Trade Liberalization and the Economy:

Stock Market Evidence from Singapore

January 11, 2009

Rasyad A. Parinduri, the corresponding author, is aliated with the Nottingham Univer-
sity Business School, the University of Nottingham, Malaysia Campus; Shandre M. Thangavelu
is aliated with Department of Economics, National University of Singapore. Mailing ad-
dress: Nottingham University Business School, Jalan Broga, 43500 Semenyih, Selangor Darul
Ehsan, Malaysia; phone: +6(03) 8924 8287, fax: +6(03) 8924 8019; e-mail addresses:
[email protected] (Rasyad A. Parinduri), and [email protected] (Shandre
M. Thangavelu).
1
1 Introduction
This paper examines the eect of free trade agreements (FTAs) on the value of
rms using event study analysis. The central questions are the followings: Do
FTAs increase the value of rms as perceived by investors? Which industries win or
loose? Can we say rms prot from FTAs, and, hence, economies gain from trade
liberalization?
We consider the United StatesSingapore Free Trade Agreement (USSFTA). An
event study analysis of Singapore and this FTA is interesting for several reasons.
First, Singapore has been adopting liberal trade policies, which means the gain from
complete trade liberalization through FTAs may not be large. Second, the USSFTA
requires Singapore to set zero taris for all U.S. exports immediately. The U.S., on
the other hand, is required to eliminate 92% of current taris on Singapore exports,
while the rest within eight years. Singapore rms would have to compete with
cheaper imports immediately, while some of their competitors in the U.S. will enjoy
protection for a few more years. Third, Singapore is perhaps the most aggressive
country that has been pursuing FTAs, which makes the negotiation moves of the
Government of Singapore more predictable.
1
These features would perhaps make the identication of the eect of FTA in
event study analysis more dicult. If we could show that FTAs increase the value
of rms in a free market economy whose government eagerly and predictably pursues
FTAs, we would perhaps provide some evidence that FTAs do increase the value of
rms.
We estimate the eect of FTA-related news using a modied market model, from
which we could use the residuals as the measures of the abnormal returns (1s) or
1
To date, Singapore has concluded 13 FTAs and is currently negotiating nine others. See
http://www.fta.gov.sg/.
2
cumulative abnormal returns (C1s) associated with the FTA. First, we consider
one industry at a time, and we constrain all rms in the industry to have the same
responses to the FTA news. Second, we relax this assumption by parameterizing
the 1s or C1s with measures of comparative advantage and economies of scale.
2
We nd that many of the event coecients that we introduce in the model are
insignicant statistically. Only one event that investors seem to think surprising,
i.e., the removal of the last obstacles to the USSFTA deal in January 2003. On
average, the value of rms in two or three out of six industries that we consider
increases 2-5% because of this event. Other industries may prot from the FTA
too; their event coecients are not statistically signicant, however.
We do some robustness checks. First, we use Straits Times Index as the market
portfolio rather than the FTSE Strait Times All Share Index to exclude equities that
are infrequently traded from the market index. Second, we examine rms in several
sectors rather than industries to make rms in each sample more similar. Third, we
include sector stock index to allow each rm to have dierent sector specic risks.
Fourth, we drop all events except the one that is signicant statistically in the basic
specication. Overall, our results are quite robust.
To the best of our knowledge, there are three papers that analyze the eect
of FTA on the value of rms using event study analysis, i.e., Thompson (1993,
1994) and Rodriguez (2003). The focus of Thompson (1994) and Rodriguez (2003),
however, is on whether the market responses to news about, respectively, Canada
United States Free Trade Agreement and North American Free Trade Agreement,
are consistent with comparative advantage and economies of scale as predicted
by the HeckscherOhlin model and the New Trade Theory. While we control for
2
Controlling for these two characteristics in the market model is proposed by Thompson (1994),
in which she develops a theoretical model of the value of a rms capital as a function of, among
others, the relative factor shares, the relative factor prices, and the relative plant scales.
3
Singapores comparative advantage and the rms plant size in some specications,
we focus more on whether the FTA increases the value of rms.
3
This paper is more closely related to Thompson (1993). She considers six
CanadaUnited States Free Trade Agreement related events, and she nds that
only one event in which the 1s are both statistically signicant and consistent
with the hypothesis about the eect of FTA.
The paper proceeds as follows. Section 2 presents the methodology. Section
3 describes the events and data. Sections 4 and 5 discuss the empirical results.
Section 6 concludes.
2 Methodology
We examine investors expectation about the eect of USSFTA on the value of
Singapore rms using event study analysis. Under the assumption that market is
ecient, equities will be priced so that they yield "normal"-risk adjusted expected
rate of return. When investors nd out that the Government of Singapore concludes
an FTA with the U.S., and if this agreement increases the expected stream of future
prots and, hence, the value of Singapore rms, they will bid up the share prices
immediately. We could then use the response of share prices to the FTA to estimate
the change in the value of rms that results from the FTA announcement.
We model the expected return of securities using the following market model:
:
it
= c
i
+ ,
i
:
mt
+ c
it
. t = 1. .... 1. (1)
where :
it
is the return to security i at time t; :
mt
is the return to the market portfolio
at time t; ,
i
is the systematic risk of security i, and c
it
is a stochastic term, which
is assumed to be homoskedastic across equities and serially uncorrelated. We use
3
Thompson (1994) and Rodriguez (2003) do not present the event coecients in their papers.
4
the residuals of the market model as a measure of the abnormal returns (1s) of a
security.
4
Our primary interest is the sum of the abnormal returns of security i over an
event window : (the cumulative abnormal returns (C1)), which is the measure
of the eect of event : on the market value of security i. If we nd that C1
is
is positive and statistically signicant, we would then reject the null hypothesis
that event : does not change investors expectation about rm is stream of future
prots.
Because we have multiple events, to simplify the estimation of the 1s, we
introduce several event-day indicator variables to Equation (1) as follows
:
it
= c
i
+ ,
i
:
mt
+
X
s

its
1
its
+ c
it
. t = 1. .... 1. (2)
where 1
its
is an indicator variable equals one for the tth day in event window : and
zero otherwise. The estimated coecient of
its
would then be identical to the 1s
of security i at time t associated with the event : estimated using Equation (1).
We make two adjustments to Equation (2). First, we allow the systematic risk of
securities after Event 1 to be dierent from the risk during the estimation window.
Second, we control for leveraging across rms.
5
The magnitude of the change in
share prices that results from an FTA announcement depends on the debt-to-equity
ratio of rms. Because we will assume that rms have similar responses to an
announcement, we divide the event indicator variables by the share of equity in the
total rm value. Our working model would therefore be as follows:
:
it
= c
i
+ ,
i
:
mt
+ c
0
i
+ ,
0
i
:
0
mt
+
X
1
(1 1c/t)

its
1
its
+ c
it
(3)
4
For surveys on event study analysis, see MacKinley (1997) and Binder (1998). For its appli-
cation in the economics of regulation, see Binder (1985) and Lamdin (2001).
5
This adjustment to the market model is used by Rose (1985).
5
where c
0
i
and ,
0
i
are the cs and ,s of the return to the market portfolio after
Event 1, and (1 1c/t
i
) is the share of equity in rm i. We estimate this system
of equations using Zellner (1962)s Seemingly Unrelated Regression.
6
An unconstrained estimation of this model has low power against the null hy-
pothesis. Therefore, we impose some structures on this system of equations. First,
we group rms according to their industries or sectors, and constrain that all rms
within a group have the same response to an event. If the rms in a group are quite
similar so that the events aect the rms similarly, the constraint is justiable.
Second, we relax the equality constraint by parameterizing
its
, the share price
responses, as functions of rms or industries characteristics. Following Thompson
(1994)s theoretical model of the value of a rms capital, we use two characteristics
to parameterize
its
, i.e., loan to capital ratio as a measure of comparative advan-
tage, and plant scales as a measure of economies of scale. The
its
in Equation (3)
is then modelled as follows:

its
= o
1s
+ o
2s
1C1
i
+ o
3s
oC11. (4)
where 1C1 is the relative labor intensity and oC11 is the relative plant scale.
3 Events and Data
3.1 Events
The Governments of Singapore and U.S. initiated the negotiation on November
16, 2000, and the FTA was signed three years later on May 6, 2003. The long
negotiation of the FTA, and the both governments strong support for it, make it
dicult to identify events that investors nd unexpected. Nevertheless, based on
news reported in the Straits Times, Business Times, and Channel News Asia, we
6
In cases in which we use three-day- or ve-day event windows, we use a modied version of
Equation (3) proposed by Salinger (1992) to account for both the intertemporal and contempora-
neous correlation of estimated residuals.
6
consider ten events that may change investors perception on the probability that
the FTA will be concluded and implemented. Table 1 presents these events.
[INSERT TABLE 1 HERE]
The rst event is the joint statement on a U.S.Singapore Free Trade Agreement.
Since then, both parties had held several rounds of negotiation, each lasted a few
week long. Both parties were typically optimistic about the negotiation before each
round, and signicant progress was typically reported at the conclusion. We do not
think that these rounds of negotiation and their results are surprising, however.
The second event is the rejection of the Trade Promotion Authority (TPA) bill,
which would facilitate the passage of FTAs in Congress, by the U.S. Senate. Most
observers did not think that this rejection would harm the FTA so that this event
may not be surprising either. We still include it as one of the events, however. We
also include the clearance of the TPA bill by the U.S. House of Representatives a
few months later as the third event, and the announcement of broad agreement over
the FTA in November 2002 as the fourth event.
The fth event is a major milestone in the negotiation of the FTA. On January
15, 2003, it was announced that Singapore and U.S. overcame the last obstacle to an
FTA deal, i.e., the free transfer of capital. Later that month, Bush administration
notied Congress of plan to sign FTA with Singapore, which we include as the sixth
event.
The last four events are as follows: The FTA text was made online, the news
that Goh Chok Tong to meet Bush to sign the FTA, the signing of the FTA itself,
and the approval of the FTA by the U.S. House of Representatives. We are not
sure whether investors nd these events surprising. However, we include them all
and let the data tell us which events are unexpected, if there are any.
7
3.2 Data
We get the daily stock market data from the Bloomberg. The period of analysis
starts one year before the rst event, and ends right after the last event, i.e., from
September 1, 1999 to July 26, 2003. We include rms whose stocks were listed in
the Singapore Exchange during the entire period of analysis. In addition, these
rms must be located in Singapore.
We use the dividend-adjusted stock prices and the FTSE Strait Times All Share
Index to calculate the stock return, :
t
, and the return to the market portfolio, :
mt
,
respectively. Because the share of the market value of debt in total rm value, 1c/t,
is not available, we use the book value of debt as a proportion of total assets as a
proxy.
The relative labor intensity, 1C1, which is dened as the ratio of labor income
to capital income, is not available. As a proxy, we use the ratio between the number
of employees and the value of the rms capital in 2000. We calculate the relative
plant scale, oC11, as the dierence between the average sales of rms in a sector
in Singapore and those in the U.S.
We include six out of the ten industries as dened by Bloomberg. They are
Basic Materials-, Consumer Goods-, Financial, Technology, Industrial- and Health
Care industries. Because rms included in the "Industrial" are large and diverse,
we exclude some sectors such as Industrial Engineering and Industrial Services.
7
Out of the 753 rms listed in the Singapore Exchange, 456 equities satisfy our
requirements. Many securities are infrequently traded however. The total number
of rms that we use in all regression is 144 if we include all ten events. In some
specications we include only one event (i.e., Event 5); in this case the total number
7
Bloomberg does not provide the SIC codes of rms listed in the Singapore Exchange. Indus-
tries that we exclude are Consumer Services, Oil & Gas, Telecommunications, and Utilities.
8
of rms in the samples is 210.
4 Results
4.1 Constrained Results
We estimate the constrained version of Equation (3) for six industries using ve-
day-, three-day-, and one-day event windows. We nd that, in all cases, almost all
estimates of the event coecients are insignicant statistically. To reduce the noise
that may arise in longer event-window periods, at the risk of putting some events
outside of event windows, we decide to use one-day event window. Table 2 presents
the constrained results.
[INSERT TABLE 2 HERE]
The estimates are typically small and statistically insignicant. We nd only
one event coecient that is signicant statistically at 5% level, i.e., the estimate of
Event 5 for Health Care industry, which includes Health Care Equipment & Services
and Pharmaceuticals & Biotechnology sectors. Investors think that Event 5the
removal of the last obstacle to an FTA dealincreases the expected future prots
of rms in this industry by 2%.
The only other estimate that is statistically signicant at 10% level or lower is
the estimate of Event 5 for Basic Materials industry. The estimate is much smaller,
however: Overcoming the last hurdle increases rms values in this industry by 0.3%.
4.2 Cross-Firm Heterogeneity
Perhaps the characteristics of companies in each of the six industries are quite dif-
ferent so that the FTA does not aect the expected future prots of rms similarly.
If this is the case, estimating the average responses of rms using the constrained
version of Equation (3) will lead to ambiguous results. Therefore, to allow each
9
rm to respond dierently to the FTA-related announcements, we estimate Equa-
tion (3) in which the
its
is parameterized with two rms characteristics, i.e., 1C1
and oC11. Table 3 presents the results.
8
[INSERT TABLE 3 HERE]
Like the constrained results, almost all event coecients are insignicant statis-
tically. Only the estimates of Event 5 that are signicant statistically at 5% level,
i.e., the estimates for Basic Materials and Health Care industries.
9
According to
these estimates, the removal of the last obstacle to an FTA deal increases the val-
ues of Basic Materials companies by 11.4% and those of Health Care companies by
4.2%. The only other estimates that are statistically signicant at 10% level are
the estimate of Event 5 for Technology Hardware & Equipment industry and that
of Event 8Prime Minister Goh Chok Tong to meet Bushfor Consumer Goods
industry.
4.3 Further Analyses
We do some robustness checks. First, the FTSE Strait Times All Share Index,
which we use as the market portfolio, may include securities that are infrequently
traded. To make sure that the market model provides unbiased systematic risks of
securities and 1s, we use the Straits Times Index, an index based on the stocks
of 30 representative companies, as the market portfolio. These two indices are
highly correlated, however; and we get estimates that are quite similar to the basic
results.
10
8
The estimates of 1C1 and oC11 are not provided to save space. They are available from
authors upon request.
9
Basic Materials industry includes Chemicals-, Forestry & Paper-, Industrial Metals-, and
Mining sectors.
10
To save space, we do not present all results of robustness checks. They are available upon
request from the authors, however.
10
Second, the companies in each of the six industries that we consider in our basic
specications may be too diverse, and this may make the constraint that all rms
have the same response to an event less justiable. To rule out this possibility, we
estimate Equation (3) for twelve sectors, two from each industry. Table 4 presents
the results.
[INSERT TABLE 4 HERE]
Almost all estimates for all ten events are insignicant statistically. Only two
estimates are signicant statistically at 5% level of signicance or less, i.e., the Event
5 estimates for Chemicals- and Technology Hardware & Equipment sectors. Both
estimates are positive, which indicate that overcoming the last hurdle to an FTA
deal increases the expected future prots of rms in the Chemicals- and Technology
Hardware & Equipment sectors by 0.6% and 2.6%, respectively. Two estimates of
Event 5 are negative; only one of them is signicant statistically at 10% level,
however, i.e., the estimate for Banking sector. The market anticipates 1.4% lower
prots of banks as a result of the agreement on the free transfer of capital.
Overall, these estimates are in line with the basic results: The statistically
signicant results for Basic Materials- and Health Care industries seems to be driven
by companies in some of their sectors, in particular Chemicals- and Technology
Hardware & Equipment sectors, respectively.
Third, we do not account for the sector specic risks in our basic specication.
To allow each rm in a sector to have dierent sector specic risks, we introduce a
sector stock index into Equation (3) as follows.
:
it
= c
i
+ ,
i
:
mt
+
i
:
int
+ ... + c
it
. t = 1. .... 1. (5)
where :
int
is the sector : stock index of rm i at time t. This modication does not
change our basic results much, however.
11
Fourth, perhaps we include too many events that investors nd unsurprising,
and, hence, do not aect the price of securities. Even though this inclusion does
not cost us unbiasedness, it may make our estimates less precise. To investigate
this possibility, we drop all events except Event 5, which is the only event whose
coecients are signicant statistically at 5% level. Table 5 presents the results for
the constrained model and cross-rm heterogeneity in Panel 1 and 2, respectively,
using one-day- and three-day event windows.
[INSERT TABLE 5 HERE]
Overall, the basic results are quite robust. They are sensitive to the choice of
event windows, however. Estimation of constrained model using three-day event
window provides estimate of the C1 of Basic Materials industry that is statis-
tically insignicant, though still positive. Allowing cross-rm heterogeneity using
three-day event window makes the estimate of the C1 of Basic Material industry
signicant statistically at 10% level only.
We nd that most of the coecient of 1C1 and oC11 are insignicant sta-
tistically. These results may suggest that Singapore and U.S. economies are not
that dierent, or, perhaps more likely, we do not have sucient variation in the
data to reject the null hypothesis that labor to capital ratio and plant scales are
important determinants of the eect of FTA. If they are signicant statistically,
1C1 is negative while oC11 is positive, which mean the more labor intensive
Singapore rms are, the smaller the eect of FTA is; the larger Singapore rms
compared to their competitors in the U.S. are, the larger the eect of FTA is.
12
5 Discussion
We do not expect that the market would strongly respond to all ten FTA-related
announcements. Some of the events, such as the joint statement, the meeting of Goh
Chok Tong and Bush, and the approval of U.S. House of Representatives, might be
expected long before they were in the news. We introduce them into the model to
let the data decides which announcements that investors nd surprising and which
ones that are not. Besides, Thompson (1993, 1994) and Rodriguez (2003), three
papers that are related to this paper, also nd many statistically insignicant event
coecients.
11
Nevertheless, we can think of several reasons why most events do not seem to
aect the value of rms. First, the negotiation of the FTA is perhaps too predictable
so that, unlike, for example, that of NAFTA, investors do not nd the events
surprising.
Second, some stocks in the Singapore Exchange are infrequently traded, and this
makes the estimation of the Seemingly Unrelated Regression suers fromcollinearity
problem. Many equities are then dropped from the sampleout of the 456 equities
that are available for trade during the entire period of analysis, only 144 of them
that are included in all regression. This collinearity problem, and the exclusion of
many rms from the regressions, may have compromised our results.
Third, maybe we group too diverse companies together so that the equality
constraint across rms is not justiable. We could have examined the eect of the
FTA for some subsectors of Singapore economy rather than industries or sectors,
but, given the collinearity problem mentioned above, the number of rms in each
subsector would fall dramatically. Even when we do sector by sector analysis such
11
Thompson (1994) and Rodriguez (2003)s model, which is slightly dierent from, and less
exible than Equation (4), is as follows:
its
= 0
1s
+`
s
(0
2
1C1
i
+0
3
oC11).
13
as that in Table 4, only a few rms remain in the sample.
Fourth, unlike Singapore that is required to set zero taris for all U.S. exports
immediately, the U.S. is required to eliminate 92% of current taris on Singapore
exports while the rest within eight years. This may make the event coecients
smaller, and make it more dicult to reject the null hypothesis.
Nevertheless, we still nd that two or three out of six industries prot from
the FTA.
12
Two other industries may benet too; their event coecients are not
statistically signicant, however. Only one industry that is likely to loose, if at all;
that is the Financial industries.
6 Concluding Remarks
The negotiation of the USSFTA is perhaps too predictable. Nevertheless, we nd
that the market respond positively to the USSFTA, in particular to the removal of
the last obstacle to an FTA deal in January 2003. According the market evidence,
on average, rms in the Basic Materials- and Health Care industries are the clear
winners: Their values increase by 2-5%. Our results are silent on rms in the
Consumer Goods, Industrials, Financial and Technology industries, however. Some
rms in the latter industries may gain while some other loose so that the eects
cancel out and we cannot reject the null hypothesis.
Presumably, the average eect of the FTA on the values of rms in the Basic
Materials- and Health Care industries is larger. Event 5 may be partially expected,
which makes the estimate of the eect smaller. Perhaps we can say that the conclu-
12
These results also seem to be in line with some of the sectors that the Government of Singapore
considers to prot from the USSFTA. The sectors that are part of the two industries in our sample
are electronics, chemicals, instrumentation equipment, and mineral products. We do not include
petrochemicals in our sample; and we do not have sucient evidence to conclude that processed
foods sector, which is part of Consumer Goods industry in our sample, would benet from the
USSFTA. See Info Kit: Information Paper on the US-Singapore Free Trade Agreement (USSFTA),
16 May 2003, which is available at http://www.fta.gov.sg/ussfta/info_kit_ussfta.pdf.
14
sion of the USSFTA was partially expected long before the joint statement made
by Goh Chok Tong and Bill Clinton. Moreover, we have not mentioned the eect
of the FTA on consumers, which are likely to be the big winners of the FTA.
The predictability of the FTA negotiation may contribute to the statistical in-
signicance of our results. It would be interesting to examine other Singapore FTAs
that are more controversial than the USSFTA. Looking into Singapore FTAs with
developing countries rather than developed countries like the U.S. would be inter-
esting too. The gain from these FTAs may be larger so that the identication of
the eect of FTA on the value of rms would be easier. These will the subject of
future research.
References
[1] Binder, J.J. (1985), "Measuring the Eects of Regulation with Stock Price
Data," Rand Journal of Economics, 16(2), pp. 167-183
[2] Binder, J.J. (1998), "The Event Study Methodology since 1969," Review of
Quantitative Finance and Accounting, 11, pp. 111-137.
[3] Lamdin, D.J. (2001), "Implementing and Interpreting Event Studies of Regu-
latory Changes," Journal of Economics and Business, 53, pp. 171-183.
[4] MacKinlay, A.C. (1997), "Event Studies in Economics and Finance," Journal
of Economic Literature, 35, pp. 13-39.
[5] Rodriguez, P. (2003), "Investor Expectations and the North American Free
Trade Agreement, "Review of International Economics, 11(1), pp. 206-218.
[6] Rose, N.L. (1985), "The Incidence of Regulatory Rents in the Motor Carrier
Industry," Rand Journal of Economics 16(3), pp. 299-317.
15
[7] Salinger, M. (1992), "Standard Errors in Event Studies," Journal of Financial
and Quantitative Analysis, 27(1), pp. 39-53.
[8] Thompson, A.J. (1993), "The Anticipated Sectoral Adjustment to the Canada
- United States Free Trade Agreement: An Event Study Analysis," Canadian
Journal of Economics, 26, pp. 253-271.
[9] Thompson, A.J. (1994), "Trade Liberalization, Comparative Advantage, and
Scale Economies: Stock Market Evidence from Canada," Journal of Interna-
tional Economics, 37, pp. 1-27.
[10] Zellner, A. (1962), "An Ecient Method of Estimating Seemingly Unrelated
Regressions and Tests for Aggregation Bias," Journal of the American Statis-
tical Association, 57, pp. 348-368.
16
Events Dates Descriptions
1 16/11/2000 President Clinton and Prime Minister Goh Chok Tong made a joint
statement on a United States-Singapore Free Trade Agreement.
2 17/05/2002 The U.S. Senate denied the Trade Promotion Authority (TPA) bill,
which would facilitate the passage of FTAs in Congress.
3 30/07/2002 The U.S. House of Representative cleared the TPA Bill; Senate was
likely to pass the bill later this week.
4 19/11/2002 Singapore and U.S. reached broad agreement over the FTA.
5 15/01/2003 Singapore and U.S. overcome the last obstacle to an FTA deal, i.e.,
the free transfer of capital.
6 31/01/2003 Bush administration notified Congress of plan to sign FTA with
Singapore.
7 08/03/2003 U.S.-Singapore FTA text was made available online.
8 03/04/2003 Prime Minister Goh Chok Tong to meet Bush on May 6.
9 06/05/2003 President Bush and Prime Minister Goh Chok Tong signed U.S.-
Singapore Free Trade Agreement.
10 26/07/2003 The US-Singapore Free Trade Agreement was approved in the U.S.
House of Representatives.
Table 1: Event Descriptions
17
Dependent Variable: r
t
Basic
Materials
Consumer
Goods
Financials Technology Health Care Industrial
(1) (2) (3) (4) (5) (6)
1 Made joint statement 0.000 0.001 0.001 0.004 -0.010 -0.003
(0.002) (0.004) (0.005) (0.010) (0.009) (0.004)
2 TPA was denied 0.001 0.007 -0.001 0.009 0.003 0.001
(0.002) (0.004) (0.005) (0.010) (0.009) (0.004)
3 TPA was approved 0.000 -0.003 0.000 0.004 -0.009 0.002
(0.002) (0.004) (0.005) (0.010) (0.009) (0.004)
4 Broad agreement -0.001 -0.002 0.002 0.001 -0.003 0.002
(0.002) (0.004) (0.005) (0.010) (0.009) (0.004)
5 Overcame last hurdle 0.003 0.004 -0.004 0.011 0.020 0.004
(0.002)+ (0.004) (0.005) (0.010) (0.009)* (0.004)
6 Congress was notified 0.000 0.001 0.000 -0.004 0.005 0.002
(0.002) (0.004) (0.005) (0.010) (0.009) (0.004)
7 FTA text was made online 0.000 -0.001 0.001 -0.003 -0.010 0.000
(0.002) (0.004) (0.005) (0.010) (0.009) (0.004)
8 Goh and Bush to meet 0.000 0.000 -0.003 0.003 -0.002 0.002
(0.002) (0.004) (0.005) (0.010) (0.009) (0.004)
9 FTA was signed 0.000 0.000 0.000 0.007 -0.006 0.002
(0.002) (0.004) (0.005) (0.010) (0.009) (0.004)
10 Approved by House 0.000 0.001 -0.006 -0.004 0.003 0.002
(0.002) (0.004) (0.005) (0.010) (0.009) (0.004)
Number of daily returns 7848 30411 19620 3924 15696 66708
Number of firms 8 31 20 4 16 68
Note: Standard errors are in parentheses. + significant at 10%; * significant at 5%; ** significant at 1%
Table 2: Constrained Results
18
Dependent Variable: r
t
Basic
Materials
Consumer
Goods
Financials Technology Health Care Industrial
(1) (2) (3) (4) (5) (6)
1 Made joint statement 0.026 0.005 0.004 -0.003 -0.007 -0.003
(0.035) (0.008) (0.009) (0.013) (0.017) (0.005)
2 TPA was denied 0.025 0.001 -0.002 0.013 0.011 0.000
(0.035) (0.008) (0.009) (0.013) (0.017) (0.005)
3 TPA was approved -0.026 0.010 -0.003 0.017 -0.006 0.004
(0.035) (0.008) (0.009) (0.013) (0.017) (0.005)
4 Broad agreement -0.057 -0.006 -0.008 0.001 -0.006 0.002
(0.035) (0.008) (0.009) (0.013) (0.017) (0.005)
5 Overcame last hurdle 0.114 0.007 -0.011 0.022 0.042 0.004
(0.035)** (0.008) (0.009) (0.013)+ (0.017)* (0.005)
6 Congress was notified -0.018 -0.011 -0.005 0.010 0.020 0.001
(0.035) (0.008) (0.009) (0.013) (0.017) (0.005)
7 FTA text was made online -0.005 0.008 0.008 0.001 0.010 -0.002
(0.035) (0.008) (0.009) (0.013) (0.017) (0.005)
8 Goh and Bush to meet -0.008 -0.016 0.000 0.013 0.023 0.006
(0.035) (0.008)+ (0.009) (0.013) (0.017) (0.005)
9 FTA was signed -0.013 -0.009 -0.002 0.017 -0.004 -0.002
(0.035) (0.008) (0.009) (0.013) (0.017) (0.005)
10 Approved by House 0.021 0.000 -0.002 -0.008 0.011 0.000
(0.035) (0.008) (0.009) (0.013) (0.017) (0.005)
Number of daily returns 6867 30411 17658 3924 15696 66708
Number of firms 7 31 18 4 16 68
Note: Standard errors are in parentheses. + significant at 10%; * significant at 5%; ** significant at 1%
Table 3: Cross-Firm Heterogeneity
19
D
e
p
e
n
d
e
n
t

V
a
r
i
a
b
l
e
:

r
t
C
h
e
m
i
c
a
l
s
I
n
d
u
s
t
r
i
a
l
M
e
t
a
l
F
o
o
d
P
r
o
d
u
c
t
s
H
o
u
s
e
h
o
l
d
G
o
o
d
s
B
a
n
k
s
G
e
n
e
r
a
l
F
i
n
a
n
c
i
a
l
E
&
E
E
q
u
i
p
m
e
n
t
I
n
d
u
s
t
r
i
a
l
T
r
a
n
s
p
o
r
t
.
P
h
a
r
m
a
c
e
u
-
t
i
c
a
l
s
H
e
a
l
t
h
c
a
r
e
P
r
o
d
u
c
t
s
S
o
f
t
w
a
r
e

H
a
r
d
w
a
r
e
(
1
)
(
2
)
(
3
)
(
4
)
(
5
)
(
6
)
(
7
)
(
8
)
(
9
)
(
1
0
)
(
1
1
)
(
1
2
)
1
M
a
d
e

j
o
i
n
t

s
t
a
t
e
m
e
n
t
0
.
0
0
0
-
0
.
0
1
0
-
0
.
0
0
1
0
.
0
0
3
0
.
0
0
1
0
.
0
0
1
-
0
.
0
0
1
0
.
0
0
2
-
0
.
0
0
4
0
.
0
1
0
-
0
.
0
1
1
-
0
.
0
1
0
(
0
.
0
0
3
)
(
0
.
0
2
9
)
(
0
.
0
0
5
)
(
0
.
0
0
7
)
(
0
.
0
0
8
)
(
0
.
0
0
6
)
(
0
.
0
0
6
)
(
0
.
0
0
9
)
(
0
.
0
1
4
)
(
0
.
0
1
3
)
(
0
.
0
1
4
)
(
0
.
0
1
0
)
2
T
P
A

w
a
s

d
e
n
i
e
d
0
.
0
0
0
0
.
0
3
1
0
.
0
0
8
0
.
0
1
0
-
0
.
0
0
5
0
.
0
0
0
-
0
.
0
0
2
0
.
0
0
3
0
.
0
1
0
0
.
0
0
8
-
0
.
0
0
1
0
.
0
0
7
(
0
.
0
0
3
)
(
0
.
0
2
9
)
(
0
.
0
0
5
)
(
0
.
0
0
7
)
(
0
.
0
0
8
)
(
0
.
0
0
6
)
(
0
.
0
0
6
)
(
0
.
0
0
9
)
(
0
.
0
1
4
)
(
0
.
0
1
3
)
(
0
.
0
1
4
)
(
0
.
0
1
0
)
3
T
P
A

w
a
s

a
p
p
r
o
v
e
d
0
.
0
0
0
-
0
.
0
0
2
-
0
.
0
0
4
0
.
0
0
1
-
0
.
0
0
3
0
.
0
0
3
-
0
.
0
0
4
0
.
0
0
8
0
.
0
2
0
-
0
.
0
0
8
0
.
0
0
1
-
0
.
0
1
1
(
0
.
0
0
3
)
(
0
.
0
2
9
)
(
0
.
0
0
5
)
(
0
.
0
0
7
)
(
0
.
0
0
8
)
(
0
.
0
0
6
)
(
0
.
0
0
6
)
(
0
.
0
0
9
)
(
0
.
0
1
4
)
(
0
.
0
1
3
)
(
0
.
0
1
4
)
(
0
.
0
1
0
)
4
B
r
o
a
d

a
g
r
e
e
m
e
n
t
-
0
.
0
0
2
-
0
.
0
0
1
0
.
0
0
0
-
0
.
0
0
7
-
0
.
0
0
5
0
.
0
0
4
-
0
.
0
0
3
0
.
0
0
3
0
.
0
0
0
0
.
0
0
2
-
0
.
0
0
1
-
0
.
0
0
2
(
0
.
0
0
3
)
(
0
.
0
2
9
)
(
0
.
0
0
5
)
(
0
.
0
0
7
)
(
0
.
0
0
8
)
(
0
.
0
0
6
)
(
0
.
0
0
6
)
(
0
.
0
0
9
)
(
0
.
0
1
4
)
(
0
.
0
1
3
)
(
0
.
0
1
4
)
(
0
.
0
1
0
)
5
O
v
e
r
c
a
m
e

l
a
s
t

h
u
r
d
l
e
0
.
0
0
6
-
0
.
0
0
1
0
.
0
0
3
0
.
0
0
8
-
0
.
0
1
4
0
.
0
0
1
0
.
0
0
5
0
.
0
1
1
0
.
0
2
1
0
.
0
0
4
0
.
0
0
6
0
.
0
2
6
(
0
.
0
0
3
)
*
(
0
.
0
2
9
)
(
0
.
0
0
5
)
(
0
.
0
0
7
)
(
0
.
0
0
8
)
+
(
0
.
0
0
6
)
(
0
.
0
0
6
)
(
0
.
0
0
9
)
(
0
.
0
1
4
)
(
0
.
0
1
3
)
(
0
.
0
1
4
)
(
0
.
0
1
0
)
*
6
C
o
n
g
r
e
s
s

w
a
s

n
o
t
i
f
i
e
d
-
0
.
0
0
1
0
.
0
2
6
0
.
0
0
3
0
.
0
0
0
-
0
.
0
0
2
0
.
0
0
2
0
.
0
0
1
-
0
.
0
0
2
0
.
0
0
5
-
0
.
0
1
0
-
0
.
0
0
3
0
.
0
1
3
(
0
.
0
0
3
)
(
0
.
0
2
9
)
(
0
.
0
0
5
)
(
0
.
0
0
7
)
(
0
.
0
0
8
)
(
0
.
0
0
6
)
(
0
.
0
0
6
)
(
0
.
0
0
9
)
(
0
.
0
1
4
)
(
0
.
0
1
3
)
(
0
.
0
1
4
)
(
0
.
0
1
0
)
7
F
T
A

t
e
x
t

w
a
s

m
a
d
e

o
n
l
i
n
e
0
.
0
0
0
-
0
.
0
0
1
-
0
.
0
0
3
0
.
0
0
6
0
.
0
0
9
-
0
.
0
0
4
-
0
.
0
1
0
-
0
.
0
0
8
0
.
0
0
2
-
0
.
0
0
8
-
0
.
0
2
6
-
0
.
0
0
2
(
0
.
0
0
3
)
(
0
.
0
2
9
)
(
0
.
0
0
5
)
(
0
.
0
0
7
)
(
0
.
0
0
8
)
(
0
.
0
0
6
)
(
0
.
0
0
6
)
(
0
.
0
0
9
)
(
0
.
0
1
4
)
(
0
.
0
1
3
)
(
0
.
0
1
4
)
+
(
0
.
0
1
0
)
8
G
o
h

a
n
d

B
u
s
h

t
o

m
e
e
t
0
.
0
0
0
-
0
.
0
0
1
0
.
0
0
3
-
0
.
0
0
9
0
.
0
0
1
-
0
.
0
0
4
-
0
.
0
1
2
-
0
.
0
0
4
0
.
0
1
4
-
0
.
0
0
6
-
0
.
0
1
4
0
.
0
0
4
(
0
.
0
0
3
)
(
0
.
0
2
9
)
(
0
.
0
0
5
)
(
0
.
0
0
7
)
(
0
.
0
0
8
)
(
0
.
0
0
6
)
(
0
.
0
0
6
)
+
(
0
.
0
0
9
)
(
0
.
0
1
4
)
(
0
.
0
1
3
)
(
0
.
0
1
4
)
(
0
.
0
1
0
)
9
F
T
A

w
a
s

s
i
g
n
e
d
0
.
0
0
0
-
0
.
0
0
1
0
.
0
0
4
-
0
.
0
0
7
-
0
.
0
0
5
0
.
0
0
3
-
0
.
0
0
3
0
.
0
1
6
0
.
0
2
1
-
0
.
0
0
4
-
0
.
0
0
9
-
0
.
0
0
7
(
0
.
0
0
3
)
(
0
.
0
2
9
)
(
0
.
0
0
5
)
(
0
.
0
0
7
)
(
0
.
0
0
8
)
(
0
.
0
0
6
)
(
0
.
0
0
6
)
(
0
.
0
0
9
)
+
(
0
.
0
1
4
)
(
0
.
0
1
3
)
(
0
.
0
1
4
)
(
0
.
0
1
0
)
1
0
A
p
p
r
o
v
e
d

b
y

H
o
u
s
e
0
.
0
0
0
0
.
0
0
7
0
.
0
0
4
-
0
.
0
0
1
-
0
.
0
0
1
-
0
.
0
0
8
0
.
0
0
5
0
.
0
0
5
-
0
.
0
0
4
-
0
.
0
0
4
-
0
.
0
0
5
0
.
0
0
7
(
0
.
0
0
3
)
(
0
.
0
2
9
)
(
0
.
0
0
5
)
(
0
.
0
0
7
)
(
0
.
0
0
8
)
(
0
.
0
0
6
)
(
0
.
0
0
6
)
(
0
.
0
0
9
)
(
0
.
0
1
4
)
(
0
.
0
1
3
)
(
0
.
0
1
4
)
(
0
.
0
1
0
)
N
u
m
b
e
r

o
f

d
a
i
l
y

r
e
t
u
r
n
s
2
9
4
3
1
9
6
2
1
3
7
3
4
5
8
8
6
2
9
4
3
1
2
7
5
3
2
2
5
6
3
6
8
6
7
9
8
1
2
9
4
3
1
9
6
2
1
3
7
3
4
N
u
m
b
e
r

o
f

f
i
r
m
s
3
2
1
4
6
3
1
3
2
3
7
1
3
2
1
4
N
o
t
e
:

S
t
a
n
d
a
r
d

e
r
r
o
r
s

a
r
e

i
n

p
a
r
e
n
t
h
e
s
e
s
.

+

s
i
g
n
i
f
i
c
a
n
t

a
t

1
0
%
;

*

s
i
g
n
i
f
i
c
a
n
t

a
t

5
%
;

*
*

s
i
g
n
i
f
i
c
a
n
t

a
t

1
%
T
a
b
l
e
4
:
C
o
n
s
t
r
a
i
n
e
d
R
e
s
u
l
t
s
f
o
r
S
e
v
e
r
a
l
S
u
b
s
e
c
t
o
r
s
20
Dependent Variable: r
t
Basic
Materials
Consumer
Goods
Financials Technology Health Care Industrial
(1) (2) (3) (5) (4) (6)
Panel A: Basic Specification
1. One-day event window
Event 5 (AR) 0.005 0.003 -0.004 0.010 0.021 0.006
(0.002)** (0.003) (0.003) (0.008) (0.008)* (0.002)*
2. Three-day event window
Event 5 (CAR) 0.004 0.000 -0.013 0.009 0.053 0.004
(0.003) (0.005) (0.006)* (0.014) (0.014)** (0.004)
Panel B: Firm Heterogeneity
1. One-day event window
Event 5 (AR) 0.051 0.007 -0.010 0.017 0.023 0.004
(0.026)+ (0.005) (0.006) (0.011) (0.012)* (0.003)
LCR -0.0002 -0.0002 0.0025 0.0001 0.0010 0.0003
(0.001) (0.000) (0.005) (0.001) (0.001) (0.000)
SCALE 0.0020 0.0001 -0.0004 0.0007 0.0008 0.00001
(0.001)+ (0.000) (0.000) (0.001) (0.001) (0.000)
2. Three-day event window
Event 5 (CAR) 0.034 -0.003 -0.014 0.039 0.068 0.007
(0.046) (0.008) (0.011) (0.019)* (0.020)** (0.005)
LCR -0.0002 -0.0008 0.0049 -0.0025 -0.0035 0.0000
(0.002) (0.000) (0.009) (0.002) (0.002)* (0.000)
SCALE 0.0013 -0.0004 0.0002 0.0018 -0.0002 0.0005
(0.002) (0.000) (0.001) (0.001)+ (0.001) (0.001)
Number of daily returns 3294 14274 6954 2928 12444 36966
Number of firms 9 39 19 8 34 101
Note: TAR stands for abnormal returns; CAR cumulative abnormal returns. Standard errors are in parentheses. + significant at 10%; *
significant at 5%; ** significant at 1%
Table 5: Event 5 (Overcame the Last Hurdle) Only
21