Microsoft Word - 5822-20950-1-SM-writer2-new.doc


Asian Journal of Finance & Accounting  
ISSN 1946-052X 

2014, Vol. 6, No. 2 

www.macrothink.org/ajfa 64

Is Temporary Trading Halt Mechanism for New Shares 

on The First Day of Trading Conducive to Curbing 

Speculations? 

 

Kai Shi 

School of Economics, Northeast Normal University 

2555 Jingyue Street, Changchun 130117, P.R. China 

E-mail: shik142@nenu.edu.cn 

 

Li Nie 

Graduate School of Commerce and Management Administration 

Hitotsubashi University, 2-1 Naka, Kunitachi, Tokyo 186-8601, Japan 

E-mail: brilliant.so.kai@gmail.com 

 

Received: June 17, 2014    Accepted: July 30, 2014     Published: December 1, 2014 

doi:10.5296/ajfa.v6i2.5822   URL: http://dx.doi.org/10.5296/ajfa.v6i2.5822 

 

Abstract  

IPO underpricing as well as speculation on the first day of trading is a worldwide 
phenomenon. In order to curb IPO speculation risks, Shenzhen Stock Exchange makes an 
attempt to directly restrain price volatility and turnover rate through the Temporary Trading 
Halt Mechanism. This paper employs Rubin Causal Model and Genetic Matching method to 
evaluate and analyze real effects of the Mechanism. Empirical results show that, although it 
lowers the first-day turnover rate, the mechanism does push up first-day returns and first-day 
Price-to-Earning ratio. In fact, it keeps the closing price stay in a relatively high level, but is 
helpless for curbing risks. 

Keywords: Temporary Trading Halt Mechanism for IPOs on the First Day of Trading, 
Matching Method, Rubin Causal Inference 

 



Asian Journal of Finance & Accounting  
ISSN 1946-052X 

2014, Vol. 6, No. 2 

www.macrothink.org/ajfa 65

Introduction 

In China, speculative trading about newly listed shares is a typical drinking-game in which 
thousands of irrational investors may ignore stock fundamentals and follow suit to participate 
in the hype, and the finale is being hung up for a long time and suffering great losses. 
According to the statistics of Financial Innovation Laboratory of Shenzhen Stock Exchange, 
497 of 583 newly listed shares between the beginning of 2010 and the end of February 2012, 
especially for 33 of 34 newly listings, whose first-day returns exceed 100%, have fallen 
below the first-day closing price, with an average drop of 23.15% and the maximum decline 
even outpaces 64.4%. The loss percent of individual investors in whose account the amount is 
less than 100 thousand yuan reaches up to 60.75%. 

Since the launch of Small-and-Medium Enterprise Board, especially for the beginning of IPO 
System Reform, Shenzhen Stock Exchange (SZSE) has committed to explore the effective 
ways of refraining speculation. During the initiation of Growth Enterprise Market, SZSE 
made a first trial about the Temporary Trading Halt Mechanism for Newly-listed Shares 
during First-day Trading, and punished 9 short-term speculative accounts, who frequently 
partook in the hype, by restricting free transactions. In the wake of the effects of IPO System 
Reform gradually emerging, pricing in primary market becomes more and more reasonable; 
meanwhile, newly listed shares speculations in secondary market get to intensify once again. 
From the middle of February, 2012, the returns of 8 IPOs on the first-day of trading expands 
significantly, with an average rise of 62.65%, and over 60% shares’ first-day returns growth 
outpaces 50%, including 3 shares whose rises even exceed 80%. 

In order to take precautions against the risk of newly listed shares on the first-day of trading 
as well as cooperate with IPO System Reform, SZSE made a bold move which stressed direct 
executive interference on IPO shares’ first day performance and released “Notice on Further 
Optimizing the Temporary Trading Halt Mechanism for Newly-listed Shares during First-day 
Trading” (the Notice), highlighting the role of Trading Halt Mechanisms. Although SZSE has 
rolled out a series of new supporting measures for IPO reform, analysts still insist that these 
probably encourage speculations and bring individual investors more risks. 

Among copious literature relevant to IPO underpricing not many researches pay enough 
attention to IPOs first-day risk management. Ruud (1993) investigates the distribution of 
initial returns following IPOs and find a partially unobserved left tail effect. Ruud believes 
underwriter price support may account for the skewed distribution and challenges the 
presumption that IPO underpricing is deliberate. According to his logic, any IPO regulatory 
initiatives aimed at eliminate premium is redundant. Pettway and Kaneko (1996) explore the 
linkage between the change of IPO pricing regimes and new shares initial returns. They find 
that the price limits remove and auction mechanism introduction significantly reduce initial 
returns. Thus, risk management about initial returns may rely on market-oriented reforms. 
Krigman, Shaw and Womack (1999) show that IPO flipping is predictable and underwriters’ 
pricing errors are intentional. Thus, rational pricing is the very solution to restrain the risk of 
first-day flipping. Kao, Wu and Yang (2009) examine two sets of IPO regulatory initiatives, 
pricing regulations and penalty regulations and find pricing regulations may induce firms to 
manipulate pricing-period earnings thus negatively influence the post-IPO performance 



Asian Journal of Finance & Accounting  
ISSN 1946-052X 

2014, Vol. 6, No. 2 

www.macrothink.org/ajfa 66

whereas penalty regulations prevent firms from over-optimism about earnings forecast thus 
have positive impacts. Song, Tan and Yi (2014) evaluate the relative importance of IPO 
underpricing and overvaluation and find that overvaluation has more explanatory power on 
initial return. Its implication lies in what the regulatory commission should do is to reduce 
overvaluation but not to overly depend on controlling secondary market performance.   

In this paper, we focus on testing the real effects of the Notice which in fact sets a limitation 
for the price volatility of newly listed shares, and explore whether this measure lowers the 
risks or not. To the best of our knowledge, this is the first attempt to quantify the impacts of 
price control on the secondary market performance of IPOs during the first day of trading. 
Our findings not only provide guidance for deepening China’s IPO System Reform, but also 
provide a reference for other countries. The structure of this article is as follows: section 2 
briefly reviews the attempts of SZSE in coordination with IPO System Reform; section 3 
describes data and econometric methodology; section 4 explains empirical results, and 
section 5 makes a concluding remark. 

Attempts in Coordination with IPO System Reform in China 

On 25th May, 2012, three shares landed on Small-and-Medium Enterprise Board which 
commenced the implementation of new IPO rules. The new rules bring three changes: 
increasing the allotment for institutional investors in off-line offering; removing the 
requirement for three-month lock-up; allowing new shares acquired on the first trading day to 
freely float. In order to propel reform, new IPO rules are also absorbed into Management 
Measures for Securities Issuance and Underwriting amended in line with the China 
Securities Regulatory Commission Guidelines for Further Deepening IPO Reform (the 
Guidelines). In response to changes in new shares issuance management, SZSE releases three 
follow-up measures aimed at curbing first-day speculation, inducing rational pricing and 
maintaining orders during the first-day transaction of newly-listed shares. 

First, SZSE releases “Notice on further Optimizing the Temporary Trading Halt Mechanism 
for Newly-listed Shares during First-day Trading”, which highlights the role of the trading 
halt mechanisms. In order to curb newly listed shares speculation, SZSE released “First-day 
Temporary Trading Halt System for IPO Shares” on 8th March, 2012. Some positive results 
emerged after that. First-day closing prices of 45 IPOs listed on SZSE since then has risen 
only 21.82% on average, which is obviously lower than previous. In response to the public 
concerns over shortened trading time for hot IPO shares on their first trading day, SZSE 
reformulated the trading halt triggering criterion based on the percent of price volatility and 
turnover ratio. This adjustment is part of the consistent effort to explore and optimize the 
existing trading halt mechanism through sound appraisal and extensive public consultation. 
The Notice further refines the system. First, according to the measure, a new price volatility 
criterion of 20% above or below the opening price is set to be a trigger: (1) If price 
fluctuation during intraday trading reaches or exceeds 10% above or below the opening price 
for the first time, trading will be suspended for one hour and risk alert of possible speculation 
will be issued to investors. If price fluctuation during intraday trading reaches or exceeds 
20% above or below the opening price for the first time, trading will be suspended until 2:57 



Asian Journal of Finance & Accounting  
ISSN 1946-052X 

2014, Vol. 6, No. 2 

www.macrothink.org/ajfa 67

p.m., the time that closing auction begins; (2) If turnover rate reaches or exceeds 50% during 
intraday trading, trading will be suspended for one hour. 

Second, SZSE releases guidelines to enhance regulation over the trading of newly-listed 
shares. Guidelines for Monitoring Abnormal Trading Behaviors in Periods Immediately after 
Listing (draft) covers many new points: specifying explicit definitions of abnormal trading 
behaviors on first-day and subsequent trading; clarifying specific regulatory measures to 
enhance transparency of SZSE’s regulations and procedures; offering evaluation and 
corresponding monitoring standards for member firms’ performance in managing their 
clients’ behaviors in the transaction about newly-listed shares. 

Third, SZSE provides further guidance for investors rationally participating in IPOs and 
urges members to intensify the efforts of investor suitability management. By drawing on the 
experience of investor suitability management on the ChiNext market, SZSE emphasizes risk 
alerts and education and requires institution members to help raise their clients’ awareness of 
compliance during their participation in IPOs trading as well as further enhance 
self-regulatory mechanism. 

In general, two notable measures may significantly influence the first-day transaction. First, 
under the new scheme, transactions about newly-listed shares will be halted for one hour if 
the price volatility relative to the opening price by as much as 10% during its first trading day. 
Meanwhile, the status of trading halt until the last 3 minutes before closing will take effect if 
the price deviates over 20% from the opening price. Second, transactions will also be halted 
for one hour if the turnover ratio exceeds 50% on the first day. Obviously, the distinctive 
characteristic of temporal trading halt system is that it is a non-market intervention means. In 
the view of SZSE, it is necessary for an immature market even though not everyone believes 
so. However, numerous analysts insist that although being given high expectation the trading 
halt mechanism actually loosen the regulation about speculation due to the fact that 
speculators are able to manipulate trading with fewer funds under this mechanism. Thus, it is 
helpless for risk control. 

Data and Methodology 

Data 

We aim at quantifying the effect of temporary trading halt mechanism on the performance of 
first-day trading. Data about 51 newly-listed shares upon SZSE during 1st January, 2012 and 
11th May, 2012 (including 19 shares issued before the implementation of temporary trading 
halt mechanism and 32 ones issued after that) have been selected for empirical research. We 
focus on this period for two reasons: the primary one is to keep other impact factors invariant 
as far as possible meanwhile eliminating the influence of market trends, thus 1st January, 
2012 has been choose as the beginning; the other is that since the end of May 2012, IPO in 
Chinese domestic stock market has been paused temporarily waiting for further reform 
deepening. All data come from CSMAR database. 

In the process of policy effect assessment and comparative analysis, one of the most 
considerable issues is to figure out how the same share will perform during the first day of 



Asian Journal of Finance & Accounting  
ISSN 1946-052X 

2014, Vol. 6, No. 2 

www.macrothink.org/ajfa 68

trading without temporary trading halt mechanism? It involves the comparison between 
observational and potential outcomes. Except in the realm of science fiction, where parallel 
universes are sometimes imagined to be observable, it is impossible to measure causal effects 
at the individual level. In essence, causal inference is a missing value issue. We resort to 
Rubin Causal Inference Model and Matching Method.   

Rubin Causal Inference Model 

The Rubin causal model conceptualizes causal inference in terms of potential outcomes under 
treatment and control, only one of which is observed for each unit (Rubin 1990). A causal 
effect is defined as the difference between an observed outcome and its counterfactual. 

Let Yi1 denote the potential outcome for newly-listed share i if its first trading day is after the 
release of the Notice, and Yi0 denote the potential outcome for i if its opening transaction is 
before the release of the Notice. The treatment effect for i is defined by τi = Yi1 - Yi0. Causal 
inference is a missing data problem because Yi1 and Yi0 are never both observed. In principle, 
if assignment to treatment is randomized, causal inference is straightforward, while, in our 
observational setting, covariates are almost never balanced across treatment group, in which 
members are newly-listed after the release of the Notice and thus subject to the temporary 
trading halt mechanism, and control groups, in which shares are newly-listed before the 
release of the Notice and thus independent of its regulation, because they are not ordinarily 
drawn from the same population. Thus, a common quantity of interest is the average 
treatment effect for the treated (ATET):  

( ) ( ) ( )1 01 1 1i i i iT E Y T E Y Tτ = = = − =                            (1) 

where Ti is a treatment indicator who equals to 1 when i is newly-listed after the release of 
the Notice and 0 otherwise. 

Equation 1 cannot be directly estimated because Yi0 is not observable for the treated. Thus, 
we assume that selection into treatment depends on observable covariates X. Following 
Rosenbaum and Rubin (1983), conditional on X, treatment assignment is unconfounded ({Y0, 

Y1⊥T}|X) and there is overlap: 0<Pr(T=1|X)<1. Together, they constitute the property of 

strong ignorability of treatment assignment. For ATET, the unconfoundedness assumption can 
be weakened to mean independence: E(Yij|Ti,Xi)=E(Yij|Xi), and meanwhile the support of X for 
the treated be a subset of the support of X for control observations is an enough shortcutting 
for overlap assumption. By conditioning on observed covariates X, treatment and control 
groups are exchangeable. Then the ATET can be estimated as 

( ) ( ) ( ){ }1 , 1 , 1 1i i i i i i iT E E Y X T E Y X T Tτ = = = − = =                   (2) 

where the outer expectation is taken over the distribution of Xi|(Ti=1) which is the distribution 
of baseline variables in the treated group. 

When estimating causal effects using observational data, it is desirable to replicate a 



Asian Journal of Finance & Accounting  
ISSN 1946-052X 

2014, Vol. 6, No. 2 

www.macrothink.org/ajfa 69

randomized experiment as closely as possible by obtaining treated and control groups with 
similar covariate distributions. This goal can often be achieved by choosing well-matched 
samples of the original treated and control groups, thereby reducing bias due to the covariates. 
The most straightforward and nonparametric way to condition on X is to exactly match on the 
covariates, however it usually fails in finite samples if the dimensionality of X is large. Two 
other commonly used approaches are propensity score matching and multivariate matching 
based on Mahalanobis distance. A significant shortcoming of them is that they may (and in 
practice, frequently do) make balance worse across measured potential confounders (Sekhon, 
2011). Diamond and Sekhon (2013) propose a matching algorithm, genetic matching, that 
maximizes the balance of observed covariates between treated and control groups. It is a 
generalization of propensity score and Mahalanobis distance matching. The algorithm is 
nonparametric and does not depend on knowing or estimating the propensity score. 

Genetic Matching  

The idea underlying genetic matching method is that if Mahalanobis distance is not optimal 
for achieving balance in a given dataset, one should be able to search over the space of 
distance metrics and find something better. If one has a good propensity score model, it 
should be included as one of the covariates. Under this circumstance, both propensity score 
matching and Mahalanobis matching can be considered special cases of genetic matching. 

Genetic matching is an affinely invariant matching algorithm using the distance measure d(), 
in which all elements of W are zero except down the main diagonal. 

( ) ( ) ( ) ( )
1 2

1 2 1 2,i j i j i jd X X X X S WS X X− − ′′= − − 
 

                    (3) 

where W is a k×k positive definite weight matrix and S1/2 is the Cholesky decomposition of S 
which is the variance-covariance matrix of X. 

Setting the non-diagonal elements of W to zero aims at enhancing computational power, and 
meanwhile many loss criteria can be used to choose the free elements of W. 

Genetic matching method attempts to minimize a measure of the maximum observed 
discrepancy between the matched treated and control covariates at every iteration of 
optimization. Unlike traditional matching methods, genetic matching reliably reduces both 
the bias and the mean squared error of the estimated causal effect even when the propensity 
score is incorrectly specified and the covariates are not ellipsoidally distributed (as is almost 
always the case in applied work) as long as the selection on observables assumption holds. 
Evenly when the covariates are ellipsoidally distributed and the propensity score is correctly 
specified, in finite samples, estimates based on genetic matching have lower mean squared 
error than those based on the usual matching methods such as propensity score matching 
(Diamond and Sekhon, 2013). 

 

 



Asian Journal of Finance & Accounting  
ISSN 1946-052X 

2014, Vol. 6, No. 2 

www.macrothink.org/ajfa 70

Empirical Results 

The phenomenon of high first-day returns of IPO shares in China has drawn lots of attention 
for a long time. Dozens of news reports, specialist interviews has dedicated to this topic. High 
first-day return is not only a speculation phenomenon in an immature market, but also a 
signal which implies that there are severe problems in current new shares pricing system. The 
temporary trading halt mechanism is no other than a trial to curb speculative transaction.  

Matching 

The key concept in determining which covariates to include in the matching process is that of 
strong ignorability. To satisfy the assumption, it is important to include all variables known to 
be related to both treatment assignment and the outcome. Generally poor performance is 
found of methods that use a relatively small set of “predictors of convenience” (Shadish, 
Clark and Steiner, 2008). Including variables that are actually unassociated with the outcome 
can yield slight increases in variance. However, excluding a potentially important confounder 
can be very costly in terms of increased bias. Thus, it should be liberal to include variables 
that may be associated with treatment assignment and/or the outcomes. Meanwhile, one type 
of variable that should not be included in the matching process is any variable that may have 
been affected by the treatment of interest.  

In China, under the approval system, China Securities Regulatory Commission controls the 
scale and pace of new share issue through checking submitted shares’ price, quantity and 
time to market. New share issue price is always built on price-earning ratio. Reasonable 
price-earning ratio and issuance quantity are two key points of getting issuance approval. 
Considering these, we decide to take five variables: Expected Fully Diluted Price Earning 
Ratio, Fully Diluted Price Earning Ratio, Offering Qquantity, Issuance Quantity, Circulation 
Quantity, as covariates to do the genetic matching. The full output is included in Appendix. 
As shown in the Appendix, the smallest p value across all of the variables after matching is 
0.2356 compared with the pre-matching value of 0.0998. Apparently, genetic matching 
improve the balance across covariates. The balance is excellent for all variables. 

The Impact of First-day Temporary Trading Halt Mechanism    

Not that we have achieved excellent balance, we can examine our estimate of the treatment 
effect and its standard error. In this paper, we focus on three aspects about new shares 
performance: first-day market-adjusted returns, first-day turnover rate and first-day 
price-earning (PE) ratio. All resutls are presented in the Table.  

First-day Market-adjusted Returns. The estimate of the treatment effect for the treated 
about the first-day market-adjusted returns (ATETMAReturns) is 4.79% with a Abadie-Imbens 
standard error of 0.0209. The average treatment effect about first-day market-adjusted returns 
(ATEMAReturns) is 9.71% with a standard error of 0.0449. Both of them are significant at 5 
percent significant level. 

First-day Turnover Rate. The estimate of the treatment effect for the treated about the 
first-day turnover rate (ATETTORate) is -21.85% with a Abadie-Imbens standard error of 



Asian Journal of Finance & Accounting  
ISSN 1946-052X 

2014, Vol. 6, No. 2 

www.macrothink.org/ajfa 71

0.0906. The estimate of the treatment effect about first-day turnover rate (ATETORate) is 
-21.03% with a standard error of 0.0897. Also, both the ATETTORate and ATETORate are 
significant at 5 percent significant level. 

First-day Price-earning Ratio. The estimate of the treatment effect for the treated about the 
first-day PE ratio (ATETPERatio) is 10.04 with a Abadie-Imbens standard error of 2.6952. The 
estimate of the treatment effect about the first-day PE ratio (ATEPERatio) is 9.93 with a 
standard error of 2.5873. Both the ATETTORate and ATETORate are significantly different from 0 
at 1 percent significant level. 

 

Table 1. The real effects of temporary trading halt mechanism 

 Impact Direction ATET ATE 
First-day Market-adjusted Returns + 4.79%** 

(0.0209) 
9.71%** 
(0.0449) 

First-day Turnover Rate - -21.85%** 
(0.0906) 

-21.03%**

(0.0897) 
First-day Price-earning Ratio + 10.04*** 

(2.6952) 
9.93*** 

(2.5873) 

Note: Numbers in the brackets are Abadie-Imbens standard errors. ** and *** represent being 
different from 0 at 5% and 1% significant level, respectively. “+” means the temporary 
trading halt mechanism makes the dependent variables increase and “-” means decrease. On 
average, SZSE’s First-day Temporary Trading Halt Mechanism for New Shares pushes up 
individuals’ first-day market-adjusted returns and their first-day PE ratio whereas lower down 
first-day turnover rate.  

Empirical results show that the first-day temporary trading halt mechanism for IPO shares 
has positive impacts on first-day returns and first-day PE ratio as well as a negative impact on 
first-day turnover rate, which means the initiatives that SZSE made about curbing IPO 
speculation in fact do not push up the first-day price but also limit the transaction meanwhile. 

Concluding Remarks 

The direct purpose of first-day temporary trading halt mechanism for IPO shares is to curb 
the speculative transaction that unsuccessful subscribers have high inclinations to involve in. 
Moreover, its deeper motivation lies in coordinating with the improvement of IPO pricing 
mechanism as well as narrowing down the spread between primary and secondary market. It 
is, in essence, a transitional supporting measure. In fact, under current circumstances, the 
dilemma of stock market segmentation between primary and secondary market in China is 
still serious. IPO usually descends to the tunnel of profit transferring. Restrictions about 
trading performance in secondary market without breaking the ice of segmentation are unable 
to solve the problem of IPO underpricing and first-day speculation. 

 



Asian Journal of Finance & Accounting  
ISSN 1946-052X 

2014, Vol. 6, No. 2 

www.macrothink.org/ajfa 72

Acknowledgement 

This research was financially supported by the MoE Research Funds of Philosophy and 
Social Science (11YJC790181), the Jilin Social Science Planning Funds (2014B40) and the 
Fundamental Research Funds for the Central Universities (14QNJJ039). 

References 

Abadie, A., & Imbens, G. (2006). Large Sample Properties of Matching Estimators for 
Average Treatment Effects. Econometrica, 74, 235-267. 
http://dx.doi.org/10.1111/j.1468-0262.2006.00655.x 

Diamond, A., & Sekhon, J.S. (2013). Genetic Matching for Estimating Causal Effects: A 
General Multivariate Matching Method for Achieving Balance in Observational Studies, 
Review of Economics and Statistics, 95, 932-945. http://dx.doi.org/10.1162/REST_a_00318 

Jennifer L. Kao, Donghui Wu, & Zhifeng Yang (2009). Regulations, Earnings Management, 
and Post-IPO Performance: the Chinese Evidence. Journal of Banking and Finance, 33, 
63-76. http://dx.doi.org/10.1016/j.jbankfin.2007.03.016 

Judith S. Ruud (1993). Underwriter Price Support and the IPO Underpricing Puzzle. Journal 
of Financial Economics, 34, 135-151. http://dx.doi.org/10.1016/0304-405X(93)90015-4 

Laurie Krigman, Wayne H. Shaw, & Kent L. Womack (1999). The Persistence of IPO 
Mispricing and the Predictive Power of Flipping. Journal of Finance, LIV, 3, 1015-1043. 

Richard H. Pettway, & Takashi Kaneko. (1996). The Effects of Removing Price Limits and 
Introducing Auctions upon Short-term IPO Returns: the Case of Japanese IPOs. Pacific-Basin 
Finance Journal, 4, 241-258. http://dx.doi.org/10.1016/0927-538X(96)00013-3 

Rosenbaum, P.R., & Rubin, D.B. (1985). The Central Role of the Propensity Score in 
Observational Studies for Causal Effects. Biometrika, 70, 41-45. 
http://dx.doi.org/10.1016/j.cjar.2013.12.001 

Rubin, D.B. (1990). Comment: Neyman (1923) and Causal Inference in Experiments and 
Observational Studies. Statistical Science, Vol. 5, 472-480. 

Sekhon, J.S. (2011). Multivariate and Propensity Score Matching Software with Automated 
Balance Optimization: The Matching Package for R. Journal of Statistical Software, 42(7). 

Shunlin Song, Jinsong Tan, Yang Yi (2004). IPO Initial Returns in China: Underpricing or 
Overvaluation? China Journal of Accounting Research, 7, 31-49. 
http://dx.doi.org/10.1016/j.cjar.2013.12.001 

William R. Shadish, M. H. Clark and Peter M. Steiner. Can Nonrandomized Experiments 
Yield Accurate Answers? A randomized Experiment Comparing Random and Nonrandom 
Assignments. Journal of the American Statistical Association, 103(484), 1334-1343. 
http://dx.doi.org/10.1198/016214508000000733 

 

 

 

 



Asian Journal of Finance & Accounting  
ISSN 1946-052X 

2014, Vol. 6, No. 2 

www.macrothink.org/ajfa 73

Appendix 

The full output of matching is as follows: 
***** (V1) Expected Fully Diluted Price Earning Ratio ***** 

Before Matching           After Matching 
mean treatment........     23.2380                   23.2380  
mean control..........    26.5210                   23.2330  
std mean diff.........    -58.0450                   0.0920  

mean raw eQQ diff.....     3.5811                    0.9252  
med raw eQQ diff.....    2.1100                    0.7200  
max raw eQQ diff.....      23.0600                    3.1800  

mean eCDF diff........    0.1028                    0.0443  
med eCDF diff........      0.0905                    0.0400  

max eCDF diff........       0.2611                    0.1600  
var ratio (Tr/Co).....       0.2794                    0.9535  
T-test p-value.......         0.2356                    0.9836  
KS Bootstrap p-value.       0.3230                    0.8510  
KS Naive p-value......      0.4537                    0.9062  
KS Statistic..........          0.2611                    0.1600  
 
***** (V2) Fully Diluted Price Earning Ratio ***** 

Before Matching           After Matching 
mean treatment........     30.8900                  30.8900  
mean control..........     34.7260                  30.3300  

std mean diff.........       -49.8980                  7.2907  
mean raw eQQ diff.....    4.3337                    1.4132  
med raw eQQ diff.....     2.3800                    0.9600  

max raw eQQ diff.....      30.7400                   5.6800  
mean eCDF diff........     0.0901                    0.0457  
med eCDF diff........     0.0779                    0.0400  

max eCDF diff........     0.2463                    0.1600  
var ratio (Tr/Co).....     0.2843                    0.2843  
T-test p-value.......      0.1652                    0.3027  
KS Bootstrap p-value.     0.3900                    0.8500  
KS Naive p-value.....     0.5291                    0.9062  
KS Statistic.........       0.2463                    0.1600  
 
***** (V3) Offering Qquantity ***** 

Before Matching           After Matching 
mean treatment........     2808.5                    2808.5  
mean control..........      2500.9                    2582.7  
std mean diff.........      20.816                    15.282  

mean raw eQQ diff.....    378.16                    249.16  
med raw eQQ diff.....      200                       160  
max raw eQQ diff.....      3400                      1500  



Asian Journal of Finance & Accounting  
ISSN 1946-052X 

2014, Vol. 6, No. 2 

www.macrothink.org/ajfa 74

mean eCDF diff........     0.0524                    0.0515  
med eCDF diff........     0.0463                    0.0400  
max eCDF diff........     0.1811                    0.1600  

var ratio (Tr/Co).....     2.2794                    1.3164  
T-test p-value........      0.0998                    0.5919  
KS Bootstrap p-value..     .7030                    0.8270  
KS Naive p-value......     0.8710                    0.9062  
KS Statistic..........       0.1811                    0.1600  
 
***** (V4) Issuance Quantity ***** 

Before Matching           After Matching 
mean treatment........     2808.5                    2808.5  
mean control..........     2500.9                    2582.7   
std mean diff.........      20.816                    15.282  

mean raw eQQ diff.....   378.16                    249.16  
med raw eQQ diff.....    200                       160  
max raw eQQ diff.....    3400                      1500  

mean eCDF diff........    0.0524                    0.0515  
med eCDF diff........     0.0400                    0.0463  
max eCDF diff........     0.1600                    0.1811  

var ratio (Tr/Co).....     2.2794                    1.3164  
T-test p-value........      0.0998                    0.5919  
KS Bootstrap p-value..    0.7030                    0.8270  
KS Naive p-value......    0.8710                    0.9062  
KS Statistic..........       0.1811                    0.1600  
 
***** (V5) Circulation Quantity ***** 
       Before Matching           After Matching 
mean treatment........     2249.6                    2249.6  
mean control..........     2068.4                    2005.2  
std mean diff.........      15.304                    20.648  

mean raw eQQ diff.....    302.98                    198.11  
med raw eQQ diff.....    164                       128  
max raw eQQ diff.....    2700                      1200  

mean eCDF diff........    0.0519                    0.0507  
med eCDF diff........     0.0484                    0.0400  
max eCDF diff........     0.1811                    0.1600  
var ratio (Tr/Co).....      2.2602                    1.3189  

T-test p-value........      0.1010                    0.5911  
KS Bootstrap p-value..    0.7090                    0.8280  
KS Naive p-value......    0.8710                    0.9062  
KS Statistic..........      0.1811                    0.1600  

Before Matching Minimum p.value:  0.0998 

After Matching Minimum p.value:   0.2356