The underpricing and long-term performance of Chinese IPOs listed on the Hong Kong exchange

Purpose – This study aims to inform prospective listing firms, investors and regulators of the unique drivers of Chinese initial public offering (IPO) pricing on the Hong Kong Exchange. Design/methodology/approach – Using a hand-collected IPO dataset, we investigate whether information uncertainty or investor exuberance drives underpricing and Chinese IPOs ’ performance from 2002 to 2015, including 114 state-owned enterprises (SOEs). Findings – Contrasting with the “ listing bubble ” in the China domestic stock market, generated by the overoptimism of retail investors, we highlight a “ placing bubble ” among Chinese firms listed in Hong Kong. This is driven by institutional investors ’ buoyant demand for Chinese IPO shares, particularly those of SOEs. Chinese listing firms employ discreet earnings management strategies with their working capital accounts to smooth pre-IPO earnings, which becomes apparent to the market only in the long term. Originality/value – This study is the first to examine the pricing of sought-after Chinese IPOs among international investors, who face various restrictions when investing in the Chinese domestic stock market. Additionally, it is the first study to measure earnings management using hand-collected pre-IPO data in IPO underpricing studies.


Introduction
Hong Kong is a sought-after listing destination for Chinese firms seeking equity capital because of its status as a global financial hub with strong institutions and a sophisticated global investor base.Over half of initial public offering (IPO) applicants in Hong Kong are Chinese firms [1].Amid increasing geopolitical tensions between China and the USA, Hong Kong has become even more crucial for Chinese firms raising equity capital from international investors (Qian et al., 2024).However, when Chinese companies seek listings in Hong Kong, they are subject to the scrutiny of an equity market dominated by sophisticated institutional investors and more stringent disclosure rules than the domestic market.
Information does not equally disseminate among corporate insiders, underwriters and investors in IPOs, leading to mis-valuation during pricing (Ljungqvist, 2007).This phenomenon is exacerbated by the imbalance between private and public information of IPO firms (Easley et al., 2002).Investors tend to overestimate an IPO firm's future earnings based on pre-listing disclosures, which can be manipulated by managers for personal gains (Teoh et al., 1998a;Nagata, 2013).Chinese IPOs on the domestic stock market exhibit the world's highest underpricing, averaging 120% on the first trading day owing to opaque earnings, aggressive earnings management practices (Boulton et al., 2011) and overoptimistic expectations of retail investors in the A-share market (Kimbro, 2005;Shen et al., 2014).Aggressive earnings management is prevalent among IPO firms in mainland China.
Empirical evidence supporting the "investor exuberance theory" in the Chinese domestic market and other major developed markets is well documented (Guo et al., 2011;Roosenboom et al., 2003;Teoh et al., 1998a, b).Qian et al. (2024) attribute the high Chinese IPO underpricing or "money on the table" phenomenon to the "pricing caps" imposed by the stock market regulator.This causes retail investors to frantically bid for IPO shares at the start of trading, leading to high initial trading day returns or IPO underpricing.Financial market friction resulting from regulatory intervention is a major driver of Chinese IPO underpricing in the domestic stock market.
Given Hong Kong's efficient institutions (particularly the presumed absence of Chinese regulatory intervention in the IPO process) and sophisticated institutional investors pricing IPOs, we expect that the "information uncertainty theory" prevails in the Hong Kong equity market when Chinese companies are listed.Institutional investors account for the largest share of Chinese IPO offerings in Hong Kong (Liu, 2019).Embedded information uncertainty manifested through Chinese listing firms' earnings management practices is positively associated with IPO underpricing in the Hong Kong equity market.Sophisticated institutional investors dominating the pricing and trading of IPOs in Hong Kong are believed to possess superior resources and abilities to digest the notoriously opaque accounting information released by Chinese listing firms.Therefore, they are expected to submit bids reflecting the degree of uncertainty associated with reported earnings (Gao et al., 2017;Nagata, 2013).When IPO shares begin to trade in the aftermarket, institutional investors are compensated to bear informational risk.Therefore, we expect institutional investors in Hong Kong to see through masked earnings and bid prudently based on their assessment of the information uncertainty associated with these Chinese IPO firms.
Surprisingly, we find that Chinese IPO underpricing is, on average, 9% in Hong Kong (very moderate compared to over 100% underpricing of IPOs in the Chinese domestic market) and is not significantly related to earnings quality.Thus, it appears that institutional investors in Hong Kong do not price information uncertainty related to Chinese IPOs.Interestingly, we document a "placing bubble" of the Chinese IPOs that is followed by an average of 40% value loss at the Hong Kong Exchange.This phenomenon is likely driven by institutional investors' excessive demand for sought-after investment opportunities that are otherwise difficult for foreign investors to access.Anecdotal evidence points to indirect intervention from the Chinese equity market regulator, the China Securities Regulatory Commission (CSRC), in Chinese IPO pricing, which tends to elevate the offer price in Hong Kong.This is distinctive from the "listing bubble" documented in the Chinese domestic market resulting from regulatory inefficiency (such as pricing caps) and retail investor exuberance (Gao et al., 2015(Gao et al., , 2017;;Qian et al., 2024).
Although the investor base in the Hong Kong stock market is dominated by institutional investors and sophisticated investors with a global presence [2], Chinese IPOs are still subject to overdemand or the "hot issue" problem.A typical Mainboard listing on the Hong Kong Exchange has two tranches: the "placing" tranche available to institutional investors and the "public offer" tranche available to Hong Kong retail investors.The listing documents available at the Hong Kong Exchange News website do not disclose full information of the "placing" tranche.However, we collected the percentage of overallotment options exercised by Chinese IPO firms to measure the extent of interest from institutional investors in the book-building process.Out of a sample of 177 Chinese IPOs, 120 listing firms exercise the option of issuing up to 15% of the total shares available to meet the high demand of institutional investors.In the "public offer" tranche, the average oversubscription ratio (the ratio of all valid IPO share subscriptions over the total shares available to retail investors) of the sample is 152 times.We conjecture that both institutional investors and retail investors in the Hong Kong stock market may contribute to the underpricing of Chinese IPO shares owing to the supply-demand imbalance in both "placing" and "public offer" tranches.This imbalance, coupled with the indirect influence of Chinese stock market regulators, leads to an elevated offer price for Chinese IPOs.
To capture the earnings manipulation of IPO firms when managers' incentives and potential gains from such practices are heightened, we manually collect financial data from prospectuses to construct multiple earnings quality measures.Therefore, we can avoid the mismeasurement problem highlighted by Armstrong et al. (2016), suggesting that IPO year discretionary accruals largely reflect working capital changes when IPO firms deploy newly raised equity capital instead of a result of managers' earnings manipulation.We also shed light on how reputable underwriters and auditors may assist Chinese IPO firms in mitigating accounting and information uncertainty issues, which are highly sensitive to IPO pricing (Beatty, 1989;Boulton et al., 2011;Dong et al., 2011).
The Chinese IPO sample comprises 177 nonfinancial listings on the Hong Kong Exchange from 2002 to 2015, covering eight industry sectors, with the highest industry representation in consumer goods and industries.The average issue size is HK$3.9bn.Nearly 64% of the sample in the study period are state-owned enterprises (SOEs), which are larger and more profitable than non-SOEs (statistics are not reported for brevity).According to Huang and Yeung (2013), SOEs listed on the Hong Kong stock market take political assignments to encourage structural reforms in the Chinese domestic market and further mitigate the historical inefficiencies associated with political corruption.It is unclear how these "window companies" highly regarded by the Chinese government perform post-listing in Hong Kong.
Distinct from prior studies on SOE IPO underperformance (Gu, 2003), our analysis reveals that Chinese SOEs overperform by 7.5% than non-SOE IPO firms when industry, firm with issue charateristics, and market factors are controlled.However, the favorable response of investors to the initial listing of SOEs has diminished over the long term.We find no conclusive evidence that SOEs differ significantly from non-SOE firms in terms of long-run performance.
Existing literature studying Chinese IPOs either focus on the domestic market with substantial regulatory restrictions and with limited international investor presence (Gu, 2003;Chen et al., 2004Chen et al., , 2018;;Fan et al., 2007;Gao et al., 2015Gao et al., , 2017;;Qian et al., 2024) or a small cohort of listed companies in Hong Kong such as "H shares" (Chong et al., 2010;Fung et al., 2013).How Chinese IPOs listed in Hong Kong perform as a unique listing cohort has not attracted the much-deserved research attention, despite the increasing importance of Hong Kong being one of the most accessible international capital markets for Chinese companies (Qian et al., 2024).
Therefore, this study examines how the pricing of Chinese IPOs in Hong Kong is different from the Chinese domestic stock market.The unique cohort of Chinese IPO firms listed on the Hong Kong Exchange is dominated by state-owned enterprises (SOEs), giving rise to a rare opportunity to examine whether their natural competitive advantages translate into longterm value creation when aided by more advanced institutions in the Hong Kong equity market.
This study contributes to the existing literature in several ways.We are the first to document the "placing bubble" of Chinese IPOs, which is caused by institutional investors' excessive demand boosting the offer price, resulting in a moderate level of initial trading day returns in the Hong Kong equity market.The findings of this study complement the IPO underpricing literature by disclosing fresh evidence of how IPO underpricing may be caused

Journal of Asian
Business and Economic Studies by financial market frictions other than "investor exuberance" or "information uncertainty": constrained investment opportunities and implicit regulatory intervention in an equity market with strong institutions.We show that the true earnings of Chinese IPO firms are challenging to understand in the market, even under the rigorous disclosure requirements of the Hong Kong Exchange.Only the manipulation of trade receivables, a form of "premature revenue recognition" (Stubben, 2010), is incorporated in the IPO underpricing.Other forms of earnings management, such as the manipulation of working capital accounts (e.g.delayed recognition of expenses) and non-current accounts (e.g.deceleration of depreciation), are executed in a discreet manner.Earnings management practices tend to smooth earnings instead of inflating them.Our findings provide insights to financial market regulators to strengthen IPO disclosure requirements.
The remainder of this paper is organized as follows: Section 2 provides an overview of the related literature and develops the hypotheses; Section 3 describes the sample and the construction of key variables; Section 4 presents the descriptive statistics of the sample and important variables; Section 5 reports the results of the empirical analysis and discusses their implications and Section 6 concludes the paper.

Related literature and hypotheses
Chinese IPOs listed in the domestic market exhibit the world's highest underpricing, averaging 172% on the first trading day (Qian et al., 2024).This is attributed to inefficient regulatory interventions, such as pricing caps and restricted IPO share supply, imposed by the stock market regulator CSRC (Chen et al., 2018;Tian, 2011;Qian et al., 2024).Additionally, research has suggested that unsophisticated retail investors in the Chinese domestic market tend to be overly optimistic when bidding for newly traded IPO shares in the aftermarket (Boulton et al., 2011;Gao et al., 2015;Geng et al., 2010;Kimbro, 2005;Shen et al., 2014).Moreover, Agarwal et al. (2008) have discovered that investors' excessive demand is associated with higher IPO underpricing in the Hong Kong equity market due to overoptimism in "hot issues".Beatty and Ritter (1986) argue that underpricing is positively associated with ex ante uncertainty related to IPO firm valuation.Studies supporting this argument have been recorded in several country and cross-country studies (Boulton et al., 2011;Chahine et al., 2012;Nagata, 2013).Particularly, Gao et al. (2017) reveal that institutional investors' bidding price is negatively associated with pre-IPO accrual-based earnings management in the Chinese domestic market.They argue that sophisticated institutional investors can detect IPO firms' earnings management, thereby submitting a lower bidding price.
A typical Mainboard IPO in Hong Kong comprises two tranches: a "placing" tranche for institutional investors and a "public offer" tranche for retail investors.Initially, 90% of shares are allocated to the "placing" tranche, with the remaining 10% allocated to the "public offer" tranche.When the "public offer" tranche is significantly oversubscribed, shares from the "placing" tranche are reallocated to meet retail investors' high demand, following a scale prescribed by the Listing Rules or approved by the Hong Kong Exchange [3].For instance, when the oversubscription ratio of the IPO is 100 times or more than the originally allocated shares, the "public offer" tranche will increase up to 50% of all shares on offer.The excess demand from institutional investors is met with an over-allotment option of up to 15% additional shares.The IPO offer price is determined through the book-building process of the "placing" tranche based on institutional investors' interest expressed.
Given that the Hong Kong IPO offer price is largely determined by well-informed institutional investors, we expect that they can discern the manipulated earnings of Chinese listing firms by incorporating this uncertainty into the offer price.Thus, we propose the following hypothesis.
H1. Chinese IPO firms' pre-IPO earnings quality is positively associated with IPO underpricing when the listing firm, deal characteristics and market conditions are controlled.
Chinese IPOs are enthusiastically received by both institutional and retail investors in Hong Kong.Evidence includes nearly 70% of Chinese IPO firms exercising over-allotment options, with a 152-fold oversubscription ratio for the "public offer" tranche.We expect that supplydemand imbalances of Chinese IPO shares, particularly those of SOEs, will contribute to higher IPO underpricing.Therefore, we propose the following hypothesis.
H2. Excessive demand of Chinese IPO shares is positively associated with IPO underpricing when the listing firm, deal characteristics and market conditions are controlled.
The underpricing of Chinese SOE IPO firms caused by political and managerial inefficiencies is well documented, as investors demand compensation for bearing agency costs of holding SOE shares (Chen et al., 2004(Chen et al., , 2015;;Fan et al., 2007).Chen et al. (2008) attribute this phenomenon to politically connected managers' intentions of transferring shares to related parties who control their political promotions.
Chinese SOEs seeking listing in Hong Kong must secure CSRC approval for their offer prices through "guidance" (communications between listing firms, their underwriters and the CSRC, not publicly available).According to anecdotal revelations from former investment bankers (who cannot be quoted on the record), the CSRC only approves offer prices when they are not deemed underpriced, thus leading to wealth transfers to foreign investors post-listing, particularly for SOEs.Such regulatory intervention by the CSRC is distinct from the pricing caps placed on domestic IPO shares, as revealed by Qian et al. (2024).Therefore, we believe that Chinese IPOs, particularly SOEs, are priced at or above their intrinsic values in the "placing" tranche under the "guidance" of the CSRC.The average PE ratio of our sample is approximately 32 times (38 times for SOEs and 25 times for non-SOEs) from 2002 to 2015, which is significantly higher than the various "pricing caps" placed on domestic IPOs documented by Qian et al. (2024).Institutional and retail investors in Hong Kong are inclined to accept the inflated offer price approved by the CSRC, given the otherwise restricted access to the investment opportunities.We expect this may lead to higher IPO underpricing for Chinese SOEs when shares begin to trade in the open market.Therefore, we posit that state ownership of Chinese IPO firms is positively related to underpricing or the first trading day returns, owing to excessive demand from both institutional and retail investors in Hong Kong.
H3. Chinese SOEs have higher IPO underpricing (or initial trading day returns) than non-SOE IPO firms listed on the Hong Kong Exchange when the listing firm, deal characteristics and market conditions are controlled.Roosenboom et al. (2003) and Teoh et al. (1998a) show that IPO firms with abnormally high discretionary accruals in the pre-IPO year consistently underperform in the three-year period post-IPO.They argue that investors gradually absorb information regarding the actual value of issuing firms from released financial reports, analyst analyses and the media.Consequently, investors' overoptimism fades as initial expectations of high earnings disappear, resulting in diminishing positive abnormal returns over time.We expect the "placing bubble" of Chinese IPOs listed in Hong Kong to burst over time as the market gains a better understanding of their actual earnings.Therefore, we propose the following hypothesis: H4. Chinese IPO firms' pre-IPO earnings quality is negatively associated with IPO performance in the long term, when the listing firm, deal characteristics and market conditions are controlled.

Journal of Asian Business and Economic Studies
Chinese SOEs' performance post-listing is primarily examined in the domestic stock market.Gu (2003) records substantial underperformance in the early 1990s among firms with higher state ownership.Fan et al. (2007) document substantial underperformance in partially privatized listing firms with politically connected CEOs, reflecting managerial inefficiencies caused by expected state intervention.While the Chinese central government's policy favors large SOEs (Qian et al., 2024), long-term outperformance against their non-SOE counterparts is not expected despite these policies.
H5. Chinese SOE IPOs listed on the Hong Kong Exchange do not outperform their non-SOE counterparts when the listing firm, deal characteristics and market conditions are controlled.
Some studies, such as Benveniste and Spindt (1989), Carter et al. (1998) and Dong et al. (2011), attribute IPO underpricing to underwriters, suggesting that reputable investment bankers facilitate accurate information disclosure and promote IPO issues effectively.In our analysis, we include proxy variables for both the IPO underwriter and auditor reputations of Chinese listing firms to investigate the efficacy of external parties in reducing IPO underpricing.

Methodology
We identified 263 IPOs listed on the Hong Kong Exchange Mainboard from 2002 to 2015 in the China Stock Market and Accounting Research database (CSMAR).Financial firms and IPOs with "back-door," "transfer," "introduction" and those on the Growth Enterprise Board (GEM) are excluded, as these listings are not subject to the same rules as the Mainboard.Our final sample comprises 177 non-financial Chinese IPOs.
The data required for our analysis were collected from three sources: the Hong Kong Exchange News, the CSMAR and the Wind database.We manually collected pre-IPO accounting information from the digital prospectus available on the Hong Kong Exchange News website.
We use the following ordinary least squares models to estimate the cross-sectional regressions for underpricing and long-term post-IPO performance.We correct the clustering by issue year and use heteroscedasticity-consistent standard errors for the inferences.
where UP i denotes the underpricing of IPO firm i; BHAR i denotes buy-and-hold abnormal returns of IPO firm i; EQ i denotes the vector of earnings quality proxies of IPO firm i; SOE i denotes the ownership status of the IPO firm i, which is equal to one if the issuer is an SOE and zero otherwise; ðControl UP j Þ i and ðControl LR j Þ i denote vectors of control variables for underpricing estimation and long-term post-IPO performance estimation, respectively; and ε denotes the error term.The state ownership of Chinese IPO firms is determined by the listed share class and ownership information from the WIND database.
An IPO is underpriced if the first trading day's closing price exceeds its pre-determined offer price, which is measured as follows: where UP i denotes the underpricing of IPO firm i on its initial listing day; P i1 denotes the closing price of IPO firm i at the end of the first trading day; and P i0 denotes the offer price of IPO firm i.
We adopt a long-term event study (Ritter, 1991) to construct the BHAR for assessing post-IPO performance over 12-, 24-and 36-month event windows, following prior literature (Gao et al., 2015;Roosenboom et al., 2003;Shen et al., 2014;Teoh et al., 1998a).Additionally, cumulative abnormal returns (CAR) are calculated as a robustness check.Generally, IPO research utilizes a composite market index as a benchmark portfolio (Ritter, 1991;Teoh et al., 1998a, b).The broad market index Hang Seng Index (HSI) may be a misspecified market portfolio for pricing Chinese IPOs.Therefore, we employ the Hang Seng China Enterprise Index (HSCEI) as the market portfolio in calculating share performance.HSI is employed in our robustness check.
We construct various earnings management measures for our analysis, following Teoh et al. (1998a) and Dechow et al. (1995), to establish an industry-matching group for each IPO firm in the year before listing.We estimate earnings management variables using coefficients derived from the cross-sectional regressions.Four discretionary accrual variables are constructed in this manner: Discretionary Total Accruals (DTA), Discretionary Current Accruals (DCA), Discretionary Noncurrent Accruals (DNA) and Absolute Discretionary Total Accruals (ADTA).These discretionary accruals are expected to capture earnings manipulations in working capital accounts (Teoh et al., 1998a).Additionally, we create several alternative measures to capture operating cash flow and receivables account manipulations: Cash Flow Accruals (CFA) (Dechow and Dichev, 2002), Discretionary Cash Flow (DCF) (Francis et al., 2005) and Discretionary Revenue (DREV) (Stubben, 2010).This process and the definitions of all variables are outlined in the Supplementary Online Appendix.

Descriptive statistics
Table A1 [4] presents the sample analysis and variables used to estimate Equation (1).Chinese IPO underpricing averages approximately 9%, notably lower than other markets.For example, IPO underpricing in Japan is 45% (Nagata, 2013), 14% in the USA (Ritter, 1991), 172% in the Chinese domestic market (Qian et al., 2024), and 96% in Chinese small and medium-sized enterprise (SME) IPOs (Gao et al., 2015).The significantly low level of discretionary accruals among Chinese IPOs listed in Hong Kong (averaging 1-2%) can be attributed to effective monitoring by regulators, underwriters and auditors during the pre-IPO disclosure process.However, the significantly higher level of ADTA of 8.9% suggests that the considerable swing in earnings management practices in both directions may not be captured by discretionary accruals.
Table A2 [4] summarizes the key variables in the long-term IPO performance analysis.Nine IPO firms are excluded owing to delisting in the long-term event window.Chinese IPOs exhibit an average BHAR of À47% three years post-listing.The average three-year CAR for the IPO sample is at a similar level of À45%.
We plot the average monthly returns, BHAR and CAR of the sample over each event month post-listing in Figure A1 [4].All returns are statistically significant except in the first month.Subsequently, we divide the sample by state ownership and plot their average BHAR in Figure A2 [4] and CAR in Figure A3 [4], respectively.SOEs outperform non-SOEs post-IPO if performance is measured in CAR, as shown in Figure A3 [4].The difference is statistically significant.

Chinese IPO underpricing
The regression results of Equation ( 1) are presented in Columns ( 1) to (4) Table A3 [4].Surprisingly, discretionary accruals and discretionary cash flow variables do not predict IPO

Journal of Asian
Business and Economic Studies underpricing (results are not tabulated for brevity).Only easily detectable Discretionary Revenue (a measure of premature revenue recognition) is priced in IPO underpricing, as shown in Column (1), with a negative coefficient indicating that such earnings uncertainty is priced in aftermarket trading rather than institutional investors' offer price.To access the rare investment opportunities of Chinese IPOs, institutional investors in Hong Kong do not discount prices during the book-building process based on their assessment of listing firms' earnings uncertainty.Another surprising outcome is that SOEs outperform non-SOEs by over 7.5% when issuer and issue characteristics and market sentiment are controlled (Column (1) Table A3 [4]), aligning with the earlier analysis of the excessive demand for Chinese SOE IPOs by both institutional and retail investors in Hong Kong.
To understand the sought-after traits in Chinese IPOs by Hong Kong investors, we regress the Oversubscription Ratio and Overallotment Exercised variables against the issuer and issue characteristic variables (results are not tabulated for brevity).The demand for IPO shares is stronger (or overallotment option is more likely to be exercised) for IPO firms with stronger revenue growth pre-listing and more IPO shares on offer.Conversely, demand is lower (or the overallotment option is less likely to be exercised) for issuing firms with negative pre-listing operating cash flows and dual listings on other stock exchanges.The availability of listing firms' public equity is related to the demand for Chinese IPO shares and thus the pricing of Chinese listing firms in Hong Kong.Interestingly, the demand for Chinese SOE IPOs is more than 260 times higher than that for non-SOE IPOs when other factors are controlled for [5].We suggest this excessive demand likely contributes to IPO underpricing in our analysis.
In Columns ( 2) and ( 3) of Table A3 [4], we estimate Equation ( 1) using the variables measuring the supply-demand imbalance of Chinese IPOs.Both the Oversubscription Ratio and the Overallotment Exercised dummy variable are positive and statistically significant in explaining IPO underpricing.
In Column (4), we include the interaction term between the SOE dummy variable and the dummy variables representing underwriter reputation to examine the different marginal effects of external parties on IPO underpricing.The negative and significant coefficient estimates show that SOEs benefit more than non-SOEs when employing reputable underwriters in their IPO process to reduce IPO underpricing or the discount institutional investors apply to the offer price based on their assessment of listing firms' earnings uncertainty.This finding validates the use of reputable underwriters to reduce information uncertainty in presumably opaque Chinese listing firms.

Post-IPO long-term stock performance
We estimate Equation ( 2) and report the results in Columns ( 5) to (8) Table A3 [4] to examine whether Hong Kong investors understand Chinese firms' earnings manipulation.Surprisingly, only ADTA is priced into long-term performance (the results of other earnings quality measures are not tabulated for brevity).Larger absolute discretionary accruals imply aggressive managerial efforts to smoothen company earnings, suggesting lower earnings quality or higher information uncertainty.This finding suggests that working capital and cash flow account manipulation is discreet and challenging for investors to detect.Markets' understanding of Chinese listing firms' manipulation of financial accounts takes up to three years post-IPO.A one standard deviation increase in ADTA is associated with a 9% decrease in BHAR over the 36-month event window post-IPO [6] when other influential covariates are controlled.
In Section 5.1, we predict that excessive demand for Chinese IPO shares primarily drives initial trading day returns, thus causing overoptimistic valuations by secondary market investors.IPO share value is expected to revert to its intrinsic value in the long term.However, we do not find evidence to support the prediction that the long-term loss of Chinese IPO shares is a reversal of the overoptimism of the market at IPO, since the coefficient estimates of Market Adjusted Underpricing (MUP) and PE Ratio are not statistically significant.This finding contrasts with prior evidence (Teoh et al., 1998a) suggesting that the higher the underpricing level, the worse the after-market long-term performance, as investors reverse their belief in the true earnings ability of IPO firms to a reasonable level.
The coefficient estimates of SOE in the regression analysis of Columns ( 5) and ( 6) Table A3 [4] shift signs and show weak statistical significance.Positive expectations of IPO investors in SOEs do not translate into long-term value creation.However, we find a negative and significant coefficient estimate of the interaction term between SOE and the reputation proxy for their auditors, which is equivalent to a positive significant coefficient estimate if the non-SOE dummy variable is employed.This indicates that non-SOE firms with reputable auditors tend to perform better in the long term, given similar earnings uncertainty (proxied by ADTA) when other covariates are controlled.

Robustness analysis
We employ various long-term performance measures to estimate Equation ( 2) for robustness: 36-month CAR, 12-and 24-month BHAR, and the calculation of abnormal returns using HSI instead of the HSCEI.Our results remain economically and statistically similar.To address concerns about the Global Financial Crisis distorting the IPO market and thus underpricing, we re-estimate all regression analyses with a smaller sample excluding the IPOs listed during the crisis period of 2007-2008 (11 IPOs are excluded) (11 IPOs are excluded).All results hold, and our predictions of Chinese IPO underpricing and long-term performance do not change.

Conclusions
To fill in the gap in the literature of Chinese IPO underpricing and performance, we study a unique cohort of Chinese firms listed on the Hong Kong Exchange with strong institutions and stringent disclosure requirements.We observe a "placing bubble" phenomenon when Chinese firms list their shares on the Hong Kong Exchange, which bursts over the long term.IPOs are priced close to the aftermarket equilibrium price, driven by high demand from Hong Kong investors and the regulatory influence of the Chinese stock market regulator, boosting the valuation of Chinese listing firms, particularly SOEs.Contrary to the prudent discounted price that institutional investors place in the book-building process of Chinese domestic IPOs, institutional investors in Hong Kong must accept the elevated "placing" price approved by the CSRC to secure investment opportunities, which are otherwise difficult to access as foreign investors.We disclose new findings on how financial market inefficiencies, such as constrained investment opportunities and implicit regulatory intervention, have a longlasting impact on IPO pricing in an equity market with strong institutions.Surprisingly, Chinese companies' massive value loss of over 40% three years post-IPO does not deter international investors' enthusiasm.It is unlikely that sophisticated institutional investors in Hong Kong consistently overvalue Chinese IPO firms.We suggest that the ownership of Chinese IPO firms, particularly of SOEs by international investors, is driven by strategic reasons rather than by expected returns over multiple years.Further research into how institutional investors in Hong Kong participate in the pricing of Chinese IPOs would provide more informative insights into this unique market dynamics.