Money talks? The impact of excess funding on post-IPO performance

Abstract

How does the excessive money raised from initial public offerings (IPOs) affect the firms’ operating efficiency and long run performance? Using a sample of 1621 Chinese IPOs over the period of 1999–2014, we find that the excess funding aggravates investment inefficiency. Moreover, the negative impact of excess funding on investment inefficiency is more pronounced for non-state-owned enterprises (non-SOEs) but not state-owned enterprises (SOEs). For non-SOE IPOs, excess funding has a negative impact on long-term performance, and the negative impact is further punctuated by investment inefficiency. We conclude that the excessive money infusion is detrimental to post-IPO performance.

1 Introduction

Money could be an evil that does well, or could be a good that does evil. The noted phenomenon of initial public offerings (IPOs) long-run underperformance has been extensively explored ever since Ritter (1991) and Loughran and Ritter (1995). Researchers proposed alternative explanations to this anomaly, e.g. market timing hypotheses (Loughran and Ritter 1995), overreaction (Levis 1993; Aggarwal et al. 1993), behavioral explanation (Ritter 1991; Lerner 1994; Loughran and Ritter 2000), or underwriter’s attributes (Carter et al. 1998; Chen et al. 2017a, b). Other than these explanations, Billett et al. (2011) and Huang and Ritter (2022) find that post-IPO performance is negatively affected by the frequency and size of subsequent security issuances. This seems to imply that excess funding might not be a good thing to IPO firms.

The impact of money inflow on performance should be contextually dependent. In this study, we explore how the IPO proceeds affects firm’s investment inefficiency and therefore post-IPO performance. The examination was conducted via the platform of the Chinese IPO market which was firstly noted for the high initial return.¹ This IPO anomaly was rectified after the abolishment of quota system and the imposed limit in setting offer price in 1999. After that, excess funding emerged as a new anomaly to attract numerous attentions.

Even though IPO excess funding is a global phenomenon (Loughran et al. 1994), it has not been widely noted in IPO literature because it is relatively minor in other markets and there is no universally-accepted definition of excess funding in literature. Firstly, it is not a minor issue in the Chinese IPO markets. According to Xu (2013), the excess funding reached 255% for 590 Chinese IPOs in 2010–2011 sampling period. Wan (2014) indicated that 88.22% of China IPOs in 2009–2014 were characterized as excess funding. By contrast, the excess funding ratio for U.S. SEOs is merely 3.91% (Chan et al. 2018) and − 4% for U.S. IPOs (Jeppsson 2018). We believe that our sample showing an average excess funding of 66.92% associated with 1999–2014 China’s IPOs should have a material effect on IPO firms. Specifically, we focus on its effect on the IPO firm’s follow-on investment inefficiency and long-term performance. Secondly, we define the excess funding as the difference between real IPO proceeds and the needed for investment in projects.² The data are accessible because this information disclosure is mandatory for IPO firms in China.

We firstly relate the excess funding to the agency problem and therefore investment inefficiency.³ Specifically, IPO firms obtaining funding in excess of their planned need tend to overinvest to the suboptimal level. The excess funding therefore represents a surrogate of agency costs (Jensen 1986) and is positively correlated with investment inefficiency. More importantly, we find that the adverse impact of excess funding on investment inefficiency is more pronounced for non-state-owned enterprises (non-SOEs) but not state-owned enterprises (SOEs). This is could be understood that prior to IPO listing, non-SOEs tend to confront with a severer financial constraint than SOEs in the first place (e.g., Cull et al. 2015). The incremental agency problem that is associated with excess funding and therefore take a toll on investment inefficiency is severer for non-SOEs than for SOEs.

Second, we find that IPO excess funding has a negative impact on long-run performance for non-SOEs, and the negative impact is further reinforced by investment inefficiency. Ritter (1991) and Loughran and Ritter (1995) hypothesize that long-term underperformance arises from investors' overoptimism about the issuing firm’s future performance. However, our results show that investors’ overoptimism has only a significant impact on post-one-year performance. In contrast, the investment inefficiency indebted to excess funding has far-reaching negative impacts on long-term performance up to three years post IPO.

Two advantages are associated with the use of Chinese IPOs in testing this issue. First, the Chinese government specifically requires all IPO firms to explicitly reveal the expected use of funds. With this information we could have a clear definition of excess funding. Second, as the focus of anomalies has gradually transited from high initial return to excess funding in the Chinese IPO market, we expect that the average excess funding ratio of 67% should have material effects on firms’ follow-on investment inefficiency and long-run performance. Someone might have the intuition that more money raised from IPOs benefits the issuing firms. However, we find the opposite: the excessive money infusion has adverse impact on IPO firm’s follow-on investment efficiency and performance.

The potential contributions of this study are multifold. First, we document the excess money raised from IPO implies agency cost (Jensen 1986) and therefore have adverse impact on firm’s following on investment efficiency and long-run performance. Second, we find that excess funding has more explanatory power on post-IPO performance than investor’s overoptimism (e.g., Ritter 1991; Loughran and Ritter 1995). Specifically, we document that the excess funding takes a toll on long-run performance via the intermediary role of investment inefficiency. Third, the funding-investment-performance relation is verified after taking account of the endogeneity issue and a battery of robustness checks.

We briefly compare our findings with similar studies on Chinese IPOs. Firstly, we examine the agency problems associated with excess funding, which manifest as follow-on investment inefficiencies and, consequently, inferior long-term performance. Prior research, including Xu and Xia (2012), Zhao et al. (2017), Zhang et al. (2021) and Wu et al. (2022), also highlights the negative aspects of excess IPO funding. They argue that excess funding can be detrimental due to over-investment issues (Xu and Xia, 2012; Wu et al. 2022), managerial private benefits (Zhao et al 2017), managerial empire building (Zhang et al. 2021), and acquisitiveness (Wang et al. 2023). In contrast, Cao et al. (2023) present a more positive view, suggesting that excess IPO funding can enhance post-IPO performance by alleviating a firm’s financial constraints. Additionally, we find that the negative impact of excess IPO funding is more pronounced for non-SOEs than for SOEs. Similarly, Xu and Xia (2012) report that over-investment problems related to excess funding are more severe for firms with poor internal governance mechanisms and non-SOEs. Zhao et al. (2017) also demonstrate that the adverse effects of excess IPO funding are more significant for non-SOEs compared to SOEs.

The rest of this paper is organized as follows. Section 2 describes the institutional background of the China IPO market. Section 3 reviews the related literature and develops hypotheses. Section 4 outlines the data, variables, and empirical models. Section 5 summarizes empirical findings. Section 6 concludes.

2 Institutional background of China IPOs

Ever since the Shanghai and Shenzhen stock markets were established in 1990 and 1991, the number of listed firms almost doubled⁴ in one decade. Despite its remarkable development, the China stock market has been striking a delicate balance between planned and market mechanisms. The highly regulated market has been noted by the enormous initial returns and long-term underperformance associated with IPOs. These anomalies are jointly attributed to market imperfection (including information asymmetr⁵ and overreaction) and governmental interventions (including quota system,⁶ limits on offer price-setting,⁷ split-share structure,⁸ and government subsidies⁹).

The IPO market in China has been suspended several times due to the market downturn,¹⁰ financial crisis,¹¹ and financial inspection.¹² These risks resulted in an enormous initial return.¹³ This is the first anomaly that attracted public attention.

Excess funding in China has not been widely noted until the Growth Enterprise Markets (GEM) which was initially inducted on October 30, 2009. The IPO shares listed in GEM were noted for high offer prices, price-to-earnings ratios, and IPO proceeds. The proceeds collected from GEM IPOs were almost 1.5 times the claimed investment. Xu (2013) indicate that the excess funding ratio is 255% for 590 IPOs in 2010 and 2011. More importantly, the excess funding of RMB 307 billion was not matched with corresponding projects, and 72% of cash was still held two years after IPOs. According to Wan (2014), 88.22% of IPOs in 2009–2014 are characterized as excess funding.

Someone indicates that excess funding was attributed to market cyclicality (e.g., Li 2010), indicating that it is more likely to occur in bullish than bearish markets. Moreover, after the reopening of IPOs in 2009, many industry-leading IPOs were associated with high PE ratios: 42 in the mainboard market, 50 in the small board market and 68 in the growth enterprises market. Although excess funding is a global phenomenon, it is much more prominent in China not only due to its high-level oversubscriptions but also it serving as a major channel for Chinese-listed firms to mitigate financial constraints.¹⁴

3 Hypothesis development

In this section, we develop hypotheses to illustrate how the excess funding from an IPO affects the firm’s follow-on investment inefficiency, and therefore long-term performance.

3.1 Excess funding and investment inefficiency

Is holding excess cash beneficial or detrimental to firm value? From the perspective of agency problems, excess funding aggregates agency problems associated with free cash flow that result in overinvestment (e.g., Chen et al. 2016) and empire-building (e.g., Jensen 1986; Harford 1999; Brahma and Economou 2024; Lo and Shiah-Hou 2022). If this is the case, the excess money infusion from IPO implies the increase of firm’s free cash flow, which in turns deteriorates investment efficiency. For the perspective of information asymmetry, Myers and Majluf (1984) indicate that financially constrained firms tend to be confined in investment to their internally generated funds (e.g., Hubbard 1998).

Firstly, if excess funding is mild, its net impact on investment inefficiency could be negative because the benefit from ameliorating information asymmetry due to mild excess funding outweighs the cost from aggravating agency problems. In contrast, if excess funding is severe, its net impact on investment inefficiency could be positive when the aggravation of agency problems outweighs the mitigation of information asymmetry. Given the case of China’s IPOs, the excess funding is significant. We therefore expect that the excess funding is positively correlated with investment inefficiency.

Secondly, prior studies indicate that as compared to SOEs, non-SOEs in China are confronted with severe financial constraints (e.g., Allen et al. 2005; Ding et al. 2013; Cull et al. 2015). Because SOEs have a better relationship with the government (e.g., Xu et al. 2022) and are less financially constrained than non-SOEs, the degree of excess funding is expected to be lower for SOEs than non-SOEs. Given that the excess funding has adverse impact on IPO firm’s investment efficiency, we would expect to find that the adverse impact of excess funding on investment efficiency would be much salient for non-SOEs than for SOEs.

Hypothesis 1

Excess funding would result in investment inefficiency, and the adverse impact is more salient for non-SOEs than SOEs.

3.2 Excess funding, investment inefficiency, and long-term performance

How does excess funding affect firm’s long-term performance? Prior studies such as Ritter (1991), Loughran and Ritter (1995, 2002), Baker and Wurgler (2000) and Hirshleifer (2001) indicate that IPO firms and their investment bankers tend to take timing advantage by selling overpriced shares to overly optimistic investors. Excess funding in some sense captures the timing opportunity and therefore is expected to be negatively related to long-term performance. As indicated earlier, excess funding results in investment inefficiency, which is also consistent with the observation of Fu (2010) indicating that overinvestment results in a reduction in asset productivity. The further question is how the negative relation between excess funding and IPO long-run performance is moderated by investment inefficiency. As indicated in Fu (2010), underperformance of SEO firms is the result of managers' overinvestment. We expect that investment inefficiency further punctuates the negative relation between excess funding and long-run performance. Since non-SOEs are more likely to be financially constrained than SOEs, the influx of excess funding has a greater impact on the investment inefficiency for non-SOEs than SOEs. Therefore, the moderating effect of investment inefficiency on the relation between excess funding on long-term performance is more salient for non-SOEs than SOEs.

Hypothesis 2

Excess funding is negatively correlated with long-run performance, and the negative impact of excess funding on long-run performance is further punctuated by investment inefficiency more for non-SOEs than SOEs.

Figure 1 illustrates the timeline for setting the offer price, mid-term investment inefficiencies, and long-term performance measures. Identifying this time sequence may help mitigate concerns about endogeneity.

Fig. 1

Timeline for setting the offer price, mid-term inefficiencies, and long-term performance measures

4 Sample description

Our sample consisting of 1621¹⁵ Chinese IPOs in the sampling period of 1999–2014. The reason why our sampling period starts in 1999 and ends at 2014 is mainly due to the concern that CSRC imposed quota and offer price management on China’s IPO market before 1999 and the requirement of disclosing investment amount has been abolished since 2014. In order to identify excess funding, we hand-collected the data on the expected investment, the expected funds needed, and final amount of funds raised in the IPO from IPO prospectuses and annual reports. We define excess funding as the difference between IPO proceeds and the expected investment, divided by the expected investment. This definition aligns closely with prior studies such as Xu (2013), Xu and Xia (2012), and Zhang et al. (2021). In contrast, Cao et al. (2023) and Zhao et al. (2017) define it as the difference between IPO realized proceeds and the expected raised funds. We postulate that our definition more directly measures excess funds that could potentially lead to follow-on investment inefficiencies. The other data such as accounting, financial information, and ownership are collected from China Stock Market and Accounting Research (CSMAR). Financial firms that are subject to different regulations are excluded from the sample.

Table 1 reports distribution of number of IPOs, excess funding (EF_ratio), the percentage of excess funding (D_EF(%)), and the percentage of SOE IPOs in 1999–2014 sampling period. The number of IPOs exhibited a significant increase from 31 in 1999 to 338 in 2010. Excess funding (EF_ratio) is defined as the difference between the final amounts of funds raised from the IPO and the expected investment divided by the expected investment.¹⁶ The result shows that excess funding indeed exhibited a significant surge in 2006, the year when the regulation of P/E caps was released. For example, the average excess funding rate was 25.95% in 2006, implying that IPO firms obtained about 26% more in funds than their planned investment. The average excess funding surged to 172.25% in 2010 and 118.16% in 2011. This implies that IPO firms in those two years obtained funds that were more than double than their planned investment.

Table 1 Sample distribution

We note that although excess funding exhibited a significant increase in 2006, there were sporadic cases of excess funding before 2006. Our sample, which extensively covers excess funding over a long sampling period, could provide a comprehensive picture. We also explore the subsamples beginning in 2006 and find the results are qualitatively similar. The empirical results will be provided upon request.

We also report the percentage of IPO firms that exhibit excess funding (D_EF, defined as a dummy that is assigned the value 1 when the IPO proceeds exceed the expected investment and 0 otherwise). The ratio of firm exhibiting excess funding has exceeded 90% since 2009, and was even as high as 97.63% in 2010. Nevertheless, there were sporadic cases of excess funding found before 2006. In response to the over-heated IPO market that resulted in a high percentage of firms raising much more funds than their planned investments, the Chinese regulatory entity cooled down the market by suspending the approval of IPO applications in 2013. The overall average excess funding percentage is 63.79% throughout the sampling period.

The last column of Table 1 reports the percentage of IPOs by state-owned enterprises (SOE_S). It shows that the percentage of SOE IPOs gradually decreased over the passage of time. The percentage decreased from 59% in 2001 to 4.36% in 2011.

5 Empirical results

5.1 Descriptive statistics

Table 2 reports the summary statistics of variables. Panel A reports the long-term performance measures of IPOs in the sampling period. We alternatively use the wealth relative measure proposed by Ritter (1991), the buy-and-hold returns, (i.e., market performance) returns on equity and returns on asset of post-IPO 1 through 3 years (i.e., financial performance). In reference to Ritter (1991), the wealth relative measure (WR) is calculated as follows.

$${WR}_{n}=\frac{{\prod }_{t=1}^{12*n}\left(1+{R}_{i,t}\right)}{{\prod }_{t=1}^{12*n}\left(1+{R}_{m,t}\right)} , n=\text{1,2},3.$$

where R_i,t denotes the return of IPO firm i in month t, and R_m,t denotes the corresponding market return in month t. This measure represents the IPO firm’s long-run buy-and-hold returns in excess of corresponding market returns. The mean wealth relative measures in post-1 year through -3 years are in the narrow range of 0.937–0.965, indicating that these IPO firms are slightly lower than the corresponding market index returns. Moreover, we also include alternative performance measures including buy-and-hold returns (BHR), returns on equity (ROE) and returns on asset (ROA). The buy-and-hold returns (BHR) are calculated as follows.

$${BHR}_{n}={\prod }_{t=1}^{12*n}\left(1+{R}_{i,t}\right)-{\prod }_{t=1}^{12*n}\left(1+{R}_{m,t}\right), n=\text{1,2},3.$$

(1)

where BHR_n denotes buy-and-hold returns for 1, 2, and 3 years post IPO. The mean buy-and-hold returns for 1, 2- and 3-years post IPO are − 9%, − 6%, and − 11%, respectively. The corresponding mean ROE are 9.7%, 9.6%, and 7.8%, respectively. The corresponding mean ROA are 7.4%, 6.8%, and 6.6%, respectively.

Table 2 Summary statistics

Panel B reports excess funding. We alternatively define excess funding ratio (EF_ratio) as the difference between proceeds from IPO and expected investment divided by expected investment. The average excess funding ratio is 66.92%, implies that IPO firms in general raise 66.92% more funds than their planned investments. We also explore whether IPO firms exhibit excess funding. The excess funding dummy is assigned the value 1 when the IPO firm raises proceeds more its planned investment and 0 otherwise. One average, 63.79% of the sampling IPO firms are associated with excess funding.

Panel C reports the statistics of variables for testing investment inefficiency. Both investment (INV) and cash flow (Cash) are deflated by the beginning of total assets, and are 7.8% and 7.9% on average. Leverage (Lev) is defined as the total debts divided by total assets and is 27.5% on average. Size denotes the natural logarithm of total assets and is 8.807 on average. Liquidity (Liq) is defined as current assets divided by total assets and is 76.1% on average. Wedge is defined as the difference between controlling owner’s voting rights and cash flow rights and is 4.369% on average. We also include variables of corporate governance as control variables. Ownership concentration (Own_Con), defined as the sum of the top five shareholders’ ownership, is 65.96% on average. Compensation, defined as the natural logarithm of the top three managers’ compensation, is 5.896 on average. Board independence (Ind_Dir), defined as the percentage of independent directors in the board, is 37.1% on average. Moreover, the sales growth rate is 98.7% on average. The mean earnings per share (EPS) is 0.496 RMB.

Panel D reports the control variables of underwriting and characteristics of IPO firms. The selection of these variables were jointly referred to in prior studies (e.g., Boubakri and Cosset 1998; Sun and Tong 2003; Fan et al. 2007). The mean initial return is 82%.¹⁷ The mean age¹⁸ of IPO firms, measured by years since inception, is 7.32 years. The mean underwriting expense per share is 1.23 RMB. The board on average consists of 9.00 directors. Tobin’s Q, calculated as the sum of the market value of equity and the book value of liabilities divided by the book value of assets, is 2.88 on average. Turnover, denoting the turnover rate on the listing day, is 0.63 on average. The real issuing shares divided by the maximum number of issuing shares (Real_Issue) is 0.99 on average, indicating that most IPO firms issue shares to the maximum allowable limit.

We propose two measures to determine whether offer prices and post-IPO prices are set too high or too low¹⁹: PE_offer/1 yr (defined as the ratio of PE on the offering day to the PE one year post IPO) and PE_IPO/1 yr (defined as the ratio of PE on the first listing day to the PE one year post IPO). The two measures are used as surrogates of investors' overoptimism. We find that the mean and median of PE_offer/1 yr are 1.144 and 0.956, respectively. The fact that both are close to 1 indicates that the offer price is rationally set. However, the mean and median of PE_IPO/1 yr are 1.633 and 1.570, respectively. The price-to-earnings ratio of much higher than 1 indicates that post-IPO prices are too high, and investors overreact.

Among the total sample of Chinese IPOs, 22% are from SOEs. As shown in Table 1, the percentage of SOE IPOs exhibits a decrease over the passage of time. We use the dummy of exchange (Exchange) that is assigned the value 1 for IPOs on the Shanghai Stock Exchange and 0 for IPOs on the Shenzhen Stock Exchange. On average, 67.92% of IPOs were on the Shanghai Stock Exchange. Finally, we use the dummy of Growth Enterprise Exchange Market (GEM) to indicate percentage of IPOs from GEM; the percentage is 24.86%. Finally, the average percentage of IPO issued in hot markets is 20.48%.

5.2 Regression analyses

5.2.1 Excess funding and investment inefficiency

In this section, we explore how excess funding is related to investment inefficiency. The proxies of investment inefficiency are alternatively gauged in reference to Richardson (2006), Chen et al. (2011) and Dai et al. (2018). The first one (Inefficiency) is gauged by the absolute value of the residual term (| ε_it |) in the year of IPO:

$$INV_{i} = \alpha + \beta_{1} EPS_{i} + \beta_{2} Lev_{i} + \beta_{3} Liq_{i} + \beta_{4} Cash_{i} + \beta_{5} Sales\_Gr._{i} + \beta_{6} Age_{i} + \beta_{7} Size_{i} + \, \varepsilon_{i} ,$$

(2)

where EPS denotes earnings per share; Lev denotes leverage; Liq denotes liquidity; Cash denote the cash flow from operations; Sales_Gr denotes sales growth rate; Age denotes firm’s age; and Size denotes firm’s size.

The following two measures (Inefficiency_1 and Inefficiency_2) are gauged by the absolute value of the residual term (|ε_i|) of the following two regression specifications.

$$INV_{i} = \alpha + \beta_{1} TobinQ_{i} + \beta_{2} Lev_{i} + \beta_{3} Liq_{i} + \beta_{4} Cash_{i} + \beta_{5} Sales\_Gr._{i} + \beta_{6} Age_{i} + \beta_{7} Size_{i} + \, \varepsilon_{i } ,$$

(3)

$$INV_{i} = \alpha + \beta_{1} MB_{i} + \beta_{2} Lev_{i} + \beta_{3} Liq_{i} + \beta_{4} Cash_{i} + \beta_{5} Sales\_Gr._{i} + \beta_{6} Age_{i} + \beta_{7} Size_{i} + \, \varepsilon_{i}$$

(4)

The fourth measure (Inefficiency_3) is gauged by the absolute value of the residual term (| ε_it |) of the following regression specification (Biddle et al. 2009).

$$INV_{i} = \alpha + \beta_{1} SalesGrowth_{i} + \varepsilon_{i} ,$$

(5)

In Table 3, we regress investment inefficiency on excess funding ratio and other control variables with respect to non-SOEs and SOEs, respectively. Specifically, the following regression specification is adopted.

$$\begin{aligned} & Inefficiency_{i} = \alpha + \beta_{1} EF_{i} + \beta_{2} Compensation_{i} + \, \beta_{3} Ind.\_Dir_{i} + \, \beta_{4} Own\_Con_{.i} + \beta_{5} Size_{i} + \, \beta_{6} Cash_{i} + \, \beta_{7} Lev_{i} \\ & \quad + \, \beta_{8} Liq_{i} + \, \beta_{9} Sales\_Gr_{i} + \beta_{10} EPS_{i} + \beta_{11} Tobins \, Q_{i} + \beta_{12} Wedge_{i} + v_{i,} , \\ \end{aligned}$$

(6)

where Inefficiency_i is alternatively gauged as the absolute value of the residual term from Eq. (2)–(5). Compensation denotes the natural logarithm of the top three managers’ compensation, Ind_Dir denotes the percentage of independent directors, Own_Con denotes ownership concentration and is gauged by the percentage of shareholding held by the top five shareholders, and Sale_Gr denotes sales growth. EPS denotes the earnings per share. Tobin’s Q is defined as the sum of the market value of equity and the book value of liabilities divided by the book value of assets. Wedge is the difference between controlling owner’s voting rights and cash flow rights.

Table 3 The impact of excess funding on investment inefficiency

In Panel A of Table 3, the result indicates that excess funding ratio is significantly positively correlated with investment inefficiency for non-SOEs. The corresponding regression coefficients are 0.275, 0.274, 0.275 and 0.293, respectively. However, this is not the case for SOEs (Panel B of Table 3). If anything, the regression coefficients of the excess funding ratio for SOEs are negative (− 0.292, − 0.300, − 0.297 and − 0.279, respectively), albeit insignificant. Therefore, the negative impact of excess funding is more pronounced for non-SOE IPOs than SOE IPOs. The finding is consistent regardless the alternative proxies of investment inefficiency, and therefore in supportive of Hypothesis 1 indicating that excess funding results in investment inefficiency, and the impact is more salient for non-SOEs than SOEs.

As for the control variables, we find that ownership concentration as measured by the total shareholding of the top-five shareholders is positively correlated with investment inefficiency for non-SOEs. The corresponding regression coefficients are in the range between 0.019 and 0.023. Again, this is not the case for SOEs. As indicated by Chen et al. (2017a, b), controlling owners having more shareholdings are capable of expropriating the rights of minority shareholders. They find that high-level ownership concentration is detrimental to the firm’s investment efficiency. Our empirical finding for the positive relation between ownership concentration and investment inefficiency to be more salient for non-SOEs than SOEs could be understood that as compared to SOEs, non-SOEs are not only more likely to be associated with excess funding and less centralized control. Therefore, as compared to SOEs, the influx of money without tight control for non-SOEs is more likely linked to agency problems and therefore investment inefficiency.

Moreover, we find that cash is positively correlated with investment inefficiency for non-SOEs (the corresponding regression coefficients are 0.094, 0.101, 0.099 and 0.040, respectively) but is negatively correlated with investment inefficiency for SOEs (the corresponding regression coefficients are − 0.298, − 0.299, − 0.298 and − 0.337, respectively). This could be understood that in a relative sense that a firm’s cash is loosely controlled in non-SOEs but tightly controlled in SOEs. Furthermore, we find that size is negatively correlated with investment inefficiency for non-SOE IPOs (the corresponding regression coefficients are − 0.971, − 0.989, − 0.973 and − 0.919, respectively), implying that large firms are less likely to overinvest or underinvest to the suboptimal level. Moreover, leverage is positively correlated with investment inefficiency for non-SOE IPOs (the corresponding regression coefficients are 0.045, 0.044, 0.044 and 0.044, respectively). This could be understood that debt decreases agency cost because creditors concerning about principal and interest payments would monitor the activities including investment so as to protect firms from becoming bankrupt. This effect of leverage on investment inefficiency is more salient for non-SOEs than SOEs.

Finally, we find that liquidity is negatively correlated with investment inefficiency both for non-SOEs (the corresponding regression coefficients are − 0.015, − 0.015, − 0.015 and − 0.013, respectively) and SOEs (the corresponding regression coefficients are − 0.058, − 0.058, − 0.058 and − 0.057, respectively). This finding is in tandem with prior studies (i.e., Butler et al. 2005) indicating that firms with liquid stocks enjoy a lower cost of equity, so that firm’s investment inefficiency is sensitive to market liquidity via equity issuance. Moreover, Khanna and Sonti (2004) using feedback theory indicate that firm’s investment decisions are affected by informed traders who weigh the prospect of their invested firms and use exaggerated trading to signal managers of their investment preferences.

5.2.2 The long-term effect of excess funding

Figure 2 illustrates how excess funding is related to post-IPO performance (ROA) up to three years post IPO. This figure shows that post-1-year ROA (ROA1) is higher than post-2-year ROA (ROA2) and post-3-year ROA (ROA3), implying that IPO underperformance is more pronounced as the passage of time after IPO. Moreover, the excess funding ratio (EF_ratio) started to increase in 2008 (− 0.98%) and reached to its highest level in 2010 (172.25%). In the meanwhile, the post-IPO performance (ROA1, ROA2 and ROA3) exhibit a trend of deterioration. By contrast, the excess funding ratio decreased in 2012 while the post-IPO performance gradually increased thereafter. This figure illustrates a negative relation between excess funding and post-IPO performance.

Fig. 2

Excess funding and post-IPO performance

In Tables 4 and 5, we investigate the impact of IPO funding, investment inefficiency and their interactions on long-term performance of IPO firms. The inclusion is for the purpose of investigating whether investment inefficiency has a moderating effect on post-IPO long-term performance measure. If the regression coefficient of the interactive term is significant, we could infer that the long-term underperformance of IPOs, a phenomenon that have been extensively explored in prior studies, is at least partially indebted to the impact of mid-term investment inefficiency owing to overflow of funds from IPOs. In Table 4, we use wealth relative measure (WR) proposed by Ritter (1991) as the proxy of long-term performance. In Table 5, we additionally include the buy-and-hold return (BHR) (e.g., McGuinness 2016), return on equity (ROE) and return on asset (ROA) as alternative proxies for long-run performance. To better capture the long-term effect, we trace the long-term proxies up to three years post IPO.

Table 4 Excess funding, investment inefficiency and long-term performance: SOEs and non-SOEs

Table 5 Excess funding, investment inefficiency and alternative long-term performance: SOEs and non-SOEs

The results in left part of Table 4 show that for non-SOE IPOs, both excess funding (the corresponding regression coefficients are − 0.050, − 0.077 and − 0.065, respectively) and investment inefficiency (the corresponding regression coefficients are − 0.077, − 0.102 and − 0.074, respectively) are negatively correlated with the wealth relative measure up to three years post IPO (WR₁, WR₂, and WR₃). More importantly, the interaction between excess funding and investment inefficiency (the corresponding regression coefficients are − 0.082, − 0.089 and − 0.134, respectively) is also negatively correlated to the wealth relative measure up to three years post IPO. In contrast, for SOEs, only investment inefficiency is negatively correlated (the corresponding regression coefficients are − 0.111, − 0.149 and − 0.281, respectively) with the wealth relative measure. The impact of excess funding and the impact the interaction between excess funding and investment inefficiency are insignificant.

In Table 5, the finding of using alternative performance proxy such as the buy-and-hold return up to three years post IPO (BHR₁, BHR₂, and BHR₃) is pretty similar to the finding from Table 4. That is, for non-SOEs, not only excess funding and investment inefficiency are detrimental to but also their interaction additionally take toll on post-IPO performance. For SOEs, only investment inefficiency is negatively correlated with BHR. The result of using ROE or ROA is qualitatively similar, albeit less significant. Moreover, investors might be interested in buy-and-hold return since it is what they receive from owning stocks. In general, the overall result is in supportive of our Hypothesis 2 indicating that excess funding is negatively correlated with long-run performance, and the negative impact of excess funding on long-run performance is further punctuated by investment inefficiency more for non-SOEs than SOEs.

The finding is inspiring. It shows that, at least for private-owned firms (non-SOE IPOs), excess funding results in investment inefficiency. This implies that firms with ample cash from IPOs tend to overinvest to the suboptimal level. The investment inefficiency indebted to excess funding then gradually accumulates to the detriment of a firm’s long-term performance. Moreover, the investment inefficiency further punctuates the negative impact of excess funding on post-IPO performance.

The results from control variables in Table 4 show that underwriting expense is positively correlated with the wealth relative measure up to three years post IPO, with the corresponding regression coefficients of 0.027, 0.039, and 0.068 for non-SOEs and 0.100, 0.236, and 0.369 for SOEs, respectively. If the expense paid to underwriter signals underwriter’s reputation (e.g. Beatty and Ritter 1986; Carter and Manaster 1990), prestigious underwriters for the purpose of protecting their underwriting reputation asset would monitor clients after IPO (Hansen and Torregrosa 1992). The monitoring and certification function embedded in underwriting expense is beneficial to post-IPO performance. The effect is significant for SOEs and non-SOEs.

Moreover, we find that turnover is significantly negatively correlated with WR₁ and WR₂ (the corresponding regression coefficients are − 0.182 and − 0.141) for non-SOEs and WR₁ for SOEs (the corresponding regression coefficient is − 0.277). Krigman et al. (1999) use first-day trading characteristics to predict post-IPO performance. They assume that the actions of institutional investors are useful to infer future returns. If these institutional investors hold on promising IPO stocks and therefore result in low turnover, these IPO stocks are assumed to perform better.

We find that board size is negatively correlated with WR₁ and WR₂ (the corresponding regression coefficients are − 0.102 and − 0.020, respectively). This is consistent with prior studies (e.g., Lipton and Lorsch 1992; Jensen 1993) indicating that larger boards are associated with ineffective board and therefore firm performance because of coordination and communication problems. We find that the negative impact of board size on post-IPO performance is only salient for non-SOEs but not SOEs. Tobin’s Q is positively correlated with WR₁ and WR₂ (the corresponding regression coefficients are 0.043 and 0.029) for non-SOEs, implying that non-SOE IPO firms with growth potentials are associated with higher performance measures. Finally, we do not provide explicit explanations for some controlled variables that are not consistent throughout the alternative use of performance proxies. For example, the GEM dummy (being assigned the value 1 when the IPO firm is listed in the Growth Enterprise Exchange and 0 otherwise) is significant for non-SOEs but not SOEs.

5.3 Robustness test

A reader might be concerned about endogeneity issue between excess funding and long-run performance. Firstly, from a temporal perspective, the setting of the offer price—an event that defines excess funding—occurs before the onset of post-IPO investment inefficiency and performance issues. The time gap between the independent variable (excess funding) and the dependent variables (investment inefficiency and post-IPO performance) suggests that our findings are more likely to reflect a causal relationship rather than mere correlation.

Furthermore, the positive relationship observed between excess funding and investment inefficiency is unlikely to be due to reverse causality. Underwriters are not inclined to set higher offer prices for firms already exhibiting signs of investment inefficiency. Therefore, our results support a causal interpretation of the impact of excess funding on investment inefficiency and post-IPO performance.

To further tackle the potential endogeneity issue, we use lagged industry excess funding (EF_industry_1) and initial return (IR)²⁰ as the instrument variables, and conduct two-stage least-square regression (2SLS). The 2SLS results of WR, summarized in Table 6, are qualitatively similar to those from Table 4. That is, excess funding is negatively correlated with WR, and the negative impact of excess funding on WR is further punctuated by investment inefficiency more for non-SOEs than SOEs. We also explore 2SLS for alternative long-term performance measures such as BHR, ROE, and ROA. The (unreported) findings are qualitatively similar, and will provide upon request. Finally, following Cao et al. (2023), we use a bull market dummy and initial return as instrumental variables and conduct a 2SLS analysis. The results (unreported) remain qualitatively similar.

Table 6 Endogeneity test: two-stage least-square regression (2SLS)

We also employ Propensity Score Matching (PSM) to find matched counterparts based on criteria including size, age, liquidity, wedge, Tobin’s Q, and ROA. The results presented in Table 7 reveal that the negative impact of excess funding and investment inefficiency is more pronounced for non-SOEs compared to SOEs.

Table 7 Propensity score matching

To examine whether investment inefficiency mediates the relationship between excess funding and long-term performance, we employ several alternative tests, including the Sobel test, Aroian test, and Goodman test. The results summarized in Table 8 indicate that investment inefficiency significantly mediates this relationship for non-SOEs, but not for SOEs.

Table 8 Test of mediating effect

Finally, we investigate potential endogeneity related to firm characteristics. The results summarized in Table 9. Our results show that firms with CEO/chairman duality, older firms, and those with high levels of political connections²¹ experience a more significant negative impact of excess funding on post-IPO performance. This negative effect is further amplified by investment inefficiencies. These findings largely align with previous research. For instance, Chowbury et al. (2023) found that highly powerful CEOs often reduce investment efficiency by increasing overinvestment, particularly when information asymmetry, agency problems, and product market competition are strong. Additionally, Guntoro et al. (2020) reported that firms in the later stages of their life cycle tend to have higher levels of cash flow and, consequently, overinvest. Furthermore, Chahal and Ahmad (2022) found that firms with strong political connections display higher levels of investment inefficiency compared to those with weaker connections. Overall, our findings reinforce the notion that agency problems related to IPO excess funding negatively impact long-term performance, with this adverse effect being exacerbated by investment inefficiencies.

Table 9 Firm heterogeneity analysis: CEO/chairman duality, firm age and political connections

6 Concluding remarks

In this study, we explore a recent issue of excess funding in the Chinese market. This market provides a splendid forum to examine the effect of excess funding on firm’s long-term performance measure. We especially focus on the contrast between SOE and non-SOE IPOs. Our findings are summarized as follows. First, excess funding positively contributes to investment inefficiency for non-SOE but not SOE IPO firms. Second, investment inefficiency is detrimental to long-run performance both for non-SOE and SOE IPOs. Third, excess funding takes toll on post-IPO performance for non-SOE but not SOE IPOs. More importantly, investment inefficiency further punctuates the negative impact of excess funding on long-run performance for non-SOE IPOs.

Our empirical findings from Chinese IPOs directly address both the free cash flow hypothesis and costly external financing hypothesis, and bridge the gap between cash holding and IPO performance. Moreover, we systematically explore how excess funding affects IPO firms’ investment inefficiency and long-term performance. This study sheds a light on the thread of studies that compare SOEs and non-SOEs using the platform of the China stock market, and also illustrates the infusion of money from IPO have policy implications: excess funding connotes the aggravated agency problems associated with free cash flow (e.g., Jensen 1986), that in turns has a negative impact on investment efficiency and performance. This study provides a framework for further studies that aim to deepen the exploration of emerging stock markets.

Appendix: Variable definitions

Variable	Description
A. Post-IPO performances
WRn	According to Ritter (1991), wealth relatives (WR) is calculated as $\text{WR}n=\frac{{\prod }_{t=1}^{12*n}(1+{R}_{i,t)}}{{\prod }_{t=1}^{12*n}(1+{R}_{m,t)}}, n=\text{1,2},3.$
BHRn	BHR is calculated as $\text{BHRn}={\prod }_{t=1}^{12*n}(1+{R}_{i,t)}-{\prod }_{t=1}^{12*n}(1+{R}_{m,t)}$,$n=\text{1,2},3,$ where Ri,t denotes the return of IPO firm i in month t, and Rm,t denotes the corresponding market return in month t
ROEn	ROEn denotes the returns on equity in n years after IPO, where $n=\text{1,2},3$
ROAn	ROAn denotes the returns onassets in n years after IPO, where $n=\text{1,2},3$
B. Excess funding
EFratio	EFratio is the difference between proceeds from IPO and expected investment revealed in prospectus divided by expected investment
D_EF	D_EF is a dummy that is assigned the value 1 when the IPO proceeds exceed the expected investment and 0 otherwise
C. Investment inefficiency
INV	Investment denotes total capital expenditure divided by total assets
Cash	Cash denotes the cash flow divided by total assets
Size	Size is the natural logarithm of year-end total assets
Liq	Liquidity denotes current assets divided by total assets
Wedge	Wedge is the difference of controlling owner’s voting rights from his/her cash flow rights
Own_Con	Ownership concentration denotes the top five shareholders’ shareholdings
Compensation	Compensation denotes the natural logarithm of the top 3 managers’ compensation
Ind_Dir	Independent directors denote the percentage of independent directors
Sales_Gr	Sales growth denotes the sales growth rate in two consecutive quarters
EPS	EPS denotes the earnings per share
D. Investor sentiment
INV	Investment denotes total capital expenditure divided by total assets
Plot/PR	Plot/PR the pre-IPO market closing price on IPO lottery day divided by the fundamental value
E. Underwriting and firm’s characteristics
IR	Initial return is defined as the difference between initial price and offer price divided by offer price
Age	Age denotes the number of years from inception to IPO
Expense	Expense denotes the total issuing fees divided by number of shares financing
Board Size	Board Size denotes the number of directors and supervisors
Tobin’s Q	Tobin’s Q is defined as the sum of the market value of equity and the book value of liabilities divided by the book value of assets
Duration	The number of days from IPO announcement to listing
Turnover	Turnover is the turnover rate of the listing day
Real_Issue	Real_Issue denotes the real issuing shares divided by the maximum number of issuing shares in prospectus
PEoffer/1 yr	PEoffer/1 yr is the ratio of PE in offering to the PE one-year post IPO
PEIPO/1 yr	PEIPO/1 yr is the ratio of PE in the first listing day to the PE one-year post IPO
SOES	SOES is a dummy that is assigned the value 1 when the IPO firm is a state-owned enterprise and 0 otherwise
Exchange	Exchange is a dummy that is assigned the value 1 when the IPO firm is listed in Shanghai Stock Exchange and 0 when the IPO firm is listed in Shenzhen Stock Exchange
GEM	GEM is a dummy that is assigned the value 1 when the IPO firm is listed in Growth Enterprise Exchange and 0 otherwise
Hot_M	Hot_M is a dummy that is assigned the value 1 when the new issue market is a hot market and 0 otherwise. The identification of hot and cold markets is referred to Ibboston and Jaffe (1975) and Ritter (1984). Specifically, we identify a hot market as when the number of new issues is higher than sample median (8) and the average initial return is higher than sample median (90.2%)