Test for Cumulative Average Abnormal Retums of Private Equities

In this study, the event date of private placement was set to“the date being public disclosed in Market Observation Post (MOPS) after the boards meetings approve private placements." The event study module derived from Taiwan

Chiao Da λ fanagement Review Vol. 30 No. 1, 2010 27

Table 3 Statistics summary

Variables Obs. Minimum Maximum Mean Standard Deviation

Cumulative abnormal return

ofprivate placement 199 -179.76% 207.30% 10.05% ^{的 46%}

announcement (CAR)

Discount rate (DC) 132 -67.62% 128.31% -13.82% 24.01%

Jnternal should raise ratio

132 O 100% 40.98% 41.73%

(JND)

Ratio ofprivate placement

132 O 220.27% 29.34% 28.06%

over paid-in-capital 伊C刃

Table 4

Analysis of frequency statistics

Variable Frequency Percentage Effective Cumulative Percentage Percentage

Do not change ofmanagerial right 118 89.39 89.39 89.39

Change ofmanagerial right 14 10.61 10.61 100.0

f刁nancialsoundness companies 91 68.94 68.94 68.94

Financial distress companies 41 31.06 31.06 100.0

Non-strategic partners on private

91 68.94 68.94 68.94

placement

Strategic partners on private placement 41 31.06 31.06 100.0

28 EjJects on Shareholders' 悅。lthand Premium ρiscount)

of Private Placement Announcemen

Table 5

Statistics of directors or supervisors seat investors obtained in the private placement

Number of directors or supervisors seat investors

O 2 3 4 5 6 7 8 9 Sum

obtained in the private placement (NCR)

Fnquency 84 7 14 9 6 3 3 2 132

Economic Joumal Database was adopted. The cumulative average abnonnal retums during the announcement period were calculated according to the market model with the sample. In this study, the “estimation period" is from 200 days before announcement to 60 days before 祉， with data of at least 30 days. And the

“

event period" is defined as the period from 30 days before announcement to 30

days after it. The estimation and test for the cumulative average abnonnal retums of listed and OTC companies which had launched private placements were perfonned.

Table 6

Results of Average Abnormal Return (AAR) and Cumulative Average Abnormal Return (CAAR)

Event day ^AAR t-value p-value ^CAAR t-value p-value

-30 -0.23 -1.01 0.31 -0.23 -0.95 0.34

-29 0.21 0.92 0.36 -0.05 _{。07 。.94}

-28 0.31 1.35 0.18 0.27

。“

^0.51

-27 0.87 3.83 0:∞ 1.22 2.34 ^0.02抖

-26 0.42 1.84 0.07* 1.47 2.85 o.∞*料

-25 0.13 0.57 0.57 1.53 2.81 0.01***

-24 012 0.96 0.34 1.73 2.93 ^o.∞*料

Chiao Da Management Review Vol. 30 No. 1,2010 29

-23 0.46 2.ül 0.05** 2.31 3.37 。∞料*

-22 0.36 1.57 0.12 2.62 3.66 。∞*料

-21 0.17 0.73 0.47 2.68 3.69 ^{。 ω*料}

-20 0.15 0.67 0.50 2.74 3.70 o.∞料*

-19 023 0.99 0.32 2.~ 3.81 ^o.∞料*

-18 0.62 2.72 O.ül *** 3.74 4.36 。-∞*料

-17 0.43 1.88 0.06* 4.25 4.67 ^o.∞料*

-16 0.37 1.61 0.11 4.67 ^{4到 o.∞料*}

-15 '{).78 -3.40 。ω料* 3.65 3.95 ^o.∞*料

-14 0.24 1.05 0.30 3.82 4.07 o.∞料*

-13 0.48 2.10 ^0.04料 4.47 4.42 o.∞***

-12 0.41 1.82 0.08* 4.89 4.70 ^o.∞*料

-11 0.59 2.58 0.01** 5.39 5 日 o.∞料*

-10 0.50 2.205 0.03** 5.79 5.46 ^。∞艸*

-9 0.13 0.59 0.56 5.87 5.45 。助***

-8 '{).52 -2.27 0.02** 5.49 4.89 ^o.∞料*

-7 '{).43 -1.87 0.06* 5.05 4.43 ^0.00紳*

~ '{).16 '{).71 0.48料 4.88 4.20 。∞料*

-5 0.53 2.32 0.02** 5.54 4.55 o.ω*艸

4 '{).16 '{).69 0.49 5.34 4.34 。∞*艸

-3 '{).06 。.25 0.80* 5.11 4.22 。∞***

-2 0.57 2.52 0.01** 5.68 4.58 。 ω*料

0.19 0.85 0.40 5.89 4.66 ^o.∞科*

O 0.42 1.82 0.07* 6.56 4.89 。 ω***

。.42 1.85 0.07* 6.98 5.12 o.ω***

2 0.32 1.39 0.17 7.51 527 ^o.ω*料

3 '{).29 -1.27 0.21 6.83 4.98 ^o.∞料*

4 0.49 2.14 ^0.03輛 7.40 5.26 ^o.∞*料

5 。.09 。 38 。.70 7.32 5.13 ^o.∞*料 6 0.15 0.64 0.52 7.24 5.16 。∞***

30 ξfJects on Shareholders' Wealth and Premium (Discount) 01 Private Placement Announcement

7 。 04 -0.20 0.85 6.29 5.師 。∞紳*

8 0.17 0.73 0.47 6.到 5.10 O.∞*料

9 0.38

l

“

^{0.10* 7.29 529 O.∞料*}

10 0.18 0.78 0.44 7.07 5.34 O.∞*料

11 0.37 1.63 0.10 7.64 5

.5

² _。-∞***

12 0.01 0.03 0.97 7

.5

^{9 5 船} 。ω***

13 0.16 0.68 0.50 8.13 5

.5

⁰ 。-∞*料

14 -0.41 -1.79 0.08* 7.63 5.18 O.∞*料

15 0.28 1.21 0.23 7.72 5.30 _。ω***

16 0.10 0.45 0.65 8.20 5.30 O.∞*料

17 0.09 0.38 0.71 8.24 5.30 。∞***

18 0.49 2.13 0.03** 8.10 5

.5

³ 。∞艸*

19 0.27 1.18 0.24 8.77 5.64 。∞料*

20 0.37 1.62 0.11 9.22 5.80 。∞***

21 -0.36 -1.59 0.11 9.34 5

.5

³ 。 ω***

22 0.05 023 0.82* 9.13 5

.5

¹ 。 ω***

23 0.31 1.37 0.17 9.76 5.63 O.∞料*

24 0.44 1.93 0.06* 10.37 5.83 。∞***

25 0.61

2

“

^{0.01料* 10.94 6.12 。∞***}

26 -0.31 -1.37 0.17 10.39 5.89 O.∞*料

27 。 25 -1.11 027 9.72 5.70 O.∞*料

28 -0.34 -1.48 0.14 9.15 5.47 。ω***

29 。 07 。.30 0.77 _{9 ∞} 5.39 O.∞*料

30 。.55 2.43 0.02** 9.38 5.64 O.∞林*

Note: *** significantly at 1 %; ** significantly at 5%; * significantIy at 10% (test of two-tail).

According to Table 6 and figure 1, the AAR of the period from two days before the announcement date to two days after that for the 199 sample companies are all positive. The ARRs on the day of announcement and the next day shows significant at 10%. The ARR on the 4th day after the announcement date even

Chiao Da A1anagement Review Vol. 30 No. 1,2010 31

reaches the significant level of 5%. It is obvious that the private placernent inforrnation does have effects on the day of announcernent and the next day According to Table 6 and figure 2, the CAAR of the period frorn 30 days before the event date to 30 days after that (CAAR[-30,+30]) is 9.38%. This phenornenon is cornpatible with the results frorn literatures. It irnplies 出at private placernent event rnessages still have inforrnation content. And the reaction period is longer In other words, announcernents of private placernents have positive inf1uence on shareholders' wealth

Figure 1

Average abnormal return (AAR) of event period.

。9!IlO

』閉目

Figure 2

Cumulative average abnormal return (CAAR) of event period.

l肌u刪|……^'

B叩01XXl f卡卜.一….一…-…….一….一…-一…-一….一ι-一…-一….一…-…，ι ，…….…一…-一-_..一….一.一...-……-…-‘」↓，.ι-……-一…-一-斗↓-一…-…--r-…_. -~. --_..._...,

叫i J -K?于; ( :(;(;(1 : : ( ! : ;

401XXl

200CJJ

.11 22 .14 -6 10 19 話

In Table 7, using event day as benchmark, by cornparing the curnulative average abnorrnal retum from the event period to the day before announcernent

32 EjJects ⁰¹¹ Shareholders' Wealth a^l1d Premium ρωCOUl1t) ofP.仟rrvaνP羽G仿 Pl，仿ac臼emeωn叫 11110U附ω仰nentt

(CAAR[-30,-1]) with 由at from announcement day to 30 days after that (CAAR[0,+30]), it is found that the cumulative average abnormal re仙m from the event period to the day before announcement (CAAR[-30,-1]) is 5.89%, while the cumulative average abnormal retum from announcement day to 30 days after that (CAAR[0,+30]) is 3.66%, which is lower than the 5.89% of CAAR[-30,-1], although still positive. It shows that CAAR is higher before the announcement day than after that. And the cumulative average abnormal retum since 27 days before announcement to 30 days after that reaches the significant level of 5%

This means since 27 days before submitting the proposal of private placement to the board' s meeting, the stock price of the company has already started to show a positive reaction. Because before announcement, only insiders or investors being contacted can know about the private placement while general investors can't, the significant cumulative average abnormal retum represents the information asymmetry does exist in private placement events. And those who know about it can obtain higher returns than those who know about it afterwards.

Table 7

Cumulative Average Abnormal Return (CAAR) Pre and Post Event Day

CAAR

CAAR[-30,-l]

5.89%

CAAR[O,+30]

3.66%

CAAR[-30,-30]

9.38%

In summarize, although the Financial Supervisory Commission has already started to strengthen the management in private placement cases since October, 2005, because the time asking for information disclosure is usually posterior, for those who can know about private placement before announcement such as board of directors, there are still chances to participate before announcement and end up with abnormal retums, while ordinary investors can only participate and obtain abnormal returns after announcement, their chances and opportunities are influenced by information asymmetry. It is obvious that the current regulation of private placement still need to be improved, in order to lower the differences in retums due to the timing when assessing information, and further to reach the faimess of information disclosed and market transactions.

Chiao Da A1anagement Review Vol. 30 No. 1, 2010 33

5.3. Multiple Regression of Private Placement Announcement on Cumulative Abnormal Return

The following discusses possible causes of abnormal retums on the day of announcement are explored with multiple regression analysis，的出e indication for investors to decide whether to invest or not, and also as the reference of holding period

As shown in Table 3, the lowest cumulative abnormal retum (CAR[ -30,+ 30]) of the 199 sample companies included into event analysis is

-179.76% (that is, Ta-I Technology Co., Ltd.), while the highest is 207.30%

(Prince Asset Management Corp. Limited, ex Kings Information, whose management rights were transferred to Uni President in 2007.) and the average CAR is 10.05%. 77

Table 8

Pearson correlation analysis of factors on cumulative abnormal return

IND^a NCR^b PCT CCR^d FC. SA^f

JND

入'CR -0.11 (0.11)

PCT -0.17 0.25

(.03**) (.00**)

CCR -0.20 0.37 0^.3

(.015**) ^{(.00料*) (.00* 艸)}

FC -0.09 0.13 0.28 0.18

(0.18) (0.07) (.00* 料) (.03**)

SA -0.43 _{。 20} 0.09 0.47 -0.13

(.00***) (.01 **) (0.15) (.00***) (0.09) Note: a.IND: Ratio ofinsiders.

b. NCR: supervisors seats investors obtained in the private placement.

c. PCT: numbers of private placement over total paid-in-capital d. CCR: Change of managerial rights

e. FC: FinanciaUy distress

f. SA: Strategic partners.

34 EjJects on Shareholders' Wealth and Premium (Discount) 01 Private Placement Announcement

companies' prospec仙s operation cannot successfully completed after the private placement announcement, therefore, the sample size being used in analyzing the causes of abnormal retum after private placement announcement is 122.

It can be proved by the Pearson correlation analysis in Table 8 由atthere is no relationship of high degree between each two variables. Multiple regression analysis was performed for formula (22) and derived the regression coefficients and test statistics listed under model (1) in Table 9. In addition, in order to avoid positive and negative offset effects of CAR being neutralized, the sample was divided into two samples, with CAR>O and CAR豆o respectively. Regression analyses were performed with these two samples, and the regression coefficients and test statistics are listed under model (2) and model (3) in Table 8. According to the information 仕om model (1) the D-W test statistics value for regression is 2.01, which means that the variation of the residual is stable, and there is no auto-correlation. Furthermore, the VIF values of all the independent variables less than 10, which means these independent variables are not collineari可

According to the regression coefficients in Table 9: The change of CAR is positive during the private placement period, when more directors or supervisor seats are obtained by private placement investors, when the management rights are transferred, and when there had been financial crisis for the company. This is consistent with the direction expected under hypotheses. On the other hand, the change of CAR is negative, which is the opposite of the expected direction, when more insiders of the company issuing private placement shares participate, when the percentage of private placement shares in total stock shares after private placement is higher, and when strategic pa此ners have participated þrivate placements. This result is different from the findings ofHertzel and Smith (1993) that "the ratio of private placement shares to total shares after that private placement" has positive influence on abnormal retums of private placement shares, and is also different from Folta and Janney (2004) findings that by delivering information of corporate value via private placement, there' d be higher abnormal retums for technology companies. The cause of the strategic alliance part might be that the sample companies had lowered the prices to seduce invest

Chiao Da Afanagement Rel'iew J，句1.30 No. 1,2010 35

companies with strategic alliance participating the investments. This leads to the fact that the companies with strategic alliance joining investments cannot deliver the information of growth-of-value to solve the problem of information asymmetry. The stock price cannot be push up with the growth-of-value information. A1so the regression analysis result shows that the variable "whether managerial rights have been transferred" under model (1) and (2) had reached the significant level at 5%, which is consistent with the above-mentioned “Ownership Structure Hypothesis" and Shleifer and Vishny (1986) argument: A company can attract huge extemal investors by private placement, through outside large shareholders, to supervise and manage the company, and further facilitate to promote the enhancement of corporate value

According to the regression analysis and the statistics in Table 10: the companies which had their managerial rights transferred because of the private placements have the average cumulative abnormal retum of 40.43% during the event period. As for the companies whose managerial rights had not been transferred, although their CAR values are positive, the average of CAR is 8.09%

The difference between these two averages is as high as 32.35%. This phenomenon implies that after the update of the regulation by the Financial Supervisory Commission because of the Xepex case, under the circumstance which the private placement investors have obtained the managerial rights (including backdoor listing, mergers and acquisitions, ... ), the stock prices can still rising sha中ly

The multiple regression analysis performed for formula (2月 shows that:

"whether managerial rights have been transferred" under model (1) and (2) are with significant positive abnormal retums. This proves the above-mentioned

“Ownership Structure Hypothesis". This phenomenon implies that if small private placement investors know the information that a new management team will step in because of the private placement, they can obtain excess abnormal retums by using this information to make decisions about whether to invest or not after private placement announcement

36 丘。告c臼 onShareholders' Wealth and Premium ρiscount)

of Private Placement Announcement

Table 9

Multiple Regression ofPrivate Placement Announcement on Cumulative Abnormal Return

b. NCR means Number of directors and supervisors seats investors obtained in the private

pla，臼ment.

c. PCT means the ratio of number of private placement shares over total number of paid-in-capita1.

d. CCR means the change of rnanagerial right; FC means 仕lefinancially distress; SA means strategIc partners.

Chiao Da Management Review Vol. 30 No. 1,2010 37

Table 10

Statistics summary of managerial right change and cumulative abnormal return

Cumulative abnonnal return Obs. Mean Standard _Minimum Maximum

of private placement deviation

Change ofmanagerial right 14 40.43% 56.18% -41.07% 207.30%

Do not change ofmanagerial 108 8.09% 34.41% -66.48% 120.60%

ri叉ht

CMR>O ⁷⁷ 32.25% 31.91% 0.16% 207.30%

CMR ::;;0 45 -23.89% 17.69 -66.48% -1.09%

5.4. Discussion on the Causes of Premium (Discount)

From Table 3, it can be found that the highest discount of the sample companies from private placement is 67.62%, and the highest premium is 128.31 %. Averagely, private placement share price is the reference price with

13.82% discount. The possible causes are discussed below with multiple

regression analysis.

From the Pearson correlation analysis resuIt in Table 11, it is found that there is no high degree correlation between any two of the variables. And the regression analysis result for model (1) in Table 12 shows that: the VIF values of all the independent variables are less than 10. The D-W value is 1.90, which means the independent variables are not collinearity. The variation of the residual is sable and there is no sign of auto-correlation. The resuIt of multiple regression analysis for formula (23) shows that: from the regression coefficients and test statistics of model (1) in Table 12, it is found that the discount is higher, when the ratio of insiders is higher for a private placement, when the number of directors or supervisor seats are obtained by investors of the private placement, when the ratio of number of private placement shares to number of total shares after that private placement is higher, when the number of directors or supervisors seats investors obtained in the private placement is higher, and strategic alliance partners are drawn. Additional旬， the managerial rights have been transferred for a private placement, which issued at a premium. The result is in line with the expected

38

direction ofhypothesis

EjJects on Shareholders' Wealth and Premium ρiscount)

01 Private Placement Announcement

On the other hand, when there is a financial crisis for a company, the private placement share price is higher than the reference price. This phenomenon is not consistent with the direction of hypothesis. Lin (2006) pointed out about this phenomenon that sometimes there is capital decrease during the period before or after private placement. Therefore extreme values may occur. Or the offset direction by positive and negative of premium (discount) may affect the result of regression analysis

In this study, in order to avoid offset direction by positive and negative of premium (discount) may affect the result of regression analysis, the sample was divided into two samples, discount (DC<O) and premium (DC 這0) respectively MuItiple regression analysis was performed for both samples (model (2) and model (3)) to further explore the possible factors which have influences on premium (discount). From the regression coefficient and test statistics in Table 11 and the correlation coefficient in Table 12, it is found that: in model (2) with the discount sample (DC

<

0),“the purpose of private placement is to attract strategic partners" and "number of directors or supervisors seats investors obtained in the private placement" are the important factors which facilitate the premium (discount) of private placement (both reach the significant level of 10%), and the influences effects are negative (that 凹， discount), which is consistent with the expectation of hypothesis

Wu (2004) pointed out that: the reason why the discount is higher when the private placement investors are managers, should have something to do with managers' self-dealing. When managers have only a small amount of original shares, there is a strong incentive to make self-dealing. With the method of buying in a lower price and selling in a higher price, the stockholders' weaIth is transferred from existing shareholders to managers themselves. However, it is found in this study ofthe Taiwan's private placement through the results ofmodel (1) and (3) th剖 therelationship between

“

ratio ofinsiders' share amount in

Chiao Da Management Review Vol. 30 No. 1,2010

Table 11

Pearson correlation analysis of factors affecting on premium (discount) Variables

b. NCR: Number of directors and supervisors seats investors obtained ^{in 血e}private

pla臼ment

d. PC^T: Ratio of numbers of private placement over total paid-i必capital

e. CCR: Change of rnanagerial right.

f. FC: Financially distress g. S^A: Strategic partners

Private placement to the total share amount in that private placement" and premium (discount) is not significant. The self-dealing behavior of managers cannot be proved exist significant1y

In summary, in the discount samples,“purpose of private placement is to attract strategic pa此ners" and "number of directors or supervisor seats investors obtained in the private placement" are the important significant factors which have influences on discount of private placement. The phenomenon implies that the private placement share issuing company can attract strategic alliance partners with lower prices for capital increase. In addition, the higher the "number of directors or supervisors seats investors obtained in the p討 vate placement" 肉， the higher the discount is. However, because the relationship between “ratio of insiders' share amount in private placement to the total share amount in that private placement" and premium (discount) is not significant, which means the phenomenon of enterprises using capital increase with discount to attract

"outsiders" to invest andjoining the management teams ofthe enterprises

40 EjJects ⁰¹¹ Shareholders' Wealth al1d Premium (Discoul1t)

of Private Placemel1t Al1nouncement

Table 12

MuItiple regression of private placement announcement on premium (discount)

Variabl ModelI ^{闊別leJ11 (DC}

<

⁰⁾ 加IodeJ m (DC益的

es _自 p-value VIF a p-value VIF 。 ^{p-value VIF}

C:ÒIl敘711t -11. 03 -212 ∞ 19.42 20

民V -.19 97 129 0.05 12 127 -14.27 .34 2.10

Nar

^{-1.3 23 1.19 -1.33 .仰* 1.13 -3.21 21 1.59}

PCr -.05 .54 1.19 -.∞ 97 115 -.04 .84 1.35

cat

^2.55 .75 1.41 -.15 98 1.40 -9.24 55 1.51

FC 4.35 .39 1.20 -.49 ^部 1.11 21.33 15 2.33

S.

4

-3.3 .56 1..56 “6.88 .07* 1.57 1.88 .89 2.42

F- 0.59 2.45 1.14

Pva1ue 0.74 0.03** 0.37

D-W 1.90 1.50 1.96

Adj-lf ^{-.0型 0.08 0.03}

Note: a. 的D:Ratio of insiders.

b. NCR: Number of directors and supeIVisors seats investors obtained in the private placement.

d. PCT: Ratio of numbers of private placement over total paid-in-capital e. CCR: Change of managerial right.

f. FC: Financially distress g. SA: Strategic partners

在文檔中 Effects on Shareholders' Wealth and Premium (Discount) of Private Placement Announcement (頁 26-40)