financing decisions after partial privatization in china: can a stock market quotation really...

J. Finan. Intermediation xxx (2013) xxx–xxx

Contents lists available at ScienceDirect

J. Finan. Intermediation

journal homepage: www.elsevier .com/locate/ jfi

Financing decisions after partial privatizationin China: Can a stock market quotation reallyprovide discipline?

1042-9573/$ - see front matter � 2013 Elsevier Inc. All rights reserved.http://dx.doi.org/10.1016/j.jfi.2013.08.005

⇑ Corresponding author. Address: Naamsestraat 69, 3000 Leuven, Belgium. Fax: +32 16 32 66 83.E-mail address: [email protected] (N. Huyghebaert).

1 It was not until 2000 that private enterprises started to float their shares on a more regular basis. Not surprisingly,privatized SOEs still account for a very large proportion of listed firms in Mainland China (Huyghebaert and Quan, 200

Please cite this article in press as: Huyghebaert, N., et al. J. Finan. Intermediation (2013), http://dx.d10.1016/j.jfi.2013.08.005

Nancy Huyghebaert a,⇑, Qi Quan b, Lijian Sun b

a Katholieke Universiteit Leuven, Belgiumb School of Economics, Fudan University, PR China

a r t i c l e i n f o a b s t r a c t

Article history:Received 15 April 2010Available online xxxx

Keywords:PrivatizationCapital structureAsymmetric informationDefault riskGovernance

This paper uses the framework developed by Helwege and Liang(1996) to investigate the post-listing financing decisions of 221Chinese state-owned enterprises that were partially privatizedvia the stock market in 1994–1999. First, we estimate a logit paneldata model to examine the decision to raise external funds, eitherbank loans or equity. Our results show that the probability of rais-ing external finance is positively correlated with the firm’s cashshortage, but only if non-cash working capital is included in thedefinition of the deficit. In contrast, firms with a cash surplus areless likely to tap external financial markets. Second, we examinethe type of security issued and find that default risk negativelyaffects the probability of completing a stock offering, while firm-level information asymmetries are not significant. Overall, we doc-ument some important differences between the firms where thegovernment retained its majority stake and those where it relin-quished it; we link these to differences in the firms’ governanceand access to financial markets.

� 2013 Elsevier Inc. All rights reserved.

1. Introduction

In the case of China, the IPO market has been largely dominated by state-owned enterprises (SOEs)that have undertaken a share issuing privatization (SIP).1 The basic idea behind these SIPs, it has been

partially9).

oi.org/

http://dx.doi.org/10.1016/j.jfi.2013.08.005

mailto:[email protected]


http://www.sciencedirect.com/science/journal/10429573

http://www.elsevier.com/locate/jfi



2 N. Huyghebaert et al. / J. Finan. Intermediation xxx (2013) xxx–xxx

claimed, was to subject the corresponding SOEs to the discipline of the stock market, as share priceshenceforth would reflect information on managerial decisions and firm performance. Examining thevalidity of this argument is interesting, as the risk of protracted political influence is non-trivial in Chi-nese listed SOEs. The government has indeed retained a majority stake in many of these firms, reducingits ownership stake only partially at SIP-time by allowing SOEs to sell primary, i.e. newly issued, shares toChinese domestic investors in the A share market.2

Prior research has investigated the influence of stock market quotation on the financial and oper-ating performance of Chinese SOEs, concluding that, unlike other countries, SIPs in China have had adetrimental effect on firm performance (e.g., Sun and Tong, 2003; Wei et al., 2005; Jiang et al., 2009).The underlying reasons for this decline in performance remain unsettled to date. Yet, on average per-formance deterioration was less pronounced when the government relinquished its majority stake,thereby suggesting that more was going on than just an increased exposure to industry competitionforcing SOE profit margins down over time. In this paper, we wish to gain more insights into the ques-tion whether a stock market quotation by itself can perform such a disciplining function, by examiningwhether economic forces became a fundamental driving force of the post-listing financing decisions ofChinese SOEs. We examine this research question for a sample of 221 Chinese SOEs after their listingon the Shanghai stock exchange in the period 1994–1999. A methodological contribution of our articleis that we examine incremental financing choices in the first five years after SIP. Studying incrementalfinancial decisions rather than capital structure, i.e. the cumulative cross-sectional outcome of priorfinancing choices, is instructive as the overall debt ratio of listed SOEs does not fluctuate much overtime and so conceals the huge number of transactions in debt as well as in equity markets. For thispurpose, we use the framework developed by Helwege and Liang (1996). As such, we first analyzewhat firm-specific financial factors drive the decision to raise external finance, whether in the formof bank debt or equity. Next, we examine the type of security issued. Our main contribution is thatwe consider the possibility that these two distinctive aspects of post-SIP financing decisions may beinfluenced by differences in ownership structure across partially privatized SOEs.

First, we investigate whether the decision to raise additional funds after listing is driven by a shortageof cash to finance investment projects, as is implied by the pecking order theory of capital structure (see,for example, Helwege and Liang, 1996; Shyam-Sunder and Myers, 1999; Franks and Goyal, 2003). Thisresearch question is worthwhile to explore, as the size of the primary offering in Chinese SIPs during the1990s was not determined by the need to finance investment outlays (see Sun and Tong, 2003; Bai et al.,2004; Huyghebaert and Quan, 2009). Instead, SOEs that relied heavily on state financing through loansfrom state-owned banks, through trade credit from other SOEs, and/or through subsidies, raised moreequity at their SIP. Taking these firm-level financing decisions at the occasion of the SIP as our controlgroup, we evaluate the effect of stock market quotation. Conversely, listed SOEs in which the Chinesestate remained a dominant owner after SIP may have continued to pursue political objectives, such asoffering low prices to consumers, guaranteeing employment, etc. Such expropriation of minority inves-tors could be accomplished rather easily, as property rights are poorly protected in China (Cull and Xu,2003, 2005; Allen et al., 2005; Li et al., 2008). Nonetheless, realizing political goals is expensive and, thus,may cause a shortage of cash from operations for the listed SOEs that pursue them. The latter firms couldnow also use their stock market access to raise extra funds in order to meet their extensive financial obli-gations from business operations. If this scenario arises, it is likely to apply especially to the partiallyprivatized SOEs in which the State retained its majority stake after SIP.

Second, given the decision to access external financial markets, we examine the subsequent choicebetween borrowing from banks and issuing new shares. Corporate bonds were not a viable financingalternative during our sample period. In general, two categories of firm characteristics can explain thesecurity choice. While we agree that information asymmetries between firms and investors over the va-lue of the firm should lead to a preference for debt over equity in the case of partially privatized SOEs, as isimplied by the models developed by Myers (1984) and Myers and Majluf (1984), we further contend thatvariables capturing the firm’s default risk may be negatively associated with the probability of share

2 In this paper, we use the term ‘government’ in a general meaning, including the central as well as the various localgovernments with the power to make strategic decisions in the firm.

Please cite this article in press as: Huyghebaert, N., et al. J. Finan. Intermediation (2013), http://dx.doi.org/10.1016/j.jfi.2013.08.005



N. Huyghebaert et al. / J. Finan. Intermediation xxx (2013) xxx–xxx 3

issuance. The reason is that stock prices will incorporate the firm’s poor current performance and pos-sibly also its bad future prospects. As raising additional equity can happen only at relatively low prices,issuing new shares might considerably dilute the government’s ownership stake. Chinese authoritiesmay be averse to this idea, particularly when they wish to maintain control over firms, for example topursue political objectives. In this case, borrowing from banks could be a better alternative. In China, un-like western economies, bank credit may still be available given the institutional bias of the state-ownedbanking system towards SOEs (Cull et al., 2009; Li et al., 2009). Moreover, these borrowing SOEs may fig-ure that a subsequent default on bank debt will not automatically trigger bankruptcy, as state-ownedbanks tend to bail them out. So, in contrast to the classical models developed by Myers (1984) and Myersand Majluf (1984), we conjecture that Chinese listed SOEs exhibiting high default risk may prefer to bor-row from banks, particularly when the Chinese state hangs on to majority control.

Our empirical results show that after accounting for industry and year fixed effects, the cash short-age between investments in fixed assets and internally available cash is not correlated with the prob-ability of raising external funds. We do find a significant positive relation with the external financedecision once the change in non-cash working capital is counted as part of the investment outlays.This relation turns out to be driven by the increase in trade credit (accounts receivable) granted tofirms with which our sample SOEs trade. Arguably, these listed SOEs seem to use their better accessto external finance to ease the financial constraints of their trading partners. In line with our interpre-tation that this result reflects poor financial market development rather than political interference, wedetect that the positive effect of this cash shortage variable arises regardless of whether the State re-tained its majority stake after SIP. By contrast, firms with a cash surplus are less inclined to collect ex-tra finance, especially bank loans, but only when the firm remains majority-owned by thegovernment. Possibly, these firms take into account their better access to loans from state-ownedbanks whenever such loans are needed. Next, the interaction term between the deficit variable, if po-sitive, and the ratio of cash flow to total assets is significantly positive, indicating that partially priv-atized SOEs with cash shortages arising from investments but poor performance have troubleobtaining additional finance. Interestingly, this effect only arises for equity offerings by firms in whichthe State has given up its 50% stake, suggesting once more that access to financial markets dependsupon the firm’s ownership structure. Overall, our results do not support the idea that especially listedSOEs running out of cash raise funds to meet their financial obligations from business operations. Infact, loss-making SOEs are less inclined to tap external financial markets. The latter result appears tobe driven by the firms that remain majority-owned by the Chinese government.

As to the type of security issued, the variables capturing firm-level information asymmetries donot lead to a preference for debt over equity. First, inconsistent with the asymmetric-informationhypothesis, fast-growing SOEs are more likely to issue new shares after their stock market listing.Second, larger firms tend to contract loans from (state-owned) banks, particularly when the firm re-mains majority-owned by the government. For partially privatized SOEs in which the Chinese statehas given up its 50% block, the dividend payout ratio and the number of years listed are negativelyassociated with the probability of raising equity, again in contrast to the asymmetric-informationhypothesis. Next, we demonstrate that the partially privatized SOEs with higher default risk – asmeasured by MacKie-Mason’s (1990) unlevered Z-score – prefer bank debt over equity once theytap external financial markets. Likewise, the less profitable SOEs tend to borrow from state-ownedbanks. The latter findings thus contrast with the results for listed firms in western economies butcan be explained by the fact that equity offerings by distressed firms may considerably dilute thestake held by current owners. The idea of losing control may be particularly threatening to the Chi-nese state, especially when it continues to pursue at least some political objectives after SIP. In linewith these arguments, the adverse effect of default risk only arises for the subsample of partially priv-atized SOEs in which the State hangs on to its 50% block. Finally, a high debt ratio constrains addi-tional borrowing from banks, but only when the government has given up its majority stake in thelisted SOE. The latter result again suggests differences in access to bank debt depending upon thefirm’s ownership structure.

The remainder of this article is organized as follows. In the next section, we develop our hypoth-eses, taking the specific institutional characteristics of China into account. Section 3 discusses oursample selection criteria and the basic trends in the post-listing characteristics and financing of SOEs.





Section 4 presents and discusses the empirical results on post-SIP financing decisions, while Section 5concludes the paper.

2. Development of hypotheses

In this paper, we use the framework developed by Helwege and Liang (1996) to investigate theincremental financing decisions of Chinese listed SOEs. We are particularly interested in the questionwhether economic forces have begun to drive the discrete financing choices of SOEs after their stockmarket quotation. We first analyze what exactly influences the decision to raise external finance andthen examine the determinants of the choice between bank loans and equity.

2.1. The decision to raise external finance

In the Chinese SIPs of the 1990s, the size of the fraction of shares offered to stock market inves-tors at the time of first listing was not determined by the firms’ financing needs arising from invest-ment projects (Sun and Tong, 2003; Bai et al., 2004; Huyghebaert and Quan, 2009). However, onceSOEs become subject to market forces after their stock market quotation, they may want to raise thefunds needed to exploit their growth opportunities and/or to restructure their assets and operationsin a more market-conforming way. As is pointed out by Carlin and Aghion (1996), in an economy intransition, restructuring usually requires additional capital from outsiders when the original owner,the State, is unable or unwilling to make these investments. Moreover, given that the number offirms that could start trading on the stock exchanges of Shanghai and Shenzhen was highly con-strained by the CSRC – the China Securities Regulatory Commission, the counterpart of the USSEC – under the quota system that applied up to July 1, 1999, firms not (yet) entitled to becomelisted may have obtained part of their finance from already listed SOEs. Indeed, the financial con-straints faced by listed SOEs tend to be much less severe than for unlisted firms (Li et al., 2009;Ayyagari et al., 2010). Next, Cull et al. (2009) point out that Chinese firms with better access to for-mal finance – bank loans in their study – have been extending trade credit to support their tradingpartners. Likewise, Cull and Xu (2003) find that only one-fifth of the total proceeds raised by theSOEs in their sample was invested in fixed assets. They further argue that working capital is animportant means for keeping struggling SOEs afloat, enabling them to meet their payrolls and pro-cure immediate inputs (see also Tong and Green, 2005; Cull et al., 2009). In short, when partiallyprivatized SOEs face large financing needs from investment projects, they may find it hard to fundtheir outlays on fixed assets (and possibly also working capital) out of internal cash flow. Accordingto the pecking order theory, these firms will resort to external finance. We therefore make the fol-lowing hypothesis:

Hypothesis 1. If economic forces determine incremental financing decisions after SIP, the partiallyprivatized SOEs with a deficit between investment expenditures and internal cash flow net ofdividends are more likely to raise external finance.

Conversely, some listed SOEs may continue to pursue political objectives that reduce their cashflow from operations, for example from offering low prices to consumers, from employing more work-ers than needed, etc. (see Shleifer and Vishny, 1994). As the legal protection of outside investors israther weak in China, such an expropriation of minority shareholders in listed SOEs can be accom-plished rather easily (Cull and Xu, 2003, 2005; Allen et al., 2005). Moreover, managerial incentivesare not very well aligned with those of stock market investors and hence with the overall maximiza-tion of firm value (e.g., Huang and Song, 2006). What makes matters worse is the fact that when theState remains a major owner after partial privatization, no real change takes place in the firms’ gov-ernance. So, listed SOEs still dictated by the Chinese government may continue to pursue politicalgoals after their SIP and subsequently tap external financial markets when confronted with a cashshortage. In other words, the supposedly privatized SOEs with a low or even negative cash flow fromoperations in particular may seek additional finance. In line with these arguments, Sun and Tong(2003), Wei et al. (2005), and Jiang et al. (2009) demonstrate that on average the financial and





operating performance of Chinese SOEs deteriorates after their stock market quotation. This prior re-search has further shown that this deterioration in firm performance is less severe when the govern-ment relinquished its majority stake after SIP, thus pointing out that something is involved beyondmere increased exposure to industry competition that drives down profit margins. Accordingly, we ex-pect that especially partially privatized SOEs in which the Chinese state remained the dominant ownerafter SIP are vulnerable to this type of investor expropriation (i.e. government interference for politicalreasons) and hence may collect the funds needed to meet their extensive financial obligations frombusiness operations. To ascertain whether incremental financing decisions differ between SOEs inwhich the government retained or has given up its 50% stake, we will use split-sample regressionanalysis.

2.2. The decision on the source of external finance

Given the decision to access external financial markets, we examine the subsequent choice ofinstrument, either bank loans or equity.3 Publicly issued bonds were not a viable financing alternativeduring our sample period (1994–2004). This issue is important as information asymmetries betweenfirms and investors need not result in a financing hierarchy if the array of available financing sources in-cludes hybrid instruments, such as convertible bonds (Brennan and Kraus, 1987; Noe, 1988; Constanti-nides and Grundy, 1989). So, conditional upon raising external finance, information asymmetries resultin a clear preference for debt over equity in the Chinese context.4 In line with the argument for listedfirms in developed economies, we conjecture that ex-ante uncertainty over the value of growth oppor-tunities and assets in place prompts the listed SOEs to opt for bank debt. The reason is that debt is leastaffected by the inside information held by managers and by controlling shareholders (Myers, 1984;Myers and Majluf, 1984).

Hypothesis 2. Conditional upon raising external finance after SIP, the partially privatized SOEs facinggreater ex-ante uncertainty over the value of investment opportunities and assets in place are morelikely to borrow from banks.

Finally, the literature on listed firms in western economies has argued that firms facing high de-fault risk prefer to issue new shares because risky debt offers only little advantage over equity inattenuating the mispricing of stock caused by asymmetric information. In contrast, we argue thatpartially privatized SOEs scoring high on the variables capturing default risk may dislike to completea stock offering, which could hugely dilute the government’s stake. Indeed, equity is likely to be anexpensive source of funds for firms with high default risk, as low current performance and perhapsalso bad future prospects tend to depress stock prices. When the Chinese state wishes to maintainmajority control, for example to prolong political goals, it may prefer to borrow from banks. As Chi-na’s banking system has been dominated by four large state-owned banks,5 the government couldput pressure on the banking system to lend to its favored enterprises (see also Li et al., 2009; Cullet al., 2009).6 Besides, once these SOEs can no longer meet their debt-repayment schedule, the govern-ment could compel banks to realize a bailout. So, unlike the classical models of Myers (1984) andMyers and Majluf (1984), we predict that, having decided to raise external finance, the listed SOEs with

3 There are two main forms of share issues in China: rights issues, addressed only to current shareholders, and seasoned equityofferings, open to all investors. In some cases, the Chinese state has transferred its right to participate in a rights issue to domesticstock market investors. The shares resulting from a rights issue are non-tradable and are classified as a separate category of stock.Overall, this category usually accounts for less than 1% of a listed SOE’s outstanding equity (e.g., Huang and Song, 2006).

4 In contrast, in the sample of Helwege and Liang (1996), bond issues are almost as important as seasoned equity offerings,whereas convertible bonds represent about two-thirds of all public debt offerings.

5 During our sample period, the government fully owned the Industrial and Commercial Bank of China, Agricultural Bank ofChina, China Construction Bank, and Bank of China. Many other banks were also under the control of the State.

6 The central bank – the People’s Bank of China (PBC) – decides on the reference interest rate for these loans, with a band withinwhich local bank branches have some discretion to adjust their lending rates to firm-specific credit quality. Since 1999, the bandhas been 10% below and 30% above that reference rate. For loans to large enterprises, the upper limit is restricted to 10%. Accordingto García-Herrero and Gavilá (2005), due to the banks’ lack of expertise in assessing borrower credit risk, most loans are contractedjust at or even below the PBC’s reference rate.





greater default risk may seek to borrow from banks, especially when the Chinese state hangs on to itsmajority stake. Conversely, the incentives to prevent a further dilution of state ownership likely aresmaller for listed SOEs where the State already relinquished its 50% block. The above arguments resultin the following hypothesis:

Hypothesis 3. Conditional upon raising external finance after SIP, the partially privatized SOEs facinggreater default risk are more likely to borrow from banks. If this relation exists, it should be driven bythe subsample of firms in which the government retained majority control.

3. Sample selection and description

Our database comprises the audited consolidated financial statements of 344 non-financial SOEsthat listed A shares for the first time on the Shanghai stock exchange (SHSE) between January 1994and June 1999. It was obtained from Shenyin & Wanguo Securities Company Ltd., a leading Chineseinvestment bank. We excluded financial SOEs, as they are subject to specific regulation and filetheir financial statements under a different format. We chose 1994 as the starting date because thatis when new accounting rules closer to international standards were introduced.7 We limited oursample to July 1, 1999 because that is when the quota system was abolished. Huyghebaert and Quan(2009) show that the SOEs that listed after the end of the quota system were of substantially lowerquality. This institutional change could have affected post-SIP decision processes, thus producing astructural break in the data. We follow the firms in our sample over a five-year period after theirstock market quotation, up to July 1, 2004. This ending date is well before the split-share reformof June 2005, which could have engendered another structural break in the data.8 Accordingly, byrestricting our sample of SIPs to the period January 1994–June 1999, we could avoid getting entangledin these institutional changes while still tracking all sample firms for a long-enough time after theirfirst listing.

In order to restrict our focus to former SOEs, we selected only firms in which the Chinese govern-ment – directly or indirectly via state-owned legal persons – owned at least a 20% stake before SIP (seealso La Porta et al., 1999). Detailed ownership data were collected from the SINA Finance database,which also provides information on the type of legal persons, state-owned or not. Legal-person shareswere re-classified as ‘indirect’ state shares if the Chinese government held a majority stake in the legalperson, since the latter legal persons can be presumed to look after the government’s interests. Anydistinction between these legal persons and the Chinese state is therefore likely to be artificial. Finally,many studies using Chinese stock market data have noted a time lag between stock offering and theactual listing of shares (e.g., Chan et al., 2004; Huyghebaert and Quan, 2009). In the case of stocks with‘pending historical problems’, this lag can amount to several years. So we excluded the SOEs for whichthis time lag exceeded six months, to take into account that their privatization process and thereforepossibly also their post-SIP decisions could be different. These selection criteria reduced the samplefrom 344 to 221 SOEs. On average, the Chinese government held a stake of 76.19% right before SIP(median of 80.03%).9 At SIP-time, the sample firms sold primary shares equal to an average of 24.57%and a median of 24.62% of shares outstanding. Compared to the figures reported by Jones et al.(1999), the Chinese state thus reduces its ownership to a far smaller extent at SIP. In 154 firms

7 Firms listed as of 1994 were required to adjust their pre-1994 financial statements to the new accounting rules.8 In June 2005, the Chinese state authorized 46 partially privatized SOEs to circulate their non-tradable shares in the near future.

They had to work out a feasible plan with the owners of the A shares, thus giving them some compensation in shares or in cash. Asof January 2006, most listed SOEs were allowed to begin talks towards such agreements.

9 The fact that some SOEs have less than 100% state ownership before SIP is the outcome of corporatization, which restructuredSOEs into shareholding companies. According to the Company Law of July 1994, they could take one of two forms: limited liabilityfirms with a minimum of two shareholders or limited liability shareholding companies with a minimum of five shareholders. As aresult, apart from the government, domestic institutions, including other SOEs, stock-holding companies, and non-bank financialinstitutions were allowed to participate as a co-founder or as a funds provider. The shares held by these investors are classified aslegal-person shares. Finally, in some cases, employees or managers also obtained a small stake in their firm during this process ofcorporatization.




Table 1Annual and industry distribution of the sample.

Raw data Sample

Panel A: Annual distributiona

1994 66 381995 15 51996 103 601997 85 561998 53 451999 22 17Total number of firms 344 221

Panel B: Industry distributionb

Agriculture, forestry and fishing 8 6Mining 2 2Electricity, gas and water production and supply 15 9Construction 9 6Transportation and storage 14 8Information technology 20 12Retail and wholesale trade 42 30Real estate 4 3Services 11 9Media 4 1Conglomerates 32 9Manufacturing 183 126

Food & beverages 18 15Textile 13 7Printing 10 7Petroleum, chemicals and plastic products 33 24Electronics 10 5Metal & non-metal 28 21Machines 49 34Pharmaceuticals 17 11Other manufacturing 5 2

Total number of firms 344 221

a Columns 1 and 2 report the annual distribution for the raw data and for the sample of 221 SOEs. The raw data includeinformation on 344 non-financial firms listing A shares for the first time on the Shanghai stock exchange from January 1994 toJune 1999. The sample that is used in the analyses consists of 221 SOEs that were required to have state shares or state-ownedlegal person shares accounting for at least 20% of total shares outstanding before SIP. In addition, the time lag between shareoffering and share listing should not exceed six months.

b Columns 1 and 2 report the industry distribution for the raw data and for the sample of 221 SOEs, based on the CSRCindustry classification code.


(69.68%), the government even retained its majority stake at SIP. By the fifth year after SIP, the Chinesestate still owns 50.35% of total stock in the average partially privatized SOE. Nonetheless (see alsoHuyghebaert and Quan, 2011), the stake held by the general public, typically retail investors in theChinese context, is increasing over time, to 33.89% by the fifth post-SIP year. None of the firms leftthe sample before reaching the listing age of five years, so sample attrition is not a problem in ourstudy.10 For all sample firms, we manually collected the SIP prospectus and the consolidated financialstatements from one year before to five years after SIP. Stock price data were collected from Datastream.

Table 1, Panel A shows the number of firms that listed on SHSE each year from 1994 to June 1999.The number of SIPs is fairly large in every year but 1995. Panel B displays the industry distribution ofsample firms, using the CSRC classification code. A majority of sample firms (57.01%) is active in man-ufacturing, reflecting that Chinese SOEs mainly developed from the needs of heavy industries andproducts of strategic importance, such as energy resources, basic metals, and consumer staples.

10 For comparison, Helwege and Liang (1996) find that 32% of their sample firms were no longer listed five years after IPO; half ofthem had gone bankrupt, the others had been taken over or merged into a new company.




Table 2Summary statistics over time.

Variables Year 1 Year 2 Year 3 Year 4 Year 5

DEFICIT/ASSETS 0.0521 0.0454 0.0254 0.0302 0.0304(0.0993) (0.1042) (0.0892) (0.0987) (0.1037)

DEFICIT (including NCWC)/ASSETS 0.0713 0.0641 0.0375 0.0303 0.0219(0.1222) (0.1268) (0.1209) (0.1134) (0.1372)

CAPEX/ASSETS 0.1193 0.1068 0.0832 0.0674 0.0614(0.1085) (0.1134) (0.1013) (0.1009) (0.1088)

CASHFLOW/ASSETS 0.0919 0.0795 0.0703 0.0512 0.0437(0.0572) (0.0601) (0.0589) (0.0732) (0.0749)

SALES GROWTH 0.1393 0.1243 0.1426 0.1370 0.1473(0.4208) (0.4184) (0.3839) (0.3732) (0.4312)

FIRM SIZE 20.5974 20.8011 20.9237 21.0264 21.0812(0.8132) (0.7933) (0.8033) (0.8235) (0.8655)

INTANGIBLES/ASSETS 0.0202 0.0214 0.0239 0.0260 0.0274(0.0262) (0.0279) (0.0323) (0.0383) (0.0410)

PPE/ASSETS 0.3602 0.3664 0.3647 0.3634 0.3667(0.1647) (0.1770) (0.1840) (0.1867) (0.1830)

DIVIDEND PAYOUT 0.4586 0.3736 0.3553 0.3461 0.3057(0.4520) (0.3868) (0.4350) (0.3849) (0.3814)

STOCK VOLATILITY 0.0203 0.0206 0.0194 0.0179 0.0164(0.0051) (0.0041) (0.0044) (0.0052) (0.0056)

INVERSE Z-SCORE 0.8675 0.9491 0.9367 1.0778 1.0228(0.4724) (0.6754) (0.7720) (0.8520) (0.9020)

LEVERAGE 0.3956 0.3960 0.4107 0.4348 0.4436(0.1558) (0.1604) (0.1651) (0.1777) (0.1879)

N observations 221 221 221 221 221

This table reports summary statistics (mean and standard deviation) on the test variables in each of the five post-listing years.The sample of 221 listed SOEs was selected using the criteria identified in Table 1. DEFICIT/ASSETS is the firm’s deficit ofinvestment expenditures on fixed assets (CAPEX) minus internally generated cash (CASHFLOW) net of cash dividends. SALESGROWTH is the real growth rate of sales. FIRM SIZE is the logarithm of total assets. INTANGIBLES/ASSETS is the ratio ofintangible assets to total assets. PPE/ASSETS is the ratio of property, plant, and equipment to total assets. DIVIDEND PAYOUT isdividends scaled by net income. STOCK VOLATILITY is the standard deviation of the residuals from the market model. INVERSEZ-SCORE is the inverse of Altman’s unlevered Z-score. LEVERAGE is the book value of debt and current liabilities relative to totalassets.


3.1. Trends in firm characteristics

Table 2 reports descriptive statistics (mean and standard deviation) for the firm’s deficit of invest-ment expenditures relative to internally available cash, information asymmetries, and default risk ineach of the first five listing years. All data are taken on December 31 of each year. The variables arewinsorized at 1–99%, to remove the most extreme values in either tail of the distribution. First, thedeficit of investment expenditures on intangible, tangible, and financial fixed assets11 vis-à-vis inter-nally generated cash net of dividends decreases over time, whether or not investment outlays include thechange in non-cash working capital (NCWC). Some 25–30% of listed SOEs have a negative value for thedeficit variable (i.e. a cash surplus) in the first two years after SIP; by year 5 this percentage increases to40%. The ratio of cash flow to total assets is highest shortly after SIP and declines substantially afterwards(the average drops from 9.19% in the first post-SIP year to 4.37% in year 5). Overall, we find an increasingnumber of firms with deteriorating financial conditions after SIP.

SOE sales grow considerably in each of the five years after SIP. Each year, about one third of thefirms report a decrease in sales (not reported in Table 2). The size of the average firm increases steadilyover time. Overall, we find no large changes in the ratios of intangible assets or of property, plant, andequipment (PPE) to total assets. In line with deteriorating firm performance, the dividend payout ratiodeclines firmly in the post-listing period. Stock volatility shows no strong trend. Yet, the inverse

11 Investments in financial fixed assets are negligible, in line with the idea that China lacks an active market for corporate control.




Table 3Annual distribution of external fundraising.

Year 1 Year 2 Year 3 Year 4 Year 5

Number (%) of listed SOEs that raise new bank debt 131(59%)

129(58%)

126(57%)

106(48%)

93(42%)

Number (%) of listed SOEs that make an equity offering 106(48%)

94(43%)

67(30%)

39(18%)

41(19%)

Number (%) of listed SOEs that raise external funds 172(78%)

168(76%)

153(69%)

121(55%)

115(52%)

Number (%) of listed SOEs that raise no external funds 49(22%)

53(24%)

68(31%)

100(45%)

106(48%)

A new bank loan is defined as occurring in every year in which the total amount of bank debt outstanding increases by 10% ormore. Likewise, an equity offering is defined as occurring in every year in which the total number of shares outstandingincreases by 10% or more, stock dividends not taken into account.


Z-score reveals that failure risk augments over time. Likewise, leverage trends upward. The firms inour sample rely considerably more on debt finance than those reported on in Helwege and Liang(1996). Yet, their leverage is smaller than that of the predominantly unlisted Chinese firms studiedby Li et al. (2009), showing a median debt ratio of 59%. Given that the corporate bond market is almostnon-existent in China, non-bank debt consists largely of trade credit extended by other SOEs (knownas the Chinese triangular debt problem).

3.2. Trends in financing decisions

Table 3 shows the number and the percentage of listed SOEs that raised new bank debt and equityin each of the five years after SIP. Like Helwege and Liang (1996), we define the realization of a newbank loan whenever total bank debt increased by 10% or more.12 Unlike them, however, we also includeshort-term bank loans in our calculations, to account for the large number of SOEs that lengthen thematurity of their loans after listing.13 Ignoring this feature of Chinese capital markets would definitelyinflate the number of new bank loans. An equity offering is defined analogously as an increase of at least10% in the number of shares outstanding. We do correct for stock dividends, which were paid out in 108of our firm-year observations (9.77%). All newly issued shares were sold to domestic (retail) investors inthe public equity market, typically through a seasoned equity offering. Private placements of newly is-sued shares never happened.

Table 3 reveals that nearly 60% of the sample firms increased their total bank borrowings in each ofthe first three years after SIP; this falls to 42% by year 5. Equity offerings follow the same pattern.Share issuance is quite high in the first two years after SIP – about 45% – but levels off to a moderate19% afterwards. These percentages are far higher than those reported by Helwege and Liang (1996). Intheir sample, private debt offerings declined steadily, from 40% in the first post-IPO year to under 30%five years later. Each year, less than 5% of the firms in the Helwege and Liang sample issued newshares, with no clear time pattern. Overall, the results in Table 3 show that it is necessary to studyincremental financing decisions in the case of partially privatized SOEs in China. In fact, in Table 2 thesevery significant dynamics in financial markets were concealed.

Table 4 reveals once more that bank loans are much more important than stock offerings. The tablefirst shows that the 221 partially privatized SOEs in the sample contracted 585 bank loans during oursample period. Going into more detail than the table presents, we find that 34 of the firms raised onebank loan, 50 firms raised two, 67 firms raised three, 40 firms raised four, and 18 raised five. Similarly,

12 SOEs that increased their bank debt from zero in a particular year are always classified as having obtained a bank loan (13companies).

13 Demirgüç-Kunt and Maksimovic (1999) show that firms in developing countries tend to rely more on short-term finance, inparticular short-term bank loans and trade credit. Yet, firms obtain better access to long-term funds after their stock marketintroduction. In line with this argument, Li et al. (2009) find that over 80% of listed firms in China have access to long-term debt incomparison to only 40% for the population of Chinese manufacturing firms.




Table 4Transaction and firm characteristics of listed SOEs raising external finance.

Bank debt Equityoffering

t-test, p-value

Wilcoxon p-value

Number of offerings 585 347Number of firms 221 187Proceeds raised in the offering (median, RMB) 88,060,017 57,178,824 0.0001 0.0001Proceeds/assets (median) 0.0741 0.0564 0.0001 0.0001Total assets in year prior to the offering (median, RMB) 1,228,172,129 1,023,908,270 0.0119 0.0007Sales in year prior to the offering (median, RMB) 519,859,133 491,626,285 0.6354 0.9041Growth rate of total assets in year prior to the offering

(median)0.2071 0.2242 0.0001 0.0901

Growth rate of sales in year prior to the offering(median)

0.1339 0.1471 0.1318 0.4413

Cash flow/assets in year prior to the offering (median) 0.0370 0.0886 0.0001 0.0001EBITDA/assets in year prior to the offering (median) 0.0703 0.0841 0.0001 0.0001State ownership in year prior to the offering (median) 0.5848 0.5807 0.1856 0.1479

A new bank loan is defined as occurring in every year in which the total amount of bank debt outstanding increases by 10% ormore. Likewise, an equity offering is defined as occurring in every year in which the total number of shares outstandingincreases by 10% or more, stock dividends not taken into account.


187 listed SOEs made a total of 347 equity offerings, with 79 firms completing one, 65 firms complet-ing two, 34 firms completing three, and 9 firms completing four. All the sample firms raised externalfinance at least once during their first five years after listing. Table 4 also displays some importantcharacteristics of the fundraising operations and the listed SOEs that undertook them. First, medianproceeds were RMB 88,060,017 for bank loans and RMB 57,178,824 for equity offerings.14 The largersize of bank loans in our sample is not engendered by a preference for bank debt on the part of the largestSOEs, as it persists when proceeds are expressed as a percentage of total assets. Arguably, this findingonce more stresses the importance of the (state-owned) banking sector for the financing of Chinese SOEs,even after their stock market quotation. Second, firm size and firm growth do not differ substantiallyacross partially privatized SOEs by type of security offered. We find only that the SOEs relying on bankloans have more total assets, while their one-year lagged asset growth rate is lower. However, we do ob-serve that the partially privatized SOEs that issue new shares are more profitable. Finally, the fraction ofstate ownership bears no relation with the choice of funding instrument.

4. The driving forces behind financing decisions

The main implication of the pecking order model, namely that internal funds are preferred to exter-nal funds to finance investment projects, results in an empirically testable sequential decision process,whereby firms first decide whether or not to access external financial markets. Conditional upon rais-ing funds, they decide on the type of security to offer. These two questions are covered in Sections 4.1and 4.2, respectively. Section 4.3 discusses the results of a multinomial logit model, in which the deci-sion to seek external funds and the choice of instrument are studied at the same time. Finally, we pres-ent the results of some additional robustness checks.

4.1. The decision to raise external finance

The dependent variable, External financing, in the first logit model is equal to one if the partiallyprivatized SOE either contracts a new bank loan or issues new shares in a given year, and zerootherwise (see Section 3.2). Overall, external finance was obtained in 729 out of 1105 firm-yearobservations, or 65.97% of the cases. The test and control variables are discussed hereafter. Like

14 All absolute numbers are expressed at constant 1994 prices in order to take into account the annual inflation rate.





Helwege and Liang (1996), we lag all explanatory variables by one year in the estimations, unless sta-ted otherwise in order to avoid their possible contamination by the proceeds raised in the offer.

The most important explanatory variable in the external finance logit equation is the contempora-neous value of the deficit, calculated as gross investment expenditures on fixed assets minus inter-nally generated cash net of cash dividends. The deficit is scaled by total assets to avoidheteroscedasticity problems. Under the pecking order theory, we expect a positive coefficient on thisvariable. We also construct a second proxy for the deficit, now including the change in non-cash work-ing capital (NCWC) as an additional investment outlay. The reason is that inter-firm trade credit isquite important in China (Cull and Xu, 2003; Tong and Green, 2005; Cull et al., 2009). Shyam-Sunderand Myers (1999) also include working capital in their definition of the deficit so as to recognize thatfirms may fill a cash shortfall by drawing upon financial slack built up from retained earnings. By con-trast, Helwege and Liang (1996) do not count the change in working capital when calculating the def-icit as it largely reflects previous security offerings, in particular the proceeds of the IPO. As IPOproceeds turn out to be strongly correlated with the probability of raising extra funds after listing,including working capital in the definition of the deficit could generate spurious conclusions on therelation between the deficit and the external finance decision.15 We seek to reconcile these argumentsby incorporating the change in non-cash working capital (NCWC). Overall, we find that depending onwhether the definition of the deficit excludes (includes) NCWC, external funds are raised in 71% (75%)of the firm-year observations for which a cash shortage is registered, while financial markets are tappedin only 62% (53%) of the firm years for which a cash surplus is recorded. Also, of the external financingevents, 51% (67%) are associated with a cash shortage and 49% (33%) with a cash surplus. These summarystatistics indicate that the listed SOEs that raise external funds are far more likely to report a cash short-age, while a large fraction of firms without financing needs (62% or 53%, depending on the definition) stillaccess external financial markets after SIP.

Next, we split these deficit variables into components, a positive one and a negative. So, we con-struct two types of deficit variables: one equal to the cash shortage if the deficit is positive, and equalto zero otherwise; the second equal to the additive inverse of the deficit if the firm has a cash surplus,and equal to zero otherwise. According to the pecking order theory, the coefficient on the cash short-age variable in the external finance logit model should be positive, that on the cash surplus negative.Alternatively, we separately include the ratio of capital expenditures to total assets and the ratio ofcash flow to total assets. Using the same procedure as for the deficit, we next split the cash flow var-iable into a positive and a negative component. This latter specification should allow us to explore inmore detail the argument that especially listed SOEs running out of cash from operations have anincentive to raise funds in order to fill this cash shortfall.

To allow for the possibility that partially privatized SOEs with positive deficits arising from invest-ment projects but with a low operating cash flow may have trouble raising the funds that they need,we re-estimate the external finance logit equation after including an interaction term between thedeficit variable, if positive, and the ratio of cash flow to total assets. This interaction term thus ac-counts for the possibility that listed SOEs with large cash shortages may have been shut out of finan-cial markets after their SIP, either because they lost substantial going concern value or because theirprospects worsened. We therefore expect this interaction term, if significant, to be positively relatedto the dependent variable.

In the above models, we control for firm growth, firm size, and financial smoothing. First, if salesgrowth has been slow, investors may consider the firm’s prospects to be waning. In that case, capitalexpenditures might over-state future growth prospects. Thus, firms with below-average sales growthmay be less likely to obtain external finance for a given level of the deficit. We take this idea into ac-count by including the firm’s real sales growth rate in the previous year and conjecture a positive coef-ficient on this variable. Second, large firms may have deficits that are small relative to their asset basebut nonetheless huge in absolute terms. These listed SOEs could be inclined to finance their cashshortages externally. In order to deal with this argument, we include the logarithm of total assets

15 This argument is also non-trivial in our sample, as SOEs offering a larger fraction of primary shares at the SIP are more likely totap external financial markets after their listing.





in the external finance logit equation and expect a positive sign. Third, we control for financialsmoothing by including an indicator variable equal to one for firms that obtained external financein the previous year, and zero otherwise. If firms raise funds in one year to avoid the fixed costs of hav-ing to collect finance again in the near future, the indicator variable should have a negative coefficient.Finally, we add industry and year fixed effects to all models. Year dummies, for instance, should beable to correct for market-timing behavior.

The empirical results on the external finance decision, reported in Table 5, are broadly consistentwith the pecking order theory. More importantly, they point at a number of peculiarities of Chinesepartially privatized SOEs. First, the deficit variable has a significant positive coefficient in column 1,in line with Hypothesis 1. When this variable is split into cash shortage and cash surplus components(column 2), we note that only the firms with a cash surplus are less inclined to tap external financialmarkets. The size of the cash deficit, if positive, is indeed not related to the probability of outside fund-raising.16 Yet, once the change in non-cash working capital is factored into the deficit calculation, it turnsout that the firms with larger cash shortages are more likely to raise external finance after all (comparecolumns 2 and 11). Arguably, these findings indicate that the firm’s own financing needs for expansionand/or restructuring bear no strong relation with fundraising after SIP; this conclusion is consistent withearlier findings at the time of first listing (see Sun and Tong, 2003; Bai et al., 2004; Huyghebaert andQuan, 2009). Rather, the results in Table 5 reveal that external funds are gathered at least in part to sup-port other firms via inter-firm trade credit, in line with Cull et al. (2009). Examining the various compo-nents of NCWC in more detail (not reported in Table 5), we found that the partially privatized SOEs thatextended more credit to customers (accounts receivable) tended to raise more external finance, but theirown use of trade credit (accounts payable) was not reduced.

When the deficit variable is split into investment outlays and internally generated cash (column 3),we find that in relation to firm size only the former (CAPEX/ASSETS) is significant, with a positive sign.CASHFLOW/ASSETS is indeed not significant. To further explore the argument that especially listedSOEs running out of cash from operations, for example from pursuing political objectives, may wantto collect extra funds in order to fill their cash shortfall, we split CASHFLOW/ASSETS into positive andnegative components. The results in column 4 show that for listed SOEs with a positive cash flow fromoperations, the size of this cash flow is not significantly related to the probability of tapping externalfinancial markets. In contrast, partially privatized SOEs with a negative cash flow are less inclined toraise extra funds after SIP. So, we find no supportive evidence for the idea that listed SOEs running outof cash from operations raise external finance to fill a cash shortfall. Finally, column 5 reveals that theinteraction term between DEFICIT/ASSETS P 0 and CASHFLOW/ASSETS is significant and positive.Arguably, partially privatized SOEs with a positive deficit arising from investments but with a lowoperating cash flow have trouble raising the finance they need. This significant positive interaction ef-fect is robust to the methodology developed by Ai and Norton (2003).

As for the control variables, the sales growth rate has a significant positive coefficient in Table 5, inline with Helwege and Liang (1996). When using asset growth as an alternative measure, not reportedhere, we found a similar positive effect. Firm size, in contrast with our conjectures, has a negative andsignificant coefficient. Yet a negative sign could reflect that large firms typically exhibit limited growthopportunities. This same relation emerged when we proxied firm size by log(sales) or by log(marketcapitalization). Finally, firms that raise external finance in one year are also more likely to do so in thenext. This finding is consistent with Helwege and Liang (1996), but incompatible with the idea offinancial smoothing.

As earlier research has shown that the decision to reduce government ownership below 50% hasnotable implications for firm performance after SIP, we also run the regressions for the subsamplesof firms in which the Chinese state retained and relinquished its majority stake after SIP, respectively.The results of these split-sample analyses, which are reported in columns 6–9 and 12–15, offer

16 We have checked the distribution of the cash shortage and cash surplus variables. The standard deviation of DEFICIT/ASSETS P 0 equals 0.0787 while that of DEFICIT/ASSETS < 0 equals 0.0484, significantly different under an F-test (F-value of 2.95;p-value < 0.0001). As we have dealt with outliers upfront by winsorizing the data, the results in Table 5, columns 2 and 5 areunlikely to be driven by this property of the data. If anything, it is more difficult – from a econometric point of view – to find aneffect for the variable with the lowest standard deviation.




Table 5Logit regression results on the external finance decision.

Variable Predsign

Deficit does not include NCWC Deficit includes NCWC

Full sample Government P 50% Government < 50% Full sample Government P 50% Government < 50%

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)

Constant 6.4911(0.0007)

6.9591(0.0003)

6.6663(0.0005)

7.1162(0.0002)

7.2855(0.0002)

7.3221(0.0008)

7.5628(0.0006)

4.2654(0.4058)

4.0542(0.4285)

6.5667(0.0008)

5.9305(0.0026)

6.4143(0.0039)

6.5055(0.0035)

2.7569(0.6039)

3.0092(0.5722)

DEFICIT/ASSETS + 2.0811(0.0059)

4.7358(0.0001)

DEFICIT/ASSETS P 0 + 0.8058(0.4389)

0.8481(0.4369)

0.3779(0.7651)

1.9500(0.3262)

7.3347(0.0001)

6.1963(0.0001)

5.5640(0.0015)

10.1251(0.0006)

10.1897(0.0003)

DEFICIT/ASSETS < 0 � �4.7362(0.0064)

�4.3622(0.0122)

�4.9167(0.0245)

�3.4909(0.2349)

�2.7852(0.0072)

�3.3401(0.0151)

�3.3518(0.0146)

�1.7822(0.2832)

�1.6258(0.3262)

CAPEX/ASSETS 3.4121(0.0001)

3.6190(0.0001)

3.8257(0.0005)

2.9897(0.0488)

CASHFLOW/ASSETS � 1.6769(0.1795)

CASHFLOW/ASSETS P 0 � �0.9921(0.5875)

�0.8340(0.7082)

0.0787(0.9818)

CASHFLOW/ASSETS < 0 + �5.9109(0.0204)

�7.6052(0.0427)

�4.5805(0.2246)

DEFICIT/ASSETS P 0 � CASHFLOW/ASSETS

+ 15.6396(0.0441)

11.7277(0.4998)

29.0638(0.0155)

SALES GROWTH + 1.0822(0.0001)

1.0796(0.0001)

0.8147(0.0005)

0.8254(0.0004)

0.9576(0.0001)

1.3474(0.0001)

1.0112(0.0018)

0.8312(0.0184)

0.6548(0.0670)

0.8289(0.0004)

0.8370(0.0004)

1.1370(0.0004)

1.0973(0.0008)

0.6148(0.0962)

0.5809(0.1049)

FIRM SIZE + �0.3185(0.0004)

�0.3349(0.0002)

�0.3326(0.0002)

�0.3452(0.0001)

�0.3486(0.0001)

�0.3572(0.0004)

�0.3738(0.0003)

�0.2006(0.4122)

�0.1886(0.4398)

�0.3222(0.0004)

�0.3006(0.0010)

�0.3266(0.0015)

�0.3300(0.0013)

�0.1481(0.5574)

�0.1585(0.5312)

PRIOR FINANCING � 0.5691(0.0003)

0.5630(0.0003)

0.5501(0.0005)

0.5523(0.0005)

0.5636(0.0003)

0.4966(0.0079)

0.5086(0.0074)

0.6270(0.0352)

0.5655(0.0619)

0.5422(0.0008)

0.5528(0.0006)

0.4908(0.0110)

0.4949(0.0105)

0.6037(0.0508)

0.5887(0.0571)

Number of observations 1105 1105 1105 1105 1105 709 709 396 396 1105 1105 709 709 396 396Pseudo R-square (%) 10.73 10.99 12.10 12.44 11.37 11.74 15.81 12.58 18.73 15.36 15.82 15.81 15.86 18.73 19.22

The dependent variable External financing is equal to one if a firm either raises a new bank loan or issues new shares in a given year, and zero otherwise. A new bank loan is defined asoccurring in every year in which the total amount of bank debt outstanding increases by 10% or more. Likewise, an equity offering is defined as occurring in every year in which the totalnumber of shares outstanding increases by 10% or more, stock dividends not taken into account. External financing was obtained in 729 out of 1105 firm-year observations. Theexplanatory variables are described in Table 2. All models include industry and year fixed effects. p-Values are reported between parentheses.

N.H

uyghebaertet

al./J.Finan.Intermediation

xxx(2013)

xxx–xxx

13

Pleasecite

thisarticle

inpress

as:H

uyghebaert,

N.,

etal.

J.Finan.

Intermediation

(2013),http://dx.doi.org/

10.1016/j.jfi.2013.08.005




additional insights. First, the positive effect of the cash shortage variable after including the change innon-cash working capital arises regardless of ownership structure. That is, all listed SOEs raise exter-nal finance to extend credit to their trading partners, whether or not the government retained itsmajority stake. This finding therefore suggests that this trade credit is not politically motivated butrather reflects the severe financial constraints faced by many Chinese firms, given the immature stateof Chinese financial markets. Then, the firms with better access to external finance – in this study, thepublicly listed SOEs – may provide trade credit to their customers that are denied access to formalfinancial markets, in line with the redistribution theory of trade credit (see also Cull et al., 2009). Sec-ond, Table 5 reveals that only the listed SOEs in which the government hangs on to its 50% block areless inclined to raise extra funds when confronted with a cash surplus. This outcome seems to indicatethat the listed SOEs that are still controlled tightly by the Chinese state are subject to less bindingfinancial constraints, as they can afford to forgo outside fundraising when realizing a cash surplus.Once the government has given up its majority stake, the firm may not have this same privilegeand may accordingly seek funds wherever possible. Third, listed SOEs with a negative cash flow fromoperations are less likely to tap external financial markets, but only when the State retained its major-ity stake after SIP. So, we find no evidence whatsoever for the idea that partially privatized SOEs run-ning short of cash as they (continue to) pursue political goals compensate by outside fundraising. Thelatter result also indicates that adverse selection can be ruled out as an alternative explanation for thenegative relation between CASHFLOW/ASSETS and the external finance decision. Fourth, the interac-tion term between DEFICIT/ASSETS P 0 and CASHFLOW/ASSETS is significant only for the subsampleof firms in which the government has given up its 50% block. These results, which are again robust tothe Ai and Norton (2003) methodology, thus suggest that poor performance negatively affects theprobability of obtaining external finance only when the State has opted for a more radical change inownership after SIP, by giving up majority control. Finally, firm size bears no significant relation withthe post-SIP financing decisions of SOEs in which the government relinquished its majority stake.

4.2. The decision on the source of external finance

In this section, we examine whether the partially privatized SOEs that raise external finance followsome hierarchy of funding instruments. To this purpose, we again estimate a logit model, where thedependent variable Equity now equals one if the firm issues new shares in a given year, and zero otherwise.If the firm raises both equity and bank debt in the same year, Equity is coded one.17 Overall, Equity is equal tounity for 347 of 729 (47.60%) firm-year observations where external financial markets were tapped.

We use several variables to proxy ex-ante uncertainty over the value of the firm: real sales growth,firm size (log of total assets), the ratio of intangible to total assets, the ratio of property, plant, andequipment to total assets, the dividend payout ratio, stock volatility (standard deviation of market-model residuals), a dummy equal to one if foreign investors can hold part of the firm’s stock in theB or H share market, and number of years since listing. The first six variables are specific to the firmand are one-year lagged. They capture the idea that problems of asymmetric information are more se-vere for fast-growing, smaller firms with a limited proportion of hard assets, a lower payout ratio, orgreater idiosyncratic volatility of stock returns. Also, firms whose shares are traded in the B or H sharemarket are subject to stricter information requirements (see Bai et al., 2004). Finally, when firms havebeen listed longer, more information on them tends to be available.18

17 We found it impossible to check the robustness of our results when using other definitions of Equity, given the large incidenceof bank loans in our sample. For example, it would be interesting to test the stability of our results when setting Equity equal to onein years where a stock offering was completed and no money was raised from banks.

18 The finance literature has also argued that firms subject to larger information asymmetries may underprice their IPO moresharply to signal firm quality to investors, as they expect to recoup these costs in future equity offerings. In other words, IPOunderpricing can be used to leave a good taste in the investors’ mouth. In the context of SIPs, Perotti (1995) argues thatunderpricing is needed to signal government commitment to privatization. As such, a larger reduction in state ownership at SIPneeds to be offset with more underpricing. This model has been supported empirically for Chinese SIPs by Huyghebaert and Quan(2009), who also show that ex-ante uncertainty over firm value is not related to SIP underpricing. Therefore, we do not include SIPunderpricing in the security choice model. As a robustness check, we did examine the effects of SIP underpricing and residual SIPunderpricing – after accounting for its positive association with the fraction of shares floated – and find that these variables are notrelated to the instrument choice. These results are not reported, but can be obtained from the authors upon request.





To measure default risk, we use the unlevered Z-score, which captures the portion of Altman’s(1968) Z-score that is not affected by the firm’s financing choices. It is a weighted average of EBIT,sales, retained earnings, and working capital, all scaled by total assets (see also MacKie-Mason,1990).19 Next, we include the ratio of cash flow to total assets and leverage, defined as debt plus currentliabilities over total assets. To account for the fact that firms paying high corporate taxes may have anincentive to raise debt rather than equity, we include the firm’s tax rate as a control variable. In orderto calculate the latter variable, we use the methodology developed by Booth et al. (2001).20 In the finalregression specification, we also add industry and year fixed effects. Year dummies may again pick upmarket-timing behavior.

The results of the logit regression model on the security choice are reported in Table 6, which dis-plays the results for the full model and also for a simplified model retaining only the variables thatwere significant in at least one specification. As far as the variables capturing firm-level informationasymmetries are concerned, we find only a limited effect on SOE financing choices. What’s more,the signs of the variables that are statistically significant are the opposite of those expected underHypothesis 2. First, we detect no negative relation between real sales growth and the probability ofmaking a stock offering. Rather, sales growth is positively correlated with the decision to issue extrashares, irrespective of the firm’s ownership structure. This same result is obtained using the assetgrowth rate (not reported in Table 6). Second, in contrast to the asymmetric-information hypothesis,firm size negatively affects the likelihood of selling new shares. A likely explanation is that the gov-ernment cares more about preserving control over its largest listed SOEs. Indeed, the Chinese statemay be concerned about the political ramifications of massive job losses and plant closings once itloses control over its largest SOEs, as this could jeopardize social stability and undermine public sup-port for large-scale privatization, particularly at this early stage of its SIP program. By relying on bankloans, large partially privatized SOEs can prevent a further dilution of government ownership. Consis-tent with these arguments, the negative effect of firm size arises only for the subsample of listed SOEsin which the State retained majority control. Third, the dividend payout ratio and the length of timelisted are negatively and significantly associated with the decision to issue new shares, again runningcounter to the asymmetric-information hypothesis. This relation is found only for the subsample oflisted SOEs in which the government relinquished its 50% block. The other asymmetric-informationvariables, capturing the composition of assets, stock volatility, and a B/H share market listing, arenever significant. In sum, information asymmetries bear no negative relation with the decision to issuenew shares after SIP, consistent with earlier findings at the time of first listing (see Chan et al., 2004;Huyghebaert and Quan, 2009). As the sample firms could not issue hybrid instruments, we can ruleout non-linearities as an alternative explanation for the lack of significance of information asymme-tries to explain the security choice decision (see Brennan and Kraus, 1987; Noe, 1988; Constantinidesand Grundy, 1989).

In line with Hypothesis 3, we find that MacKie-Mason’s (1990) unlevered Z-score, our most impor-tant proxy for default risk, affects the probability of issuing extra shares negatively and significantly.This result contrasts with the traditional pecking order model but can be explained by the fact thatoffering new stock would considerably dilute government ownership for firms facing substantial de-fault risk. Conversely, unlike firms in western economies, partially privatized SOEs exhibiting high de-fault risk may still be able to borrow from state-owned banks. Also, they may take into account thateven a subsequent default on this bank debt will not automatically trigger bankruptcy. Our inferencesare further supported by the positive correlation between the ratio of cash flow to total assets and the

19 The unlevered Z-score is calculated as 3.3 � EBIT/assets + 1.0 � sales/assets + 1.4 � retained earnings/assets + 1.2 �workingcapital/assets. Then, the inverse of this variable is taken as an indicator of the firm’s default risk. We recognize that the coefficientsoriginally estimated by MacKie-Mason (1990) may be less relevant in the context of partially privatized SOEs in China. But, giventhe limited number of bankruptcies of Chinese SOEs, we cannot re-estimate this model to adjust it to our sample firms.Nonetheless, we expect that less profitable and less efficient firms are also likely to face a larger probability of default/distress inChina, which tends to adversely affect their stock prices.

20 We find that the median tax rate calculated by using this methodology is only 14.92%, substantially below the official rate of33% for domestic firms. This result is not surprising, as SOE managers in China have proved quite successful in limiting their taxpayments, through negotiation and/or tax evasion (e.g., Fisman and Wei, 2004; Tong and Green, 2005; Huang and Song, 2006). Thisfinding is also in line with cross-country evidence that firms with political connections enjoy lower taxation.





decision to complete a stock offering. Interestingly, the negative (positive) effect of INVERSE Z-SCORE(CASHFLOW/ASSETS) on share issuance is found only for the subsample of listed SOEs in which theState retained its majority stake. The latter outcome is again in line with our conjectures, as theaim of preserving control is likely to be more important for the listed SOEs in which the governmenthas kept its 50% block. Finally, highly leveraged firms prefer selling new shares to raising bank debt,thus indicating that the debt ratio preceding the instrument decision constrains additional borrowingfrom banks. Yet, this outcome only applies to firms in which the Chinese state abandoned its majoritystake after SIP, suggesting once more that this category of firms is treated somewhat less favorably bythe (state-owned) banking system than those in which the State remains the majority shareholder.Finally, the tax rate is totally uncorrelated with the security choice, which is consistent with our ear-lier argument that listed SOEs manage to limit their tax bill in other ways than through the creation ofdebt tax shields. Debt is thus less needed as an instrument to reduce the corporate tax bill.

4.3. Robustness checks

In this section, we first relax the assumption that the decision to raise external funds is made inde-pendently of the choice of instrument, i.e. bank loans or equity. That is, we develop an alternativespecification in which firms simultaneously decide to collect extra finance and fix the security type.The variables determining the decision to seek funds (i.e. the deficit variables) and those affectingthe choice of instrument (i.e. information asymmetries and default risk) are thus combined into a sin-gle multinomial logit model with three choices: raise no funds, take out bank debt, or issue equity.21

The explanatory variables are those that were significant in Tables 5 and 6. The results of this additionalmodel are reported in Table 7. Although the results are largely comparable to our previous findings, theydo provide some additional insights. First, the significant positive coefficient on DEFICIT/ASSETS P 0holds for the decision to borrow from banks as well as for the decision to issue new shares. This findingonce more emphasizes the fact that, given the poor financial market development in China, listed SOEsuse their better access to external finance – whether this is bank debt or equity – to support their tradingpartners with trade credit. Second, we note that for listed SOEs in which the Chinese state retained its50% block after SIP, the cash surplus variable has a negative effect on the likelihood of bank borrowingonly. Possibly, these partially privatized SOEs count on ready access to loans from state-owned bankswhenever they need them. Third, poor-performing SOEs in which the government relinquished itsmajority stake only encounter difficulties in marketing new shares, as the interaction term betweenthe cash deficit, if positive, and the ratio of cash flow to total assets is significant only in the equity equa-tion. This result could reflect the minimum performance requirements for stock offerings imposed by theCSRC.22 In support of the latter argument, we note that CASHFLOW/ASSETS also has a positive and sig-nificant correlation with share issuance in this subsample. Although our reasoning would apply just aswell to firms in which the State retained its majority stake, in that subsample the interaction term isinsignificant.

Next, listed SOEs reporting higher sales growth tend to take out extra bank debt as well as issuenew equity. Yet, sales growth remains statistically significant at conventional levels only for the sub-sample of SOEs in which the government retained its majority stake after SIP. For this subsample, firmsize is negatively and significantly associated with the probability of raising external finance, regard-less of security type. Next, the dividend payout ratio negatively affects the odds of selling new shares,regardless of the firm’s ownership structure. Likewise, the length of time since SOE listing has a sig-nificant negative effect. Consistent with the results in Table 6, we demonstrate that default risk bearsa negative relation with the probability of completing a stock offering for firms in which the govern-

21 The multinomial logit model relies on the assumption of independence of irrelevant alternatives (IIA), which is not alwaysdesirable. We have implemented a Likelihood Ratio test on the inclusive values, obtaining a v2-value of 3.30 (p-value of 0.1921).This test statistic thus indicates that the multinomial logit model is an appropriate specification.

22 The specific criteria vary from year to year, but generally three consecutive years of profits are required to offer new shares inthe public equity market. To qualify for a rights issue, listed firms had to have an ROE exceeding 10% in each of the three previousyears, based on the stock market rules published on January 24, 1996. This performance threshold was lowered somewhat in lateryears, to a three-year average ROE of at least 10% in 1999 and to a three-year average ROE no lower than 6% in 2001. Therequirements for seasoned equity offerings were generally the same as those for rights issues.




Table 6Logit regression results on the instrument choice.

Variable Pred. sign Full sample Government stake P 50% Government stake < 50%

Constant �6.5609(0.9710)

�7.2315(0.9681)

�6.1921(0.9752)

�5.9290(0.9763)

�7.1843(0.9774)

�8.9562(0.9719)

SALES GROWTH � 0.4282(0.0266)

0.4427(0.0211)

0.5128(0.0341)

0.5033(0.0337)

0.7021(0.0977)

0.8708(0.0426)

FIRM SIZE + �0.2720(0.0476)

�0.2462(0.0539)

�0.2832(0.0890)

�0.3064(0.0457)

�0.1657(0.6134)

�0.0968(0.7372)

INTANGIBLES/ASSETS � 2.7989(0.3694)

�1.4076(0.7418)

6.7340(0.2034)

PPE/ASSETS + 0.7376(0.2445)

0.3209(0.6911)

1.8592(0.1491)

DIVIDEND PAYOUT + �0.5084(0.0418)

�0.5412(0.0274)

�0.2272(0.4622)

�0.3344(0.2662)

�1.3119(0.0070)

�1.0872(0.0208)

STOCK VOLATILITY � 2.2270(0.9207)

14.4116(0.6150)

�41.9935(0.3669)

B/H SHARE DUMMY + 0.4058(0.2742)

0.2267(0.6441)

0.5510(0.4421)

TIME LISTED + �0.1809(0.0715)

�0.1486(0.1214)

�0.0991(0.4069)

�0.0848(0.4525)

�0.4602(0.0490)

�0.3482(0.0999)

INVERSE Z-SCORE � �0.9328(0.0029)

�0.5689(0.0198)

�1.1066(0.0061)

�0.7487(0.0151)

�0.9541(0.1316)

�0.3776(0.4488)

CASHFLOW/ASSETS + 3.7647(0.0486)

4.4981(0.0212)

3.9043(0.0933)

4.6283(0.0528)

3.4844(0.3587)

3.5942(0.3285)

LEVERAGE + 1.6796(0.0186)

1.4814(0.0279)

0.8916(0.3019)

1.0625(0.1966)

4.2858(0.0052)

2.7585(0.0350)

TAX RATE � 1.8006(0.5727)

0.3133(0.9377)

5.9048(0.3541)

Number of observations 729 729 496 496 233 233Pseudo R-square (%) 16.82 16.11 15.87 18.91 23.57 21.28

The dependent variable Equity is equal to one if a firm issues new shares in a given year, and zero otherwise. An equity offeringis defined as occurring in every year in which the total number of shares outstanding increases by 10% or more, stock dividendsnot taken into account. Equity was issued in 347 out of 729 firm-year observations in which external finance was raised. Allmodels include industry and year fixed effects. The explanatory variables are described in Table 2. p-Values are reportedbetween parentheses.


ment retained majority control; the debt ratio constrains additional bank debt for firms in which theChinese state is no longer the majority shareholder.

As an alternative robustness check, we estimated the models in Tables 5 and 6 separately for eachof the five years after SIP (not reported). Our main conclusions remain valid for each of these annualcohorts. Next, we divided the sample by size, larger or smaller than the sample median. The results ofsubsequent split-sample regression analyses (not reported) lead to inferences that are qualitativelysimilar to those of splitting the sample based on government ownership.

5. Conclusions

This paper empirically investigates the post-listing financing decisions of 221 Chinese state-ownedenterprises that were partially privatized through SIP in the period 1994–1999. We track these firmsduring their first five years after listing and find that those with a larger cash shortage are more likelyto take out bank loans as well as to issue new shares, but only when the change in non-cash workingcapital is counted as an investment outlay. This correlation arises regardless of the firm’s ownershipstructure, revealing that partially privatized SOEs tend to exploit their better access to external financein order to support their trading partners. We further demonstrate that listed SOEs facing a cash sur-plus are less likely to borrow from banks, but only when the Chinese state retained its dominant own-ership block after SIP. This result suggests that these firms count on ready credit from state-ownedbanks as needed. Overall, these findings are broadly consistent with the pecking order theory, but also




Table 7Multinomial logit regression results on the external finance decision.

Variable Full sample Governmentstake P 50%

Governmentstake < 50%

Bank debt Equity Bank debt Equity Bank debt Equity

Constant 5.0991(0.0279)

8.8691(0.0008)

6.1162(0.0202)

10.0289(0.0016)

�1.6479(0.8033)

4.4348(0.4986)

DEFICIT/ASSETS P 0 6.6901(0.0001)

7.0246(0.0001)

6.3733(0.0035)

7.0847(0.0025)

10.7868(0.0005)

9.3458(0.0042)

DEFICIT/ASSETS < 0 �3.7852(0.0047)

�1.5845(0.2031)

�4.2774(0.0138)

�2.3979(0.1648)

�1.8711(0.4062)

0.3177(0.8726)

DEFICIT/ASSETS P 0 � CASHFLOW/ASSETS �1.1423(0.9402)

22.3063(0.1800)

�16.9062(0.4087)

6.0868(0.7699)

28.2780(0.2460)

51.0460(0.0680)

SALES GROWTH 0.6049(0.0203)

0.9013(0.0008)

0.8977(0.0128)

1.4661(0.0001)

0.5732(0.2412)

0.5716(0.1915)

FIRM SIZE �0.2176(0.0478)

�0.4313(0.0006)

�0.2709(0.0290)

�0.5044(0.0008)

0.1443(0.6548)

�0.1366(0.6681)

DIVIDEND PAYOUT 0.0797(0.7053)

�0.4593(0.0554)

�0.1239(0.6141)

�0.5999(0.0416)

0.1626(0.7339)

�1.0642(0.0253)

TIME LISTED �0.1959(0.0254)

�0.3674(0.0001)

�0.1760(0.0861)

�0.2715(0.0148)

�0.1529(0.4474)

�0.6006(0.0033)

INVERSE Z-SCORE �0.0264(0.8469)

�0.3129(0.0327)

�0.2499(0.1985)

�0.8833(0.0004)

0.0623(0.7794)

0.0960(0.6821)

CASHFLOW/ASSETS 0.9858(0.3192)

2.9407(0.0033)

0.8727(0.4724)

1.6103(0.1927)

�0.4154(0.8356)

5.3812(0.0074)

LEVERAGE �0.5729(0.3052)

0.3304(0.5849)

0.4558(0.5051)

0.8669(0.2616)

�3.8626(0.0013)

�1.3480(0.2540)

PRIOR FINANCING 0.4107(0.0257)

0.5650(0.0057)

0.3252(0.1384)

0.3818(0.1304)

0.1534(0.6860)

0.6052(0.1221)

Number of observations1,105 709 396

Pseudo R-square (%)15.41 16.73 19.82

The dependent variable External financing is equal to zero if no external funds were raised, equal to one if the firm took out abank loan, and equal to two if the firm issued new shares in a given year. A new bank loan is defined as occurring in every yearin which the total amount of bank debt outstanding increases by 10% or more. Likewise, an equity offering is defined asoccurring in every year in which the total number of shares outstanding increases by 10% or more, stock dividends not takeninto account. External finance was obtained in 729 out of 1,105 firm-year observations. The explanatory variables are describedin Table 2. All models include industry and year fixed effects. p-Values are reported between parentheses.


highlight some peculiarities of the Chinese context. Specifically, they emphasize the importance offinancial constraints in a developing country and the lingering role of the state-dominated bankingsystem in the financing of SOEs, even after stock market listing. Given the decline in firm performanceand firm value of SOEs after SIP, as shown in this study and in previous ones (e.g., Sun and Tong, 2003;Wei et al., 2005; Jiang et al., 2009), we conclude that this inter-firm trade finance may not have re-sulted in shareholder wealth maximization at the firms extending it. Our findings therefore imply thatstock market quotation by itself is insufficient to discipline Chinese SOEs. Yet, we find no evidencewhatsoever to support the idea that especially listed SOEs running out of cash from operations, forexample from pursuing political objectives, raise external funds to meet this cash shortfall.

Given the firm’s decision to tap external financial markets, we argue that equity is truly the lastfinancing resort for China’s partially privatized SOEs, as both information asymmetries and default risklead to a preference for debt over equity finance, particularly when the government is interested inpreserving control. Yet, we detect no negative correlation between firm-level information asymme-tries and share issuance. We do find that the partially privatized SOEs facing greater default risk aremore likely to borrow from state-owned banks, but only within the subsample of firms in whichthe State retained its 50% block after SIP. Finally, highly leveraged SOEs are less inclined to borrowfrom banks, but only when the government relinquished its dominant stake. Arguably, our results





indicate that the soft budget constraints typically enjoyed by SOEs may have been prolonged for firmsin which the government retained its majority stake. By contrast, when the Chinese state was willingto give up its dominant block, the firm’s relationships with financial markets, and particularly withstate-owned banks, seem to have developed somewhat more at arm’s length. This conclusion doesnot entail that banks have started to apply economic criteria in their lending decisions for the latterfirms. Indeed, we find no relation between variables capturing firm credit quality – including defaultrisk and cash flow generation – and bank borrowing, again indicating that stock market quotation isnot enough to subject China’s partially privatized state-owned enterprises to market forces.

Acknowledgements

The authors thank Xiaoqiang Cheng, Reinhilde Veugelers, Lihong Wang and Changqi Wu for usefulcomments on an earlier draft of this paper. Nancy Huyghebaert also sincerely thanks KU Leuven (BIL04.01) and the National Natural Science Foundation of China (NSFC-71302072) for financial support,while Qi Quan sincerely thanks for the financial support from National Natural Science Foundation ofChina (NSFC-71173043), Financial Research Center of Fudan University and Social Science Project(2012SHKXQN001) of Fudan University.

References

Ai, C., Norton, E.C., 2003. Interaction terms in logit and probit models. Economics Letters 80 (1), 123–129.Allen, F., Qian, J., Qian, M., 2005. Law, finance and economic growth in China. Journal of Financial Economics 77 (1), 57–116.Altman, E.I., 1968. Financial ratios, discriminant analysis and the prediction of corporate bankruptcy. The Journal of Finance 23

(4), 589–609.Ayyagari, M., Demirgüç-Kunt, A., Maksimovic, V., 2010. Formal versus informal finance: Evidence from China. The Review of

Financial Studies 23 (8), 3048–3097.Bai, C.-E., Liu, Q., Lu, J., Song, M., Zhang, J., 2004. Corporate governance and market valuation in China. Journal of Comparative

Economics 32 (4), 599–616.Booth, L., Aivazian, V.A., Demirgüç-Kunt, A., Maksimovic, V., 2001. Capital structure in developing countries. The Journal of

Finance 56 (1), 87–130.Brennan, M.J., Kraus, A., 1987. Efficient financing under asymmetric information. The Journal of Finance 42 (5), 1225–1243.Carlin, W., Aghion, P., 1996. Restructuring outcomes and the evolution of ownership patterns in Central and Eastern Europe.

Economics of Transition 4 (2), 371–388.Chan, K., Wang, J., Wei, K.C.J., 2004. Underpricing and long-term performance of IPOs in China. Journal of Corporate Finance 10

(3), 409–430.Constantinides, G.M., Grundy, B.D., 1989. Optimal investment with stock repurchase and financing as signals. The Review of

Financial Studies 2 (4), 445–446.Cull, R., Xu, L.C., 2003. Who gets credit? The behavior of bureaucrats and state banks in allocating credit to Chinese state-owned

enterprises. Journal of Development Economics 71 (2), 533–559.Cull, R., Xu, L.C., 2005. Institutions, ownership and finance: the determinants of profit reinvestment among Chinese firms.

Journal of Financial Economics 77 (1), 117–146.Cull, R., Xu, L.C., Zhu, T., 2009. Formal finance and trade credit during China’s transition. Journal of Financial Intermediation 18

(2), 173–192.Demirgüç-Kunt, A., Maksimovic, V., 1999. Institutions, financial markets, and firm debt maturity. Journal of Financial Economics

54 (3), 295–336.Fisman, R., Wei, S.J., 2004. Tax rates and tax evasion: evidence from ‘missing imports’ in China. Journal of Political Economy 112

(2), 471–496.Franks, M.Z., Goyal, V.K., 2003. Testing the pecking order theory of capital structure. Journal of Financial Economics 67 (2), 217–

248.García-Herrero, A., Gavilá, S., Santabárbara, D., 2005. China’s banking reform: An assessment of its evolution and possible

impact, Bank of Spain working paper.Helwege, J., Liang, N., 1996. Is there a pecking order? Evidence from a panel of IPO firms. Journal of Financial Economics 40 (3),

429–458.Huang, G., Song, F.M., 2006. The determinants of capital structure: evidence from China. China Economic Review 17 (1), 14–36.Huyghebaert, N., Quan, Q., 2009. Share issuing privatizations in China: sequencing and its effects on public share allocation and

underpricing. Journal of Comparative Economics 37 (2), 306–320.Huyghebaert, N., Quan, Q., 2011. Ownership dynamics after partial privatization: evidence from China. Journal of Law and

Economics 54 (2), 389–429.Jiang, G., Yue, H., Zhao, L., 2009. A re-examination of China’s share issue privatization. Journal of Banking and Finance 33 (12),

2322–2332.Jones, S.L., Megginson, W.L., Nash, R.C., Netter, J.M., 1999. Share issue privatizations as financial means to political and economic

ends. Journal of Financial Economics 53 (2), 217–253.La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 1999. Corporate ownership around the world. The Journal of Finance 54 (2), 471–

517.


http://refhub.elsevier.com/S1042-9573(13)00036-3/h0005













































Li, H., Meng, L., Wang, Q., Zhou, L.A., 2008. Political connections, financing and firm performance: evidence from Chinese privatefirms. Journal of Development Economics 87 (2), 283–299.

Li, K., Yue, H., Zhao, L., 2009. Ownership, institutions, and capital structure: evidence from China. Journal of ComparativeEconomics 37 (3), 471–490.

MacKie-Mason, J.K., 1990. Do taxes affect corporate financing decisions? The Journal of Finance 45 (5), 1471–1493.Myers, S.C., 1984. The capital structure puzzle. The Journal of Finance 39 (3), 575–592.Myers, S.C., Majluf, N., 1984. Corporate financing and investment decisions when firms have information that investors do not

have. Journal of Financial Economics 13 (2), 187–221.Noe, T., 1988. Capital structure and signaling game equilibria. The Review of Financial Studies 1, 331–356.Perotti, E., 1995. Credible privatization. American Economic Review 85 (4), 847–859.Shleifer, A., Vishny, R.W., 1994. Politicians and firms. Quarterly Journal of Economics 109 (4), 995–1025.Shyam-Sunder, L., Myers, S., 1999. Testing static trade-off against pecking order models of capital structure. Journal of Financial

Economics 51 (2), 219–244.Sun, Q., Tong, W.H.S., 2003. China share issue privatization: the extent of its success. Journal of Financial Economics 70 (2), 183–

222.Tong, G., Green, C.J., 2005. Pecking order or trade-off hypothesis? Evidence on the capital structure of Chinese companies.

Applied Economics 37 (19), 2179–2189.Wei, Z., Xie, F., Zhang, S., 2005. Ownership structure and firm value in China’s privatized firms: 1991–2001. Journal of Financial

and Quantitative Analysis 40 (1), 87–108.























financing decisions after partial privatization in china: can a stock market quotation really...

Documents