hot issue or political issue by tyndale

VIEWS: 11 PAGES: 18

									                          Hot Issue or Political Issue:
    A politically induced cycle in Chinese IPO market ♦
                                             Yifeng Shen

                                              Peigong Li
             The School of Management, Xiamen University, Xiamen, CHINA, 361005

                                          Chin-tan Huang
                                Ming Chuan University, Taipei, Taiwan



Abstract

      It is evidenced that there is a cycle of hot and cold in Chinese IPO market. However,
business cycle and investor sentiment hypotheses provide no explanation to these cyclical
fluctuations. With socialist law origin, the Chinese IPO market was strongly influenced by
political policies by government intervention. This paper argued that political policy is the
primary reason for the cyclical fluctuations in Chinese IPO market. Results of this study
indicate that changes of IPO quotas from political hot seasons to cold seasons lead to the
cycles of hot issue markets in Chinese Market. It shows that an initial sharp decline in the
volume of IPOs over 9 quarters directly after the national congress of The National Congress
of Communist Party (CCP hereafter), followed by a relatively flat period in the next 4
quarters and a gradual increase in successive quarters where the volume of IPOs peaks in the
quarter of national congress of CCP. The paper also finds that the reason for central
government officials manipulate economic policy is to cater to the local government officials,
from whom the central government officials gain political support.

JFL Classification: G32; D72; K22
Key words: hot issue markets; politics; socialist law origin; political business cycle


♦
  We gratefully acknowledge the helpful comments from Hui (Michael) Li, Aydin Ozkan, Jerry T. Parwada, P. J.
Wang and seminar participants at Hull University in Hull, UK and conference participant of 13th Finsia-MCFS
Banking and Finance in Melbourne, Australia. Helpful suggestions were also received from Center for Financial
Studies, the School of Management, Xiamen University. The authors would also like to acknowledge the
financial support by Natural Science Foundation of China (No. 70632001).
Corresponding author: Peigong Li, No.422, Siming Rd., South. Xiamen, China, 361005. Email address:
alex_leeblue@hotmail.com.
1. Introduction

   There is a considerable body of literature on that the initial public offering (IPO) markets
has exhibited cyclical fluctuations from hot to cold (Ibbotson and Jaffe, 1975; Ritter, 1984;
Ibbotson, Sindelar, and Ritter, 1988, 1994; Lowry, 2003; Helwege and Liang, 2004). The
evidences also show that the number of IPOs, as well as initial returns in Chinese IPO
market varies substantially over past 17 years. Figure 1 illustrates this cyclical phenomenon
in Chinese IPO market.
   Why does this anomaly named “hot issue markets” exist? Researchers have contributed
different explanations from business cycle from business cycle hypothesis and information
asymmetry hypothesis to investor sentiment hypothesis. Among of them, business cycle and
investor sentiment have been recognized the two most influential factors while information
asymmetry only bears the secondary importance if it is also important (Lowry, 2003;
Helwege and Liang, 2004).
   Quite different from other mature IPO markets, the Chinese IPO market is severely
regulated by government together with rigid authority intervention (Chan, Wang and Wei,
2004; Fan, Wong, and Zhang, 2007; Megginson and Tian, 2007; Wan and Yuce, 2007).
Entrepreneurs are unable to time their IPOs. It used to be a IPOs mechanism that the
government makes a plan of annual quota for the total number of new shares to be issued in
a year firstly, and then distributed the quota to each local government or the ministries of
central government by whom the affiliated firms will be chosen to go public. The China
Securities Regulation Commission (CSRC) assumes the full responsibilities of the timing of
IPOs according to the objective of government. The time lag between offering date and
listing date usually ranges from 3 days to over 12 years (Chen, Choi and Jiang, 2007). Why
does Chinese central government rigidly control the IPO market? Chi and Padgett (2005, p.)
give a word: “in privatization, the success of any IPO not only affects the individual
company’s reputation, but also the government’s credibility. The government cannot afford
any possible failure in the IPO markets”.
   It is impossible to interpret the cyclical phenomenon without the analysis of incentives of
government officials who make most of the decisions in Chinese IPO market. In democratic
economies, the government officials are nominated by presidents or prime ministers. The
officials only care about their specialties and their personal reputation in labor market.
Unlike their counterparts, the Chinese government officials have very different incentives.
Li and Zhou (2005) note that the term limits of government officials in China is relatively
invisible and instable, because government officials can be rotated to other districts or
promoted to higher positions occasionally. More importantly, there is no partisan
competition in China, and the government officials can maintain their power even when the
chairman or prime minister retires. So the primary consideration of Chinese government
officials is to keep their power and get promotion. The purpose of this paper attempts to link
political incentives of government officials to cyclical fluctuations in Chinese IPO market.




                                               2
                          The distribution of IPO volume and initial returns
                     80                                                                                20
                     70                                                                                15
    Volume of IPOs

                     60                                                                                10




                                                                                                             Initial Return
                                                                                                       5
                     50
                                                                                                       0
                     40
                                                                                                       -5
                     30
                                                                                                       -10
                     20                                                                                -15
                     10                                                                                -20
                     0                                                                                 -25
                          1990.12
                          1991.12
                          1992.12
                          1993.12
                          1994.12
                          1995.12
                          1996.12
                          1997.12
                          1998.12
                          1999.12
                          2000.12
                          2001.12
                          2002.12
                          2003.12
                          2004.12
                          2005.12
                          2006.12
                          2007.12
                           1991.6
                           1992.6
                           1993.6
                           1994.6
                           1995.6
                           1996.6
                           1997.6
                           1998.6
                           1999.6
                           2000.6
                           2001.6
                           2002.6
                           2003.6
                           2004.6
                           2005.6
                           2006.6
                           2007.6
                                       Volume of IPOs            Initial Returns

Fig. 1. Time series of monthly IPO volume and initial return, 1990-2007. We delete the initial return of December
1990 because it is extremely high up to 289.69%. The discrete distribution of initial returns is due to the discrete
distribution of IPO volume.


   It is well documented by economists that the cycles of economic activities are not only the
sum of micro-behavior of individual firms, but substantially affected by macro-political
factors. This gives rise to the political business cycle theory (PBC) since Kramer (1971),
Tufte (1975) and Nordhaus (1975). Nordhaus’s model predicts that the incumbent tends to
manipulate economic policy to cater to voters in order to be reelected the presidency. The
economic condition turns well as the election are approaching and reach peaks in election
period, then declines afterwards. This opportunistic behavior-led cycle repeats in every 4
years in coincidence with the president election cycle.
   Similarly, we find one of the most compelling reasons for the dramatic swings in issuance
in Chinese IPO market to be opportunistic behavior of government officials. The socialist
law origin of China makes it possible and legal for government officials to intervene IPO
market by policy manipulation, which in turn lead to the cyclical fluctuations of IPO market
in China. In order to gain political support and consensus, the central government officials
try to satisfy the needs of local authorities, especially those in districts with strong economic
influences. Unlike incentives of individual firms, the peculiar incentive behind officials
forms unique scenario in China: the number of IPOs climbs before the national congress of
Communist Party of China and declines afterwards. 1 The IPO cycle spans 5 years in
coincidence with the CPC congress cycle. Evidence indicates that an initial sharp decline in
the number of IPOs over 9 quarters directly after the National Congress of CPC, followed by
a relatively flat period in the next 4 quarters and a gradual increase in successive 7 quarters
where the number of IPOs peaks in the quarter of CPC congress.
   In addition to the PBC literature, this study also relates to law and finance literature.
Regulations of financial markets are deeply rooted in the legal structure of each country and

1
 There is no president election in China. The National Congress of CPC, however, plays the similar role. We
will discuss it in detail in section 3.

                                                          3
in the origin of its laws, while law origin plays a significant role in the development of
financial markets (La Porta et al., 1997, 1998, Demirguc-Kunt and Maksimovic, 1998). The
socialist law origin features and facilitates the intervention of government officials whose
incentives may not coincide with social welfare. This legitimate intervention distorts IPO
mechanisms heavily and hence retards the development of Chinese financial markets. To the
best of knowledge, this is the first paper to empirically study Chinese hot issue markets from
the perspective of political perspective.
   The remainder of the paper is organized as follows. Section II reviews the previous
studies on IPO cycles. In section III, we describe the institutional background of Chinese
IPO market, and introduce political business cycle theory. Section 4 describes the data and
variables. In section 5 we report the results of empirical findings. Section 6 concludes.



2. IPO Cycles: Theory and Evidence

   Despite the large literature on IPOs, including Ibbotson and Jaffe (1975), Ritter (1984)
and Ibbotson, Sindelar and Ritter (1988), have shown that there are pronounced cycles in the
average initial return of new issues and the IPO volume, little attention has been addressed
to the underlying cause of this variation. In a recent study, Lowry and Schwert (2002) also
find periods of high initial returns tend to be followed by spurts of IPOs, which are
themselves followed by periods of relative lower initial returns. They confirm with other
researchers that there is an apparent IPO cycles. Pastor and Veronesi Benninga develop a
model which predicts that IPO waves are preceded by high market returns, followed by low
market returns and accompanied by high stock prices. Their paper links IPO volume to
market returns directly. Choe, Masulis and Nanda (1993) first link the magnitudes of the
excess announcement period stock returns to the business cycle. They find that the relative
volume of equity issues in the announcement period is significantly positively related to the
excess announcement period stock returns. And business cycle variables or stock issue
activity and have significant explanatory power in accounting for the magnitudes of the
abnormal stock returns. Helmantel and Sarig (2005) report that changes in macroeconomic
conditions simultaneously affect multiple industries, as a result, the market exhibits strong
cross-sectional correlation in the profitability of firms. When one firm finds it optimal to go
public, it is also optimal for other firms to do so. The clustering in time of IPOs, the industry
concentration of IPO waves, and the coincidence of IPO waves with relatively high market
prices are proposed to explain the long-term underperformance of IPO firms. On the other
hand, Lowry (2003) proposes three hypotheses for variation in aggregate IPO volume: the
capital demands hypothesis, the information asymmetry hypothesis, and the investor
sentiment hypothesis. He then argues that companies’ demand for capital and the level of
investor sentiment are the two major factors that help to explain significant amount of the
variation in IPO volume. 2
   Loughran and Ritter (1995), Rajan and Servaes (1997) and Lerner (1994) attribute IPO

2
  The business cycle story underlines the firms’ demand of capital to finance their new projects when economy
expands. However, there is conflict evidence on the usage of proceeds raised from equity issuance. Pagano,
Panetta and Zingales (1998) and Loughran, Ritter and Rydqvist (1994) find no evidence to support that firms
issue equity to finance their projects, while Kim and Weisbach (2008) present opposite evidence.

                                                      4
cycles to investor sentiment, which provides “transient windows of opportunity” for
managers to issue overvalued equity. The investor sentiment hypothesis argues that invetors
are overly optimistic during certain periods and are willing to pay a high price for equity
than they are worth. As a result, it is relatively cheap for firm to go public. Consequently,
many firms may find it optimal to issue equity during these periods. Similar to these
arguments, Pagano, Panetta and Zingales (1998) suggest that companies incline to issue new
shares when the average market-to-book ratio (M/B) of public firms in their industry is
higher. They interpret their findings as managers’ taking advantage of industry-wide
overvaluation. In a survey conducted by Graham and Harvey (2001), 62.6% of respondents
think resent stock price rise is very important when they make decisions about issuing
common stock.
   Other researchers explain cyclical behavior of equity issuance from information
asymmetry perspective. In Maksimovic and Pichler (2001)’s model, pioneering IPO firms in
an emerging industry bear the costs of information spillover, especially when the entry risk
is high. If they decide to enter this industry, other firms will follow them with zero initial
cost of research. Alti (2005) argues that firms find it optimal to go public when mispricing is
minimized. High offer price realizations for pioneering IPO firms help reflect informed
investors’ information, hence trigger a clustering of subsequent IPOs. According to the
information spillover models, IPOs should cluster in industry level during hot issue markets.
Helwege and Liang (2004)’s findings, however, contradict with these models. They report
that the five most frequent industries of IPO clustering are identical, no matter the market is
hot or cold; hot issue markets do not show a heavier clustering than cold market. They
conclude that information asymmetry does not help explain hot issue markets.
   These theories and empirical evidence provide profound insights in IPO cycles though our
understanding of these mixed findings is limited. Chinese IPO market, however, is rigidly
controlled by central government. This unique institutional background is prominently
distinct from the base on which these theories and evidences are created. This paper seeks to
review whether these theories can apply directly to Chinese IPO market or we should look
more specifically at other alternative explanations.



3. Chinese IPO market: intervention and public business cycle

  An official IPO market emerged in China with the establishment of Shanghai Stock
Exchange in December 1990 and Shenzhen Stock Exchange in April 1991. 3 With the
socialist law origin, the Chinese IPO market was strongly intervened by government from its
very early stage, (Megginson and Tian, 2007)

3.1. Government Intervention and IPO volume

  In the early 1990s, the newborn Chinese IPO market was under the regulation of Shanghai
and Shenzhen municipal governments. The central government regained its power on IPO

3
  A number of firms had their IPOs in 1980’s before the market establishment. The shares offered during this
period were mostly by private placement without establishment of secondary markets.

                                                      5
market since the establishment of the State Council Securities Commission (SCSC) and the
China Securities Regulation Commission (CSRC) in October 1992. 4 Both the Interim
Regulations on Share Issuing and Trading issued in April 1993 and the Company Law
approved in December 1993. They all indicated clearly that share issuance must be approved
by government authorities. The annul amount of new shares to be issued was decided by the
State Planning Committee (SPC) and the SCSC, the later then distributed the quotas to each
province and ministry of central government who own a great number of state owned
enterprises (SOEs). Local governments and ministries selected firms and filed them to
CSRC for final approval. This “quota system” is slightly changed since 1996 to “direct
volume indicator management”. Under the new system, the SPC and the SCSC decided the
annual amount of new shares to be issued by the same mechanism. On the basis of this
pre-determined number of shares, the CSRC distributed specific number of IPOs rather than
amount of shares to local governments and ministries. The criteria for distribution by the
CSRC is based on certain quotas or volume indicators, as Chen, Wang and Wei (2004)
describes, in the process of development of some key industries and the local economy, the
CSRC has to balance across industries and local governments. During this period, a good
many companies bribed local officials or even national officials in order to get the limited
place.
   “Authorization System” formally replaced “direct volume indicator management” in April
2001.5 Under this system, investment banks were empowered to recommend companies that
satisfied the listing standards to go public under the approval of CSRC. However, most of
the financial intermediaries were controlled or even owned by local governments who had
strong incentives to have their affiliated SOEs to go public. Under the direction of local
governments, these financial intermediaries were greatly stimulated to expand. In order to
make a profit, these intermediaries even make up false financial reports for unqualified
candidates. As a result, the number of IPO candidates rose disproportionately to their quality
on profitability at one time. The CSRC was forced to recentralize its control on quotas, and
rationed every qualified investment bank a certain number of volume indicators. Only when
one company succeeded to go public, the investment bank could undertake another IPO
application. The operation of “Authorization System” means that the central government
failed to transform into market-centered direction system. During this period, the investment
banks began to focus on their reputation although the IPO process is still under direct control
of central government.

3.2. Government Intervention and IPO pricing

   There was no regulation on pricing of new shares before the two main stock exchanges
were established, and the new issues often offered at their par value. Started from 1994,
however, several new fixed price methods were authorized by CSRC and introduced into the
Chinese IPO market. According to these pricing methods, the offer price is the product of
after-tax profit per share and price-earnings ratio (P/E ratio). Usually, an issuer had

4
  Before SCSC and CSRC merged and retain the old named of CSRC, the SCSC was the highest regulatory body
in China, while the CSRC became the executive branch of SCSC.
5
  The volume indicator of year 1997 validated until 2001 although the CSRC did not announce new volume
indicator since the Securities Law became effective in 1999.

                                                    6
discretion to set the P/E ratio based on the similar firms already listed on stock markets.
However, as Megginson and Tian (2007) point out that the offer price, especially the P/E
ratio, has to be authorized by CSRC who set the ceiling cap of P/E ratio between 15 and 20,
while the average P/E ratios for the period from 1993 to 1999 were 32.19 and 29.85 for
Shanghai and Shenzhen Stock Exchanges, respectively (Wan and Yuce, 2007).
   The Securities Law became effective in July 1999, and stipulated that the pricing methods
of IPOs must take bookbuilding or other auction-like mechanisms. The fixed price methods
became increasingly uncommon since then.

3.3. The political business cycle theory

   It is well documented by economists that the cycles of economic activities are not only the
sum of micro-behavior of individual firms, but substantially affected by macro-political
factors. That is why political business cycle theory gains increasing popularity since Kramer
(1971), Tufte (1975), Nordhaus (1975), Lindbeck (1976), McRae (1977) and Fair (1978).
   Although a serious description to this consideration can be traced back to Kalecki (1943),
the first theoretical model was presented by Nordhuas (1975).6 In his opportunistic model,
the incumbent will lower the unemployment rate until the election eve, after which the
unemployment rate reaches its “purely myopic point”. After the election, the winner will
immediately raise unemployment to a relatively high level in order to combat climbing
inflation. As the next election approaches, this politically induced cycle will repeat itself.
   Unlike presidential election in other democratic economies, there is no partisan
competition in China. However, in every 5 years, there is a National Congress of CPC in
China, which plays the similar role as presidential election in other democratic economies.
Key positions and national strategies are nominated and negotiated on the congress. The
criteria based on which the officials get promotion is their performance and political
popularity amongst government officials. So they will probably manipulate economy
policies to cater to these influential officials. IPO quota is a subcategory of monetary policy
in a sense. So we should expect a politically driven IPO quota cycle in China as
opportunistic model predicts. Specifically, the volume of IPOs climbs up as the national
congress approaches, peaks in the period of CPC congress and declines directly afterwards.



4. Data and Variable Construction

   Our database covers 1576 listed companies that issued A-shares in Shanghai Stock
Exchange and Shenzhen Stock Exchange from December 1991 to December 2007. Data on
delisted companies will be automatically deleted from original database. Since the Chinese
central authority start to intervene securities market from the date of October 1992, we
screen out our IPO data before that date. All IPO data are resorted from China Center for
Economic Research (CCER) database and a cross-reference from the China Stock Market
and Accounting Research (CSMAR) database. These commercial databases are based on

6
  Almost at the same time, Lindbeck (1976) and McRae (1977) presented similar models, which are finally
named opportunistic PBC model.

                                                   7
company reports, audited by charted accountants. To obtain the quarterly data on GDP and
fixed investment, we used the Wind database provide by Wind Company. We also manually
collect the data missing from these databases from IPO prospectuses and listing
announcements filed with the CRSC.

4.1. Initial returns

   Along with Ibbotson and Jaffe (1975), Ritter (1984), and Ibbotson, Sindelar, and Ritter
(1988, 1994), the initial return equals the percentage change between the offer price of new
issues and the first closing price. We first calculate the equal-weighted monthly IPO initial
returns ( IR EW ) adjusted by market index.
             t
                                                    7
                                                        This procedure is described as follows,

                 1 n  CPit − OPit CI it − OI it          
      IRtEW =      ∑
                 n i =1  OPit
                                  −                       ,
                                                                                                          (1)
                                      OI it              

where CPit and CI it are the first day closing price and market index, respectively, while

OPit and OI it are the offer price and market index on the offer day, respectively.

n denotes the total number of IPOs in month t .

4.2. IPO Volume

  Both Helwege and Liang (2004) and Alti (2006) use the three-month centered moving
average technique to smooth seasonal variation on IPO volume. Lowry and Schwert (2002)
uncover that IPO volume is non-stationary and suggest to use the first difference of IPO
volume or deflate IPO volume by the total number of public firms at the end of the prior
period. In our sample, the augmented Dickey-Fuller test rejects the null hypothesis that IPO
volume is non-stationary. So we turn to the procedures of Helwege and Liang (2004) and
Alti (2006).

4.3. Political hot issue

   Four National Congresses of CPC have been held at October 1992, September 1997,
November 2002 and October 2007, respectively since Chinese IPO market established in
December 1990. We first define the national congress as political hot issues in which key
positions and national strategies will be nominated and negotiated. Then we term the periods
of 12 months before and after these political hot issues (totally 24 months) as political hot
seasons, while the periods between two political hot seasons are political cold seasons.




7
 Researchers also use value-weighted monthly initial returns. However, the existence of non-tradable shares
makes it difficult, if not impossible, to calculate the value of public firms in China. So we do not consider this
measure of monthly initial returns.

                                                         8
4.4. Descriptive statistics

   Table 1 provides descriptive statistics on IPO monthly initial returns and volume. In total,
these 1576 firms raised amount of X billion yuan caipial. The average proceeds per IPO was
8.44 million (median=6 million) and the average initial return adjusted by market index was
197% (median=126%). Consistent with Ibbotson and Jaffe (1975), Ritter (1984), and
Ibbotson, Sindelar, and Ritter (1988, 1994), monthly initial returns as well as IPO volume
per month exhibit strong autocorrelation. The first order autocorrelation coefficients for the
time series monthly initial returns are 0.54. The autocorrelation for monthly volume of IPOs
is even stronger, with a first order autocorrelation coefficient of 0.69. All autocorrelation
coefficients are significantly different from zero. The high degree of autocorrelation for both
monthly initial returns and volume of IPOs indicates the existence of hot issue markets in
Chinese IPO market.

Table 1
Descriptive statistics on IPO
This table provides descriptive statistics for a sample of 1576 IPO firms between 1991 and 2007. The mean,
median, standard deviation, minimum, and maximum of the number of IPOs per month (NIPO) and the
equal-weighted percentage initial returns (IRtEW). Generally, the initial return is the percentage return from the
offer price to the closing price on the first trading day. ρ is the autocorrelation coefficient for 3 lags (ρ1 to ρ3).

Variables        Mean      Median     Std. Dev     Max.        Min.        ρ1           ρ2           ρ3         Obs.

 NIPO             8.44        6         7.62         49         0       0.69***      0.47***       0.38***      183

 IRtEW            1.97      1.26        1.97        9.97       -0.58    0.54***      0.33***       0.21***      162

***, **, * Significance at the 1%, 5% and 10% levels in two-sided significance test, respectively.




5. Empirical Results and Discussion

  China stock market has a very distinctive institutional background and it gives a great
challenge to the rapidly growing literature on the IPO cycle. In this section, we first test
whether the popular explanations provided in the literature apply to Chinese IPO market.
Given the influential studies conducted by Lowry (2003) and Helwege and Liang (2004),
two hypotheses of investor sentiment and business cycle will be tested. Then we turn to the
hypothesis of politically led cycle based on opportunistic PBC models.

5.1. Investor sentiment, business cycle and the IPO cycle in China

Following Lowry and Schwert (2002), we construct a VAR (3) to test the Granger causality
between monthly initial returns and volume of IPOs. The results are shown in table 2. The
first column and third column of Table 2 show that both time series of IR and NIPO pass the
augmented Dickey-Fuller test. We find that neither initial returns nor volume of IPOs is the
Granger cause to the other. Evidence seems to suggest that investor sentiment hypothesis can
not explain the IPO cycle in China.




                                                           9
Table 2
Vector autoregressive on IPO volume and IRew
We construct a VAR (3) model on monthly IPO volume and IRew from October 1992 to December 2007 to test
the Granger causality between NIPO and IRew.

                                                      IR EW
                                                         t
                                                                                               NIPO
Dependent variable
                                         Co-efficient             t-stat          Co-efficient             t-stat
                                              (1)                  (2)                 (3)                  (4)
     Constant                                0.578                2.10                2.726                2.52
     AARt-1                                  0.416                5.05                0.326                1.01
     AARt-2                                  0.053                0.61               -0.167                -0.48
     AARt-3                                  0.036                0.47                0.427                1.42
     NIPOt-1                                 0.019                0.90                0.642                7.52
     NIPOt-2                                -0.002                -0.03              -0.139                -1.37
     NIPOt-3                                 0.010                0.47                0.118                1.38
     R2                                       0.27                                    0.40
Sample Size                                             143                                      143
Granger F-tests
     Lagged NIPO                                        0.62
     (p-value)                                          0.60
     Lagged AAR                                                                                  1.40
     (p-value)                                                                                   0.25




   Lowry (2003) employs GDP growth as the proxy for the aggregate demand of individual
firms. She terms GDP growth as the percentage change in GDP between quarter t and
quarter t+3. Different from Lowry, we simply construct a VAR (3) to test the Granger
causality between GDP growth and NIPO. We first calculate the GDP growth from quarter 4
of 1992 to quarter 4 of 2007. Unfortunately, augmented Dickey-Fuller test provides some
evidence that the new time series is non-stationary, so we take the second difference. We
report VAR results in table 3. Evidence indicates that there is no Granger causality between
these two variables. It seems that business cycle contributes nothing to IPO cycle in China.

5.2. Political cycle and the IPO cycle in China

  By the Chapter of CPC, the National Congress is held every five years. To more
specifically, there are 60 months (5 years) between every two national congresses. 8 We first
divide the 60 months into two halves and each half includes 30 months. The later is
partitioned into another two periods where first period is comprises 12 political hot months,
while period two consists of 18 political cold months. The scheme is described in figure 2.




8
  In fact, almost all the national congress is held in the 4th quarter every year, but not fixed in a specific month.
We roughly use the number of 60 months for convenience.

                                                          10
Table 3
Vector autoregressive on IPO volume and GDP growth ( ∆ GDP)
We construct a VAR (3) model on quarterly IPO volume and         ∆ GDP from 4th quarter, 1992 to 4th quarter, 2007
to test the Granger causality between NIPO and ∆ GDP. Time series of ∆ GDP is non-stationary and do not pass
the unit root test, so we first take the second difference of ∆ GDP to form stationary series.
                                                        ∆ GDP                                  NIPO
Dependent variable
                                            Coef.               t-stat              Coef.               t-stat
Regressors
     Constant                               0.000               0.001              10.523               2.527
     ∆ GDPt-1                              -0.998              -135.7              -0.969              -1.807
     ∆ GDPt-2                              -0.994              -104.4              -0.441              -0.636
     ∆ GDPt-3                              -0.992              -131.9               0.168               0.306
     NIPOt-1                                0.001               0.492               0.561               4.092
     NIPOt-2                               -0.001              -0.375               0.094               0.599
     NIPOt-3                                0.000              -0.078              -0.082              -0.592
       2
     R                                      0.999                                   0.380
Sample Size                                           59                                      59
Granger F-tests
     Lagged NIPO                                     0.10
     (p-value)                                       0.96
     Lagged AAR                                                                              1.25
     (p-value)                                                                               0.30




      The opportunistic PBC models posit an IPO policy cycle. Under this hypothesis, we
should observe that monthly volume of IPOs rises from political cold seasons to hot seasons
as the national congress approaches, and declines directly after the congress.
      We employ Mann-Whitney U test to examine if there is an IPO policy cycle in China.
The results are reported in table 4. Panel A of table 4 shows that IPO volume climbs up high
from political cold to hot seasons before National Congress of CPC, and more monthly IPO
volume in period 1 than that of period 2 with one percent statistically significance. Panel B,
however, suggests that there is no significant difference between monthly IPO volume in
political hot and cold seasons after the congress. The results are conflicted with PBC
hypothesis. Panel C shows that more monthly IPO volume in pre-congress period than that
in post-congress period, significantly at ten percent statistical level. If IPO volume is
symmetrically distributed, then the peak in IPO volume can also be evidenced in the
pre-congress period. As can be seen from our data, it hardly exhibits symmetrical
characteristics between political hot and cold seasons. In summary, the results across the
Table 4 are relatively strong in support the hypothesis that there is an IPO cycle in China.




                                                        11
                          The first half                                                The second half

         Time period 1              Time period 2                      Time period 3              Time period 4


National                                                                                                            Next
                   12th       month                 30th      month                 12th month before
Congress                                                                                                            National
                   after congress                   after congress                  next congress
of CPC                                                                                                              Congress
                                                                                                                    of CPC
                                                          60 months
  Fig. 2. The Scheme of the National Congress of CPC, 1990-2007. The 60 months are separated into two halves
  and each half includes 30 months. Each sub-period is further partitioned into another two periods where first
  period is comprised of 12 political hot months; while second period consist of 18 political cold months.




     Another interesting finding, as is reported in table 4, is that monthly initial returns are
  significantly higher in post-congress political hot seasons than in the next 18 months of cold
  political seasons over 15 years. This finding might suggest that individual investors in the
  secondary market have not recognized this cyclical phenomenon in that long period or even
  though have they identified this cycle, but still chosen overreaction in political hot seasons.



  Table 4
  Mann-Whitney U Test on political cycle hypothesis, 1992- 2007
  We first divide every 60 months into four sub-periods as graphed in figure 2. Then we calculate monthly IPO
  volume and initial returns in these sub-periods. Mann-Whitney U test is employed to compare medians of NIPO
  and IRew in every two consecutive sub-periods.
  Panel A The second half of 60 months during two consecutive National Congress of CPC
                                                                        NIPO                           IR EW
                                                                                                          t
                  Time periods                     Obs.
                                                              Median           Z-value         Median          Z-value
  12 month before National Congress                 39           10                             1.47
                                                                               -3.60***                         -0.62
  Another 18 months in the second half              52           4                              1.53
  Panel B The first half of 60 months during two consecutive National Congress of CPC
                                                                        NIPO                           IR EW
                                                                                                          t
                  Time periods                     Obs.
                                                              Median           Z-value         Median          Z-value
  12 month after National Congress                  42           6.5                            1.22
                                                                                -0.68                          -2.77***
  Another 18 months in the first half               53           6                              0.89
  Panel C 12 months before and after national congress respective
                                                                        NIPO                           IR EW
                                                                                                          t
                  Time periods                     Obs.
                                                              Median           Z-value         Median          Z-value
  12 months before National Congress                36           10                             1.48
                                                                               -2.32**                          -0.52
  12 months after National Congress                 38           7                              1.22
  ***, **, * Significance at the 1%, 5% and 10% levels in two-sided significance test, respectively.


                                                            12
5.3. A U-shaped IPO volume cycle

   Since each period might exhibits sinusoidal fluctuations in IPO volume, there may be a
monotonicity within each half of the 60 months. To avoid this problem, we will follow Grier
(1987, 1989), who uses a 15-quarter polynomial distributed lag (PDL) on an election
dummy variable to estimate the optimal cyclical pattern on monetary policy and finds that a
U-shaped cycle where money growth peaks in the quarter of presidential election.9
   Firstly, we construct a dummy variable (CPC) that equals one in the quarter of CPC
National Congress, and zero otherwise. Secondly, we estimate a second-order, 19-quarter
polynomial distributed lags on CPC dummy. This yields a unique lag coefficient for each
quarter of the 5-year congress cycle. Considering the critique of Heckelman and Wood
(2005), we also estimate third and fourth order PDLs to avoid imposing the restriction of a
single turning point. All these two sets of polynomial lag coefficients have similarly general
shape as the second-order one, but the second-order PDL give the best statistical fit and
further results will based on it.
   Using quarterly IPO volume data from the fourth quarter of 1992 to the last quarter of
2007, we estimate the IPO volume cycle with second order PDL technique. Results are
reported in table 5 and graphed in figure 3. In Table 5, the results show that there is a U
shaped IPO volume cycle, statistically significant for each quarter. As can be seen from
Figure 3, there is an initial sharp decline in the number of IPOs over 9 quarters directly after
the National Congress of CPC, followed by a relatively flat period in the next 4 quarters and
a gradual increase in successive 7 quarters where the number of IPOs peaks in the quarter of
CPC congress. This cycle repeats every 5 years with the CPC congress cycle.




9
  Grier (1987) posits that “by using a dummy variable/ polynomial distributed lag technique, …the data can
‘speak for itself’ about the existence , shape, and periodicity of the election cycle”. Besides, PDL dummy
technique can help preserve degree of freedom and avoid multi-colinearity compared with using many lagged
variables in the regression equation.

                                                    13
Table 5
Lagged distribution of IPO quotas on CPC dummy
We construct a dummy variable “CPC” that equals one in the quarter of CPC National Congress, and zero
otherwise. We estimate a second-order, 19-quarter polynomial distributed lags on CPC dummy. This yields a
unique lag coefficient for each quarter of the 5-year congress cycle.
Variable                                                                                Coefficient                                                                                         T-Statistic
PDL01                                                                                           19.15                                                                                                    6.01
PDL02                                                                                           -0.03                                                                                               -0.09
PDL03                                                                                            0.08                                                                                                    1.23
Lagged distribution of dummy variable ”CPC”
Variable                            Coefficient                                                 T-Stat.                                            Variable                                 Coefficient                                           T-Stat.
CPC-19                                         27.06                                             5.06                                              CPC-9                                           19.15                                           6.01
CPC-18                                         25.52                                             5.83                                              CPC-8                                           19.27                                           6.17
CPC-17                                         24.16                                             6.67                                              CPC-7                                           19.55                                           6.47
CPC-16                                         22.96                                             7.35                                              CPC-6                                           19.99                                           6.89
CPC-15                                         21.92                                             7.63                                              CPC-5                                           20.60                                           7.29
CPC-14                                         21.05                                             7.45                                              CPC-4                                           21.38                                           7.44
CPC-13                                         20.34                                             7.01                                              CPC-3                                           22.32                                           7.15
CPC-12                                         19.80                                             6.56                                              CPC-2                                           23.42                                           6.46
CPC-11                                         19.42                                             6.21                                              CPC-1                                           24.69                                           5.64
CPC-10                                         19.20                                             6.03                                               CPC                                            26.12                                           4.88
 2
R = 0.037                    sum of lags = 437.91 (T-stat = 10.40)




     29
     27
     25
     23
     21
     19
     17
     15
     13
                                                                                                                                                                                                                        3rd before
                                                                                                                                                                                                                                     2nd before
                                                                                                                                      9th before
                                                                                                                                                     8th before
                                                                                                                                                                  7th before
                                                                                                                                                                               6th before
                                                                                                                                                                                            5th before
                                                                                                                                                                                                           4th before
           1th after
                       2th after
                                   3th after
                                                4th after
                                                            5th after
                                                                        6th after
                                                                                    7th after
                                                                                                 8th after
                                                                                                             9th after
                                                                                                                         10th after




                                                                                                                                                                                                                                                  1st before
                                                                                                                                                                                                                                                               NCCPC




Fig. 3. The policy cycle on IPO quotas, 1990-2007. This U-shaped curve is graphed according to the lagged
distribution coefficients in table 5. “ith after” represents ith quarter after National Congress of CPC, while “ith
before” represents ith quarter before National Congress of CPC. NCCPC represents the quarter in which National
Congress of CPC is held.


                                                                                                                         14
5.4. Who benefit from this IPO policy cycle?

   The opportunistic presidents manipulate economic policies to cater to voters in order to
raise the probability to be reelected. So to whom does the IPO policy cycle cater to?
Basically, there are two categories of target people can be benefit from IPO policy cycle.
The first category is the investors in primary market, and the second category is the local
government officials. The investors in primary market gain economically while local
government officials benefit politically. In Chinese IPO market, especially at its early stage,
the central government severely controlled the IPO quota as well as the offer price. For those
Investors playing actively in primary market usually have special relation with government
officials. Did the officials lower the offer price during political hot seasons to help this group
of vested interests to make a profit in primary market?
   Before the Securities Law became effective in 1999, the offer prices were under the price
caps set by central government officials with PE ratios between 15 and 20, far below the
actual after-market PE ratios.

Table 6
Offer prices in primary market, 1990-2007
We first divide our sample into two periods. The first period is between January 1994 and June 1999 and the
second period is between July 1999 and December 2007. Following the scheme described in figure 2, we further
partition the two periods into 4 sub-periods.
                                          offer price/book value per share
                                                        Before 1999.7                       After 1999.7
                Time periods
                                                   Median            Z-value           Median          Z-value
12 month before National Congress                    0.84                                1.07
                                                                     -2.18**                           -2.34**
Another 18 months in the second half                 0.67                                1.27
12 month after National Congress                     0.86                                1.18
                                                                     -1.97**                               -0.44
Another 18 months in the first half                  0.73                                1.14
12 months before National Congress                   0.82                                1.06
                                                                      -0.52                                -1.27
12 months after national congress                    0.81                                1.16
***, **, * Significance at the 1%, 5% and 10% levels in two-sided significance test, respectively.




   However, the Security Law effectively constrains the intervention of price control. It is
argued that for investors to gain in the primary market, government officials may
deliberately lower the offer prices in political hot season. We therefore divide our sample
into two periods. The first period is between January 1994 and June 1999 and the second
period is between July 1999 and December 2007.10 Following the scheme described in figure
2, we further partition the two periods into 4 sub-periods. As can be seen from Table 6,
during the period before July 1999, the offer prices sanctioned by government officials are
statistically higher in political hot seasons than that in cold seasons. While during the period

10
   As discussed in section 3.2, fixed price methods were authorized by CSRC and introduced into the Chinese
IPO market in 1994. So our analysis on offer price starts from January 1994, a little different from the starting
point of analysis on IPO volume and initial returns.

                                                        15
       of low government intervention after 1999, offer prices were lower in political hot seasons
       than that in cold seasons, which contradicts with our hypothesis.
          One of the purposes to found IPO market is to raise deadly needed capitals for SOEs. A
       local company can allocate the resources to fuel its development if going public successfully.
       It is also a good deal for local government officials to encourage more companies to go
       public, not only for economic consideration, but political promotion for his/her superior
       performance. However, the IPO quotas were under direct control of central government
       officials who ration quotas to 31 provinces. This process leads severe unbalance between the
       demand of local governments and supply of central government. As predicted by
       opportunistic models, central government officials tend to manipulate economic policies to
       cater to voters who are local government officials in China. The distribution of political and
       economic power amongst provinces is unbalance. Usually, economical-boom areas have
       stronger political influence in China. During the period before National Congress of CPC,
       the central government officials might bribe the influential local officials for political
       support, and hence give them more IPO quotas.
          Figure 4 illustrates IPO quota ratios amongst 31 provinces from political hot to cold
       seasons. We find that the first 6 provinces with the highest quota ratios are identical both in
       political hot and cold seasons, and also the most prosperous provinces in China. Similarly,
       the last 3 provinces with the lowest quota ratios are identical, and they are also the least
       developed provinces. The graph also suggests that there are 13 provinces with similar quota
       ratios during political hot and cold seasons, while 8 provinces with higher quota ratios
       during political hot seasons, and 10 provinces with higher quota ratios during political cold
       seasons. In the 8 provinces, only Henan is less developed with other 7 provinces strong at
       economy as well as politics, while 8 out of the 10 provinces with lower quota ratios during
       political hot seasons are less developed. Clearly, the central government officials ration more
       IPO quotas to prosperous districts during political hot seasons while less to depressed areas.



                                   IPO quota ratios from political hot to cold seasons amongst provinces in China


0.12

0.10

0.08

0.06

0.04

0.02

0.00
                                                                                                    Neimenggu
                                Gansu




                                                   Hainan




                                                                             Jilin




                                                                                                                                                        Xinjiang


                                                                                                                                                                   Zhejiang
                                                                                                                                              Tianjin
        Anhui




                                         Guangxi




                                                            Shanxi


                                                                     Hubei




                                                                                          Jiangxi




                                                                                                                Qinghai


                                                                                                                          Shanxi


                                                                                                                                   Shanghai
                 Chongqing




                             IPO quota ratios in political hot seasons               IPO quota ratios in political cold seasons


       Fig. 4.           IPO quota ratios from political hot to cold seasons amongst provinces in China.


                                                                              16
6. Conclusion

   This paper provides a new explanation to IPO cycles. We find that the Chinese hot issue
markets coincide with political hot seasons. This stylized fact can be explained by
opportunistic behavior of incumbent government officials, while the popular explanations of
investor sentiment and business cycle may not directly apply to Chinese IPO market.
   Impressive prosperity right before the National Congress of CPC must help government
officials keep in office or even get promotion. They should have strong incentive to
manipulate policies to boost economy. IPO quota is clearly one of these policies. The
empirical evidence of this study shows that the government officials manipulate IPO policy
during political hot seasons, and there is an initial sharp decline in the number of IPOs over
9 quarters directly after the National Congress of CPC, followed by a relatively flat period in
the next 4 quarters and a gradual increase in successive 7 quarters where the number of IPOs
peaks in the quarter of CPC congress.
   We also find that the central government officials give priority to developed provinces in
terms of IPO volume. These provinces often have more political influence which is the
central government officials want to cater to. This finding is also consistent with
opportunistic model. But no evidence supports that central government officials cater to
investors in primary market.




REFERENCES

Alti, A., 2005. IPO market timing, Review of Financial Studies 18, 1105-1138.
Alti, A., 2006. How persistent is the impact of market timing on capital structure. Journal of Finance 61,
      1681-1710.
Benninga, S., M. Helmantel, and O., Sarig, 2005. The timing of initial public offerings. Journal of Financial
      Economics 75, 115-132.
Chan, K, John K. C. Wei and J. B. Wang, 2004. Under-pricing and long-term performance of IPOs in China.
      Journal of Corporate Finance 10, 409-430.
Chen, Z. H., J. J. Choi, and C. Jiang, 2007. Private benefits in IPOs: evidence from state-owned firms, working
      paper.
Chi, J and C, Padgett, 2005. Short-run underpricing and its characteristics in Chinese initial public offering
      markets. Review of International Business and Finance 19, 71-93.
Choe, H., R. Masulis, and S, Prowse, 1993. Common stock offerings across the business cycle. Journal of
      Empirical Finance 1, 1-29.
Fair, R. C., 1978. The effect of economic events on votes for president. Review of Economics and Statistics 60,
      159-173.
Graham, J. R. and C. R. Harvey, 2001. The theory and practice of corporate finance: evidence from the field.
      Journal of Financial Economics 60, 187-243.
Grier, K. B., 1987. Presidential elections and Federal Reserve policy: an empirical test. Southern Economic
      Journal, 54, 475-486.
Grier, K. B., 1989. On the existence of a political monetary cycle. American Journal of Political Science 33,
      376-389.
Helwege, J. and N. Liang, 2004. Initial public offerings in hot and cold markets. Journal of Financial and
      Quantitative Analysis 39, 541-569.
Ibbotson, R. G., and J.F. Jaffe, 1975. “Hot issue” markets. Journal of Finance 30, 1027-1042.
Ibbotson, R. G., J. L. Sindelar, and J. R. Ritter, 1988. Initial public offerings. Journal of Applied Corporate
      Finance 1, 37-35.
Ibbotson, R. G., J. L. Sindelar, and J. R. Ritter, 1994. The market’s problems with the pricing of initial public
      offerings. Journal of Applied Corporate Finance 7, 66-74.
Kalecki, M., 1943. Political aspects of full employment, reprinted in Bruno S. Frey, Political business cycles.
      Cheltenham, Edward Elgar Publishing..
Kim, W. and M. Weisbach, 2008, Motivations for public equity offers: an international perspective, Journal of

                                                        17
     Financial Economics, 87, 281-307.
Kramer, G. H., 1971. Short-term fluctuations in U.S. voting behavior, 1896-1964. American Political Science
     Review 65, 131-143.
La Porta, R., F.Lopez-de-Silanes, A. Shleifer and R. Vishny, 1997. Legal determinants of external finance.
     Journal of Finance 52, 1131-1150.
La Porta, R., F.Lopez-de-Silanes, A. Shleifer and R. Vishny, 1998. Law and finance. Journal of Political
     Economy 106, 1113-1155.
La Porta, R., F.Lopez-de-Silanes, A. Shleifer and R. Vishny, 2000. Investor protection and corporate governance.
     Journal of Financial Economics 58, 3-27.
Lerner, J., 1994. Venture capitalists and the decision to go public. Journal of financial Economics 35, 293-316.
Li, H. B. and L. A. Zhou, 2005. Political turnover and economic performance: incentive role of personnel control
     in China. Journal of Public Economics 89, 1743-1762.
Lindbeck, A., 1976. Stabilization policy in open economies with endogenous politicians. American Economic
     Review Papers and Proceedings 66, 1-19.
Loughran, T., and J. R. Ritter, 1995. The new issue puzzle. Journal of Finance 50, 23-51.
Loughran, T., J. R. Ritter, and K. Rydqvist, 1994. Initial public offerings: international insights. Pacific-Basin
     Finance Journal 2, 165-199.
Lowry, M., 2003. Why does IPO volume fluctuate so much? Journal of Financial Economics 67, 3-40.
Lowry, M. and G. W. Schwert, 2002. IPO market cycles: bubbles or sequential learning. Journal of Finance 57,
     1171-1200.
McRae, D. C., 1977. A political model of the business cycle. Journal of Political Economy 85, 239-263.
Maksimovic, V. and P. Pichler, 2001. Technological innovation and initial public offerings. Review of Financial
     Studies 14, 459-494.
Megginson, L. W. and L. H. Tian, 2007. Extreme Underpricing: Determinants of Chinese IPO Initial Returns.
     Unpublished working paper.
Nordhaus, W. D., 1975. The political business cycle. Review of Economic Studies 42, 169-190.
Pagano, M, F. Panetta, and L. Zingales, 1998. Why do companies go public? An empirical analysis. Journal of
     Finance 53, 27-64.
Rajan, R., and H., Servaes, 1997. Analyst following of initial public offerings. Journal of Finance 52, 507-529.
Ritter, J. R., 1984. The ”hot issue” market of 1980. Journal of Business 57, 215-240.
Ritter, J. R., and I. Welch, 2002. A review of IPO activity, pricing and allocation. Journal of Finance 57,
     1795-1828.
Tufte, E., 1975. Determinants of the outcomes of midterm congressional elections. American Political Science
     Review 69, 812-826.
Wan, J. Y. and A. Yuce, 2007. Listing regulations in China and their effect on the performance of IPOs and SOEs.
     Research in International Business and Finance 21, 366-378.




                                                        18

								
To top