, ET AL. market return variables BH12 and BH24 as the dependent variables, respectively, and examine the impact of the opening of high-speed rail on post-IPO market performance. We find the coefficients on POST are significantly negative in Column 1 and 2 of Table 13 but significantly positive in Column 3 and 4 of Table 13, indicating that companies succeeding to list after the opening of the high-speed rail have better post-IPO account- ing performance and market performance. The results of Table 13 show that the improvement of transportation infrastructure reduces the cost of access to company information by regulators, intermediaries and investors, allowing them to allocate financing opportunities to better-quality companies and improve the efficiency of capital market resource allocation. 5. Conclusions For a developing country with an underdeveloped financial market, financing through initial public offering (IPO) has a crucial impact on the company’s own development and the economic growth of the region in which it operates. The existing research mainly focuses on some factors that influence companies’ going public from the perspective of the soft environment such as legal system, social relation and government intervention. But surprisingly little attention has been paid to the role of hard factors such as transpor- tation infrastructure. In this paper, taking the opening of high-speed rail of cities in China as a quasi-natural experiment, using the company going public during 2004–2017 as an experimental scenario, we apply the differences-in-differences model (DID) to investigate how the improvement of transportation infrastructure affects the company’s IPO beha- viour, which, on the one hand, reveals the impact of the hardware environment on the allocation of capital market resources, on the other hand, reveals an important role of transportation infrastructure to promote economic growth from the micro-level of the company. Our results show that: First, compared with other cities that have not yet opened high-speed rail, the number of local companies’ IPO applications increases significantly after the opening of high- speed rail in one city, and the conclusion holds after a series of robust checks, indicating that the improvement of transportation infrastructure improves companies’ financing activities. We also find the above results are more evident in cities other than the three major city groups, cities that have a high-speed rail going straight to the big four cities and cities without airlines. Second, after the opening of high-speed rail in a city, not only the companies’ motiva- tion applying for listing is strengthened, but also the approval probability of the com- pany’s IPO application increases significantly. We further investigate the channels through which high-speed rail opening affects the company’s IPO application, and we find that the opening of high-speed rail reduces the cost of company information acquisition for intermediaries and venture capitalists, making it easier for companies to obtain venture capital investment and hire high-quality brokerages or auditors, thereby increasing the motivation and approval probability of IPO application. Besides, the opening of the high- speed rail line helps intermediaries and venture capitalists to pick out better-quality companies. Our empirical results confirm this expectation: the opening of the high- speed rail significantly improves the performance and growth of applicant company. CHINA JOURNAL OF ACCOUNTING STUDIES 295 Third, the opening of the high-speed rail can enhance the motivation of companies to go public, increase the approval probability of IPO applications, and allow regulators to allocate listing resources to better-quality companies. So, how is the post-IPO perfor- mance of companies succeeding in issuing shares? The results show that after the open- ing of the high-speed rail, companies have better post-IPO accounting performance and market returns. Further, the opening of high-speed rail promotes information dissemina- tion, reduce information asymmetry between investors and companies, thereby reducing the company’s IPO underpricing and the cost of corporate financing. In brief, taking the opening of China’s high-speed rail as a natural experiment, we find that the improvement of transportation infrastructure in a region can improve allocation efficiency of the capital market resource. Acknowledgments We acknowledge the financial support from the Fundamental Research Funds for the Central Universities of China (‘Confucian culture and income Distribution: Evidence from the firm-level data’ (Grant No. JBK2005003)). Disclosure statement No potential conflict of interest was reported by the authors. References Allen, F., & Faulhaber, G.R. (1989). Signaling by underpricing in the IPO market. Journal of Financial Economics, 23(2), 303–323. https://doi.org/10.1016/0304-405X(89)90060-3 Atanassov, J. (2013). Do hostile takeovers stifle innovation? Evidence from antitakeover legislation and corporate patenting. The Journal of Finance, 68(3), 1097–1131. https://doi.org/10.1111/jofi. Ball, R., Robin, A., & Wu, J.S. (2003). Incentives versus standards: Properties of accounting income in four East Asian countries. Journal of Accounting and Economics, 36(1–3), 235–270. https://doi.org/ 10.1016/j.jacceco.2003.10.003 Bernstein, S., Giroud, X., & Townsend, R.R. (2016). The impact of venture capital monitoring. The Journal of Finance, 71(4), 1591–1622. https://doi.org/10.1111/jofi.12370 Bertrand, M., Duflo, E., & Mullainathan, S. (2004). How much should we trust differences-in- differences estimates? The Quarterly Journal of Economics, 119(1), 249–275. https://doi.org/10. 1162/003355304772839588 Bertrand, M., & Mullainathan, S. (2003). Enjoying the quiet life? Corporate governance and manage- rial preferences. Journal of Political Economy, 111(5), 1043–1075. Cai, W., Hu, Z., & He, F. (2013). Political connections, venture capital investments and IPO opportu- nities: Evidence from start-ups in China. Journal of Finance and Economics, 39 (5), 51–61. (in Chinese). Chan, L.H., Chen, K.C.W., Chen, T.-Y., & Yu, Y. (2012). The effects of firm-initiated clawback provisions on earnings quality and auditor behavior. Journal of Accounting and Economics, 54(2–3), 180–196. https://doi.org/10.1016/j.jacceco.2012.05.001 Chemmanur, T.J., Hul, T.J., & Krishnan, K. (2014). Expertise or proximity in international private equity? Evidence from a natural experiment. (Working paper). Chen, K.C.W., & Yuan, H. (2004). Earnings management and capital resource allocation: Evidence from China’s accounting-based regulation of rights issues. The Accounting Review, 79(3), 645–665. https://doi.org/10.2308/accr.2004.79.3.645 296 Z. JIN, ET AL. Chen, Q., Chen, X., Schipper, K., Xu, Y., & Xue, J. (2012). The sensitivity of corporate cash holdings to corporate governance. Review of Financial Studies, 25(12), 3610–3644. https://doi.org/10.1093/rfs/ hhs099 Chen, Y., Zheng, D., & Li, L. 2014. Non-SOE IPO firms’ relationship with stock issuance examination and verification committee, IPO verification and future performance. Accounting Research, (2), 12–19+94. (in Chinese). Cheng, Q., Du, F., Wang, B.Y., & Wang, X. (2019). Do corporate site visits impact stock prices? Contemporary Accounting Research. Choi, J.-H., Kim, J.-B., Qiu, A.A., & Zang, Y. (2012). Geographic proximity between auditor and client: How does it impact audit quality? Auditing-a Journal of Practice & Theory, 31(2), 43–72. https://doi. org/10.2308/ajpt-10241 Chu, J., & Fang, J. (2016). Margin-trading, short-selling and analysts’ forecast optimism. Economic Research Journal, 51 (5), 143–158. (in Chinese). Coval, J.D., & Moskowitz, T.J. (2001). The geography of investment: Informed trading and asset prices. Journal of Political Economy, 109(4), 811–841. https://doi.org/10.1086/322088 Dai, Y., Pan, Y., & Chen, J. (2014). Dual sponsors reputation, social trust and IPO approval. Journal of Financial Research, 6, 146–161. ((in Chinese).). DeFond, M.L., Hung, M., Li, S., & Li, Y. (2015). Does mandatory IFRS adoption affect crash risk? The Accounting Review, 90(1), 265–299. https://doi.org/10.2308/accr-50859 Du, X., Lai, S., & Du, Y. (2013). Issue committee relationship, hidden rules and resource allocation efficiency of IPO market. Journal of Financial Research, 3, 143–156. ((in Chinese).). Fan, J.P.H., & Wong, T.J. (2002). Corporate ownership structure and the informativeness of account- ing earnings in East Asia. Journal of Accounting & Economics, 33(3), 401–425. https://doi.org/10. 1016/S0165-4101(02)00047-2 Giroud, X. (2013). Proximity and investment: Evidence from plant-level data. The Quarterly Journal of Economics, 128(2), 861–915. https://doi.org/10.1093/qje/qjs073 Han, B., Kong, D., & Liu, S. (2018). Do analysts gain an informational advantage by visiting listed companies? Contemporary Accounting Research, 35(4), 1843–1867. https://doi.org/10. 1111/1911–3846.12363 Huang, Z., Liu, J., & Ma, G. (2016). Geographic location, China railways high-speed and informa- tion: Evidence from China’s IPO market. The Journal of World Economy, 39 (10), 127–149. (in Chinese). Imbens, G.W., & Wooldridge, J.M. (2009). Recent developments in the econometrics of program evaluation. Journal of Economic Literature, 47(1), 5–86. https://doi.org/10.1257/jel.47.1.5 Kedia, S., & Rajgopal, S. (2011). Do the SEC’s enforcement preferences affect corporate misconduct? Journal of Accounting and Economics, 51(3), 259–278. https://doi.org/10.1016/j.jacceco.2011.01.004 Kraft, A.G., Vashishtha, R., & Venkatachalam, M. (2018). Frequent financial reporting and managerial Myopia. The Accounting Review, 93(2), 249–275. https://doi.org/10.2308/accr-51838 Li, M., & Liu, F. (2012). Social capital, property right nature and listing qualification: Empirical evidence from IPO of small and Medium-sized board. Management World, 11, 110–123. ((in Chinese).). Lin, S., & Wei, M. (2000). Earnings management during initial public offerings of Chinese A-share firms. China Accounting and Finance Review, 2 (2), 87–108. (in Chinese). Liu, J., & Zhu, M. (2015). Will geography affect the cost of IPO in China?: Based on the perspective of “soft information asymmetry”. Economic Management, 37 (10), 31–41. (in Chinese). Liu, Y., & Li, Y. 2017. High-speed rail construction and city economic growth in China. Journal of Financial Research, (11), 18–33. (in Chinese). Ljungqvist, A. (2007). IPO Underpricing. Handbook of Empirical Corporate Finance, 1, 375–422. Long, Y., Zhao, H., Zhang, X., & Li, Y. (2017). High-speed railway and venture capital investment. Economic Research Journal, 52 (4), 195–208. (in Chinese). Malloy, C.J. (2005). The geography of equity analysis. The Journal of Finance, 60(2), 719–755. https:// doi.org/10.1111/j.1540-6261.2005.00744.x CHINA JOURNAL OF ACCOUNTING STUDIES 297 Petersen, M.A., & Rajan, R.G. (2002). Does distance still matter? The information revolution in small business lending. The Journal of Finance, 57(6), 2533–2570. https://doi.org/10.1111/1540-6261. Shen, W., Ye, X., & Xu, W. (2013). Is there timing behavior in VC-backed IPOs: An empirical evidence from China. Nankai Business Review, 16 (2), 133–142. (in Chinese). Stein, J.C. (2002). Information production and capital allocation: Decentralized versus hierarchical firms. The Journal of Finance, 57(5), 1891–1921. https://doi.org/10.1111/0022-1082.00483 Wong, T.J. (2016). Institutions, governance and accountability: A review of corporate governance research on listed firms in China. Foundations and Trends in Accounting, 9(4), 259–326. https://doi. org/10.1561/1400000039 Yang, Z. (2013). Do political connections add value to audit firms? Evidence from IPO audits in China. Contemporary Accounting Research, 30(3), 891–921. https://doi.org/10.1111/j.1911-3846.2012. 01177.x Zeng, Q., Chen, X., & Hong, L. (2016). VC backing, approval probability of IPO application for the first time, and durations of IPO process: Evidence from Chinese SME and GEM. Journal of Management Sciences in China, 19 (9), 18–33. (in Chinese). Zhang, J. (2017). High-speed railway construction and county economic development: A study based on satellite light data. China Economic Quarterly, 16 (4), 1533–1562. (in Chinese). Zhang, M., Yu, F., Zhong, C., & Lin, F. (2018). High-speed rail networks, market access and business productivity. China Industrial Economics, 5, 137–156. ((in Chinese).). Zhang, T., Xu, D., & Zhu, C. (2012). Political relations and IPO opportunities of non-SOE firms. Journal of Audit and Economics, 27 (3), 76–85. (in Chinese). Zhao, J., Huang, J., & Liu, F. (2018). The opening of high-speed rail and stock price crash. Management World, 34 (1), 157–168+192. (in Chinese). Zhong, Q., & Lu, Z. (2018). Can the opening of capital markets improve the information content of share prices?: Based on the empirical test of the Shanghai-Hong Kong Stock Connect Effect. Management World, 34 (1), 169–179. (in Chinese). Zhou, H., & Zheng, X. (2012). Transport infrastructure quality and economic growth: Evidence from China’s railway acceleration. The Journal of World Economy, 35 (1), 78–97. (in Chinese). Zhu, H., & Qian, Y. 2010. Pricing efficiency view” or “Rent distribution view”? Management World, (6), 28–40. (in Chinese).
Journal
China Journal of Accounting Studies
– Taylor & Francis
Published: Apr 2, 2020
Keywords: Transportation infrastructure; company going public; high-speed rail opening; information asymmetry
