The effect of antidumping and countervailing investigations on the market value of firms

International Review of Financial Analysis 36 (2014) 97–105

Contents lists available at ScienceDirect

International Review of Financial Analysis

The effect of antidumping and countervailing investigations on the market value of firms☆ Wanli Li a,b, Ziqiao Yan a,⁎, Wei Sun a a b

Xi'an Jiaotong University, PR China Shanghai University of International Business and Economics, PR China

a r t i c l e

i n f o

Available online 3 September 2014 JEL classification: F13 G14 Keywords: Antidumping Countervailing Stock market reaction Government assistance International business strategy

a b s t r a c t Using a sample of listed Chinese firms between 2006 and 2012, we analyze the effect of international business strategy and government assistance on the stock market response to antidumping and countervailing investigations. We find a significantly negative abnormal return surrounding the announcements of antidumping and countervailing investigations. Furthermore, the establishment of a plant in a non-subject or “non-named”1 country and government assistance are positively related to the abnormal returns of antidumping and countervailing investigations. Our results suggest that government assistance is as important as strategic restructuring to offset the negative effect of trade remedy investigations. © 2014 Elsevier Inc. All rights reserved.

1. Introduction In an increasingly competitive international market, antidumping and countervailing actions are used as motivation to protect domestic firms from possible injury by unfair trade (Brander & Spencer, 1981; Gayle & Puttitanun, 2009; Jung & Lee, 2003; Reynolds, 2006). With more than 40 World Trade Organization (WTO) member countries initiating 4230 antidumping investigations and 302 countervailing investigations from 1995 to 2012,2 these types of trade remedy actions against imports have become widespread (Feinberg & Reynolds, 2008; Schuler, Rehbein, & Cramer, 2002). Previous research on antidumping and countervailing investigations shows that antidumping and countervailing duties result in a significant change in business operation by exporting firms (Peng, Wang, & Jiang, 2008). These trade remedy actions lead to exporting firms' dramatically increasing the unit values of the named products to reduce the dumping margin (Avsar, 2013). However, the increasing price results in the loss of their competitive advantage (Brenton, 2001; Ganguli, 2008). As ☆ The financial support of the National Social Science Foundation of China (10BGL014) is acknowledged. ⁎ Corresponding author at: Xi'an Jiaotong University, No. 28, Xianning West Road, Xi'an, Shaanxi 710049, PR China. Tel.: +86 13991943159. E-mail addresses: [email protected] (W. Li), [email protected] (Z. Yan), [email protected] (W. Sun). 1 As antidumping and countervailing investigations are country-specific, antidumping and countervailing duties are imposed only on imports from countries named in the petition. Following Ganguli (2008), we call such country a named or subject country. In contrast, the country not named in the petition is called a non-named or non-subject country. 2 The data are from the semi-annual reports of WTO Members to the WTO Committee.

http://dx.doi.org/10.1016/j.irfa.2014.08.006 1057-5219/© 2014 Elsevier Inc. All rights reserved.

antidumping and countervailing investigations can directly affect a firm's profitability, it is important to examine the reaction strategy to these temporary trade barriers. Once the antidumping or countervailing duties are imposed, investigated firms may respond by locating production within the destination market to bypass tariff barriers (Belderbos, 1997; Blonigen, 2002; Blonigen, Tomlin, & Wilson, 2004). Expanding other markets is also an effective way to avoid a trade barrier (Bown & Crowley, 2007). In this study, we analyze the effect of international business strategy and government assistance on the stock market response to antidumping and countervailing investigations. We address two main questions. How does the stock market in an exporting country react to antidumping and countervailing investigations? What factors affect the stock market reaction to antidumping and countervailing investigations? We use a sample of listed Chinese firms that suffered antidumping and countervailing investigations between 2006 and 2012, to examine stock market reactions. China is a particularly useful starting point for a variety of reasons. First, China was the first leading target of antidumping and countervailing actions filed by other countries from 1995 to 2012.3 Specifically, there were 442 antidumping cases and 59 countervailing cases launched against Chinese firms from 2006 to 2012. Second, considering that China is a non-market economy country, antidumping agencies are likely to impose higher duties on Chinese firms (Zeng & Liang, 2010). This makes it necessary to find a way to offset the negative effect of trade remedy investigations.

3

These data can be accessed at http://www.wto.org/.

98

W. Li et al. / International Review of Financial Analysis 36 (2014) 97–105

Petition

Initiation

ITC

DOC

DOC Final

ITC Final

Preliminary

Preliminary

Determination

Determination

Determination

Determination

45 Days

140 Days*

215 Days*

45 Days* from

from

from

from

from

DOC Final

Petition

Petition

Initiation

Initiation

Determination

20

Days*

* These dates may be extended under certain circumstances. Fig. 1. Statutory time frame for antidumping investigations in the United States. This figure presents the statutory timeframe for antidumping investigations in the United States (US). Source: United States Department of Commerce (http://www.commerce.gov).

We find a significantly negative abnormal return surrounding the announcements of antidumping and countervailing investigations. When firms establish at least one manufacturing plant in a non-named country or obtain more government assistance, the market responds more favorably to antidumping and countervailing investigations. This result suggests that government assistance is as important as strategic restructuring when firms undergo antidumping and countervailing investigations. This study contributes to the literature in theoretical and practical ways. First, this study focuses on investors' expectations of the change in future firm value caused by antidumping and countervailing investigations. The influence on the export prices of products, trade diversion, and technology adoption decisions is considered in previous research (Avsar, 2013; Crowley, 2006; Prusa, 2001). In comparison, this study directly assesses how investors in the exporting country respond to antidumping and countervailing investigations and what factors affect this reaction. Our findings suggest that investors perceive the announcement of antidumping and countervailing investigations as an indication of the increasing operation risk and decreasing profitability in these firms. Second, the findings of this study also have important practical implications. Our results are consistent with those of Belderbos, Vandenbussche, and Veugelers (2004) and Peng et al. (2008), who investigated how firms may react to antidumping and countervailing duties using foreign direct investment to bypass tariff barriers. We find that, along with a timely adjustment of international business strategy, government assistance is another important way to offset the negative influence of antidumping and countervailing investigations. These findings significantly add to our knowledge about the role of government in antidumping and countervailing investigations. Furthermore, as antidumping and countervailing investigations are worldwide phenomena, the findings from this study can be extended to other counties that are a target for antidumping and countervailing actions. China's experience of offsetting the negative influence of trade remedy actions may be of benefit to other countries. The remainder of this paper is organized as follows. Section 2 develops the hypotheses and Section 3 lays out the methodology. In Section 4, we present a discussion of our results. Section 5 concludes the paper. 2. Hypothesis development To protect domestic producers from unfair trade, antidumping duties and countervailing duties are imposed on foreign export firms if they export a product at a price lower than normal value or receive specific subsidies (Gallaway, Blonigen, & Flynn, 1999; Wang, 2010). Antidumping and countervailing investigations comprise three basic events: petition, initiation, and decision. For example, Fig. 1 presents the

statutory time frame for antidumping investigations in the United States (US). A petition is filed by a domestic industry injured by dumped imports. If the United States Department of Commerce (USDOC) determines that a petition satisfies all requirements under the law, it will publish a notice of initiation. The United States International Trade Commission (USITC) determines whether the imports are materially injuring domestic industry within 45 days. If both the USDOC and USITC make affirmative findings, the USDOC assesses duties against imports of that product as a percentage of the value of the imports and publishes a preliminary decision. The USDOC final determination comes out 75 days later and the final event is the USITC final determination, which leads to the imposition of antidumping or countervailing duties. If one country launches its own investigation and ultimately charges extra duty on imports that are found to be hurting domestic producers, the price of these imports will increase. For example, Avsar (2013) finds that to reduce the dumping margin and avoid that threat, Brazilian exporters respond to antidumping duties by increasing the prices of their products to the target country. However, the increasing price results in them losing their competitive advantage. The market share of these exports decreases, and they are replaced by domestic products or other countries' products (Nieberding, 1999). Brenton (2001) and Ganguli (2008) focus on the trade diversion caused by antidumping cases and show that antidumping investigations divert trade flow from named countries. For example, in the case of US antidumping and countervailing investigations of photovoltaic modules and components from China, after the US made a ruling of 18.32%–249.96% antidumping duties and 14.78%–15.97% countervailing duties, the exports from China's photovoltaic producers to the US declined by 80% in 2012.4 As antidumping and countervailing investigations have a profound effect on firm business operation and profitability, exporting firms suffering the antidumping and countervailing investigations give a signal to the market that their firm performance is likely to be influenced by temporary trade barriers. Therefore, investors may perceive the announcement of antidumping and countervailing investigations as an indication of an increasing operation risk and decreasing cash flows in these firms. Thus, hypothesis 1 predicts that the announcement of antidumping and countervailing investigations is associated with a negative stock market reaction. Uncertainty about antidumping and countervailing investigations makes it difficult for investors to identify the earning prospects of investigated firms. Therefore, investors will look for alternative sources of information to differentiate the quality of firms. We argue that effective strategies responding to temporary trade barriers can serve as a strong signal to ascertain the prospects of the investigated firm.

4

The data is from the Ministry of Commerce of China (http://www.mofcom.gov.cn/).

W. Li et al. / International Review of Financial Analysis 36 (2014) 97–105

Using foreign direct investment (FDI) to bypass temporary trade barriers is an effective way to react to the antidumping and countervailing investigations (Belderbos & Sleuwaegen, 1998; Belderbos et al., 2004; Peng et al., 2008). As the antidumping and countervailing investigations are filed to exports from named countries, the extra duties will not be imposed on investigated firms when the products are manufactured in a non-named country. In a word, the establishment of a plant abroad can offset the negative effect of punitive duties. Blonigen (2002) demonstrates that FDI motivated by avoiding a temporary trade barrier can eliminate or reduce the positive influence of the trade remedy police on domestic firms and increase competition. Blonigen et al. (2004) also focus on the effect of tariff-jumping FDI on the result of antidumping cases. They show that the average abnormal gain of domestic firms is 1.5% and not statistically significant when tariff-jumping FDI does occur, while such gains increase by more than 50% and are significant when no tariff-jumping is involved in the case. Therefore, the investor's perceptions of the investigated firm's prospects may be influenced by the international business strategy adopted by the investigated firm. As a result, the market reaction to antidumping and countervailing investigations will be more favorable when the subject firms have established a plant in a non-named country. Thus, hypothesis 2 predicts that the negative stock market reaction to antidumping and countervailing investigations is weaker for investigated firms with at least one manufacturing plant in a non-named country. Another way to respond to antidumping and countervailing investigations is to increase exports to other countries. Due to the punitive duties imposed on products, the investigated firms have to face a reduction in their market share. In this case, they can export named products to other countries to deflect the threats of antidumping and countervailing duties. Bown and Crowley (2007) examine the effect of US safeguard and antidumping duties on Japanese exports of the same products to third countries and find that the extra duties lead to a 5–7% increase in Japanese exports to the third markets. Similarly, after the European Commission (EU) launched an antidumping and countervailing investigation into photovoltaic modules and components from China in 2012, Jiangxi Risun Solar dropped overseas sales to a third of the firm's production and began shifting its business focus to Asia. This strategic restructuring effectively offset the effect of the EU antidumping and countervailing investigation (Yang, 2013). If the market takes this information into account, the stock market will react more positively to antidumping and countervailing investigations when the investigated firm shifts its business focus to other country markets. Thus, hypothesis 3 predicts that the negative stock market reaction to antidumping and countervailing investigations is weaker for investigated firms with a higher share of sales to a third country. Similarly, increasing sales of other products can also be used to avoid temporary trade barriers. Antidumping and countervailing duties are imposed only on named imports, and have no effect on other products manufactured by the investigated firms. As a result, increasing nonnamed product sales is a way to relieve the negative influence of antidumping and countervailing investigations on operations. Thus, hypothesis 4 predicts that the negative stock market reaction to antidumping and countervailing investigations is weaker for investigated firms with a higher share of non-named product sales. In addition to the strategic restructuring of exporting firms, government assistance is also an important factor affecting the prospects of the investigated firm. Historically, trade remedy investigations have been found to be susceptible to political pressure (Baldwin & Moore, 1991; Finger, Hall, & Nelson, 1982). Lee and Baik (2010) find that petitioning firms gain more proceeds from antidumping cases if they spend more on lobbying. This result indirectly indicates the effect of government on injury cases. In the case of the EU antidumping and countervailing investigation of Chinese photovoltaic modules and components, the EU made a preliminary ruling of 11.8% antidumping duties due to the Chinese government's diplomatic effort. This duty was much lower than expected, and was

99

canceled when the EU reached an agreement with China on minimum import prices. The government of a named country not only plays an important role in negotiations, but can also use a series of measures to boost a sagging sector due to subdued overseas demand, including offering financial support. For example, in the case of the Chinese photovoltaic industry, the government introduced a new stimulus package to offset the negative influence of the EU's preliminary ruling. This stimulus package involves subsidizing on-grid solar electricity for 20 years and providing tax breaks to photovoltaic companies that are participating in mergers, acquisitions, or reorganization (Yang, 2013). In such instances, the antidumping and countervailing investigations may be perceived as having a weak negative effect on investigated firms. Therefore, government assistance can be used by investors to judge the effect of antidumping and countervailing investigations. Thus, hypothesis 5 predicts that the negative stock market reaction to antidumping and countervailing investigations is weaker for investigated firms with more government assistance. 3. Methodology 3.1. Sample and data The sample in this study contains companies trading in the Shanghai and Shenzhen Stock Exchange that released announcements about antidumping and countervailing investigations between 2006 and 2012. We obtain the financial and market information from the Chinese Stock Market and Accounting Research database (CSMAR). We supplement government subsidy data from the firms' annual reports. We search the antidumping and countervailing announcements on the web sites of the Shanghai and Shenzhen Stock Exchange (http://www. sse.com.cn and http://www.szse.cn). A set of key terms, including antidumping, dumping, countervailing, countervailable subsidy, trade remedies, sunset review, and administrative review, is used in the search. We eliminate the announcements reporting Chinese companies filing an antidumping or countervailing petition to foreign firms, repeated announcements, and those that might be confounded by any other information announcements, such as dividends and earnings, around the announcement period. Our final sample consists of 97 announcements. 3.2. Measures 3.2.1. Dependent variable We use an event study methodology to measure the stock market reaction to antidumping and countervailing investigations with cumulative abnormal returns (Blonigen et al., 2004; Marsh, 1998). We estimate abnormal returns (AR) for firms from announcements of antidumping and countervailing investigations after controlling for general market movements, AR it ¼ Rit −Rmt

ð1Þ

where Rit is the return of stock i on day t, and Rmt is the weighted average returns of the Shanghai or Shenzhen stock exchange. The Shanghai synthesis index return is used as Rmt if the firms are listed in the Shanghai stock exchange, while the Shenzhen composition index return is used for firms listed in the Shenzhen stock exchange. The cumulative abnormal return (CAR) over any time period [t1, t2] is defined as, CARi ðt1 ; t2 Þ ¼

t2 X ARit :

ð2Þ

t¼t1

Following previous research (Hendricks & Singhal, 2008; Im, Dow, & Grover, 2001; Jacobs, Singhal, & Subramanian, 2010), a two-day event

100

W. Li et al. / International Review of Financial Analysis 36 (2014) 97–105

period is used in our study, consisting of the announcement date (day 0) and the day before the announcement (day −1). If the announcement is made on a non-trading day, the next trading day is set to the announcement date. While our primary focus is on the abnormal returns for day −1 to day 0, the abnormal returns for day −5 to 5 and day −1 to 1 are also reported. For robustness, we use the Fama–French three-factor model (Fama & French, 1993) to estimate the abnormal return and cumulative abnormal return in sensitivity analysis. We begin by estimating the three-factor model for each firm, Ri −R f ¼ ai þ bi ðRm −R f Þ þ si SMB þ hi HML þ εi

ð3Þ

where Ri − Rf is the excess return over the risk-free return rate on stock i, and Rm − Rf is the excess market return over the risk-free return rate. SMB is the difference between a small firm return and big firm return. HML is the difference between the return on a high book-to-market equity group and a low book-to-market equity group. For each sample firm, the estimation period is 200 trading days from day − 210 to day − 11. Consistent with past studies (Hendricks & Singhal, 2008), we leave two weeks (10 trading days) before the announcement date to ensure that the estimates are shielded from the announcement. 3.2.2. Independent variables We capture the establishment of a plant in a non-named country using a dummy variable, plant abroad, which is equal to 1 if the investigated firms have established at least one manufacturing plant in a nonnamed country. As it is difficult to collect data on sales in the country that filed the antidumping and countervailing investigation, we use the share of overseas sales to examine hypothesis 3. If the export ratio, which is the proportion of exports in revenues, is high, we expect the negative stock market reaction to antidumping and countervailing investigations to be stronger. The investigated product ratio is calculated as the proportion of investigated products in revenues, and used to test hypothesis 4. We measure government assistance using government subsidy, which is calculated as government subsidies scaled by revenue. All of the firm characteristics are based on the data at the beginning of the year. 3.2.3. Control variables We include the following control variables to rule out alternative explanations for abnormal returns. First, the effect of antidumping and countervailing announcements on the market depends on how much additional information they reveal (Belderbos et al., 2004). The new information contained in announcements is different at each stage of the investigation. The petition and initiation are the beginning of the investigation and are likely to be regarded as a strong signal to market. However, the uncertainty of the final determination makes it difficult for investors to identify the earning prospects of the investigated firms (Blonigen et al., 2004). The preliminary and final decisions include dumping or subsidy margins, which lead to the imposition of antidumping or countervailing duties and present new information to the market. Compared with these events, administrative review, sunset review, and other events may not convey much new information to the market. Therefore, we use dummy variables for initiation and determination to distinguish the effect of stage on market reaction. Second, different kinds of news contained in announcements may affect the market reaction. Blonigen et al. (2004) include antidumping duty to control this problem. Similarly, we create a dummy variable, bad news, to capture whether the information mentioned in the announcement is favorable or unfavorable, expecting bad news to be related to negative abnormal returns. Unfavorable results of investigations include the existence of material injury or higher duty rates, while favorable results include petition withdrawal, no material injury, lower duty rates than before, and no effect on business operations.

Table 1 Sample distribution by year and industry. This table reports sample distribution by year and by industry. The sample consists of 97 announcements of antidumping and countervailing investigations released by Chinese public firms from 2006 to 2012. Panel A: distribution by year Year

Number

Percentage

2006 2007 2008 2009 2010 2011 2012 Total

3 4 14 18 18 23 17 97

3.09% 4.12% 14.43% 18.56% 18.56% 23.71% 17.53% 100%

Industry

Number

Percentage

Wood, cork and articles Paper, paperboard and articles Products of the chemical and allied industries Machinery and electrical equipment Base metals and articles Miscellaneous manufactured articles Total

9 5 19 37 20 7 97

9.28% 5.15% 19.59% 38.14% 20.62% 7.22% 100%

Panel B: distribution by industry

Lastly, to control unobserved macroeconomic conditions and industry characteristics, we also include year and industry dummies.

4. Empirical results 4.1. Descriptive statistics Table 1 represents the distributions of our sample events by year and by industry. The distribution by year in panel A shows that the number of antidumping and countervailing announcements increased after the financial crisis of 2008. This phenomenonis consistent with empirical evidence provided by previous research showing that antidumping and countervailing investigations increase during periods of macroeconomic weakness (Takacs, 1981; Leidy, 1997; Knetter & Prusa, 2003; Niels & Francois, 2006; Bown, 2011). From an industry-based point of view, panel B shows that machinery and electrical equipment firms reported the most number of antidumping and countervailing announcements, followed by the base metals and articles industry. The descriptive statistics and correlations for the key variables are reported in Table 2. Over the two-day event window, on average there are negative abnormal returns for firms undergoing antidumping and countervailing investigations. Twenty-three percent of investigated firms have at least one manufacturing plant in a non-named country. The average export ratio is 0.41, and the average investigated product ratio is 0.65. The average government subsidy provided to these firms is 0.02. Plants in non-named countries and government subsidy have a significantly positive correlation with cumulated abnormal returns. Table 3 reports the abnormal returns and cumulative average abnormal returns for windows surrounding the event day. We examine the sign of the means with a t-test and that of the medians using the Wilcoxon signed-rank test. The mean of abnormal returns for the day of announcement (day 0) is negative and significantly different from zero at the 1% level. The firms lose an average of 0.91% of their market value with the significant level of 5% in day −1 to day 0 and 0.86% in day −1 to day 1.The medians of abnormal returns for day 0 and cumulative abnormal returns for day −1 to day 0 and day −1 to day 1 are also significantly negative. This result supports hypothesis 1 and suggests that investors perceive the announcement of antidumping and countervailing investigations as a signal of the poor future performance of a firm.

W. Li et al. / International Review of Financial Analysis 36 (2014) 97–105

101

Table 2 Descriptive statistics and correlations. This table reports means, standard deviations, median, minimum, maximum and correlations for key variables. CAR (−1,0) is the cumulative abnormal return over day −1 to day 0 (estimated from Eq. (2)). Plant abroad is equal to 1 if the investigated firms have established at least one manufacturing plant in non-named countries. Exports ratio is the proportion of exports in revenues. Investigated product ratio is calculated as the proportion of investigated products in revenues. Government subsidy is calculated as government subsidies scaled by revenues. Bad news is a dummy variable which is equal to 1 if the information mentioned in announcement is unfavorable. All of the firm characteristics are based on the data at beginning of the year. The sample consists of 97 events from 2006 to 2012. Correlations larger than 0.17 are significant at the level of p b 0.05, and those larger than 0.23 are significant at the level of p b 0.01. Variables

Mean

S.D.

Median

Min

Max

1

2

3

4

5

6

1. CAR(−1,0) 2. Plant abroad 3. Exports ratio 4. Investigated product ratio 5. Government subsidy 6. Bad news

−0.01 0.23 0.41 0.65 0.02 0.74

0.03 0.42 0.26 0.28 0.02 0.44

−0.01 0.00 0.43 0.71 0.01 1.00

−0.11 0.00 0.01 0.05 0.00 0.00

0.08 1.00 0.99 1.00 0.13 1.00

1.00 0.20 0.06 0.11 0.20 −0.25

1.00 0.02 0.17 −0.11 0.04

1.00 0.47 0.00 0.01

1.00 −0.04 0.03

1.00 −0.05

1.00

4.2. Multivariate regression results Table 4 reports the multivariate regression results with the cumulative abnormal return over the two-day event window as the dependent variable. In this table, model 1 includes the control variables only, models 2 to 5 include the predictor variables separately, and model 6 includes all the predictor variables. These models explain 27.8% to 40% of the variation in the cumulative abnormal return. Model 2 contains the plant abroad variable to test hypothesis 2. As predicted, the establishment of a plant in a non-named country is positively associated with the cumulative abnormal return, at a significance level of 1% (b = 0.026, p b 0.01). This result supports hypothesis 2 that antidumping and countervailing investigations are perceived as having a weakly negative effect on investigated firms with at least one plant in a nonnamed country. This finding is consistent with those of Belderbos et al. (2004) and Peng et al. (2008), who suggest that tariff-jumping FDI is an effective way to bypass temporary trade barriers. Hypotheses 3 and 4 predict that a higher share of non-named product sales and a higher share of sales to other countries will positively affect the firm's abnormal stock returns associated with antidumping and countervailing investigations. Models 3 and 4 separately test these two hypotheses. We do not find that the stock market responds more positively when firms have higher non-named product or third market sales. Therefore, hypotheses 3 and 4 are not supported. This result is

Table 3 Abnormal return and cumulative average abnormal return for windows surrounding the event day. This table reports the abnormal returns and cumulative abnormal returns for windows surrounding the event day. Abnormal returns are estimated from Eq. (1). Cumulative abnormal returns are estimated from Eq. (2). The sample consists of 97 events from 2006 to 2012. Panel A: abnormal return Event day

Mean

Median

T-test

Wilcoxon signed-rank test

−5 −4 −3 −2 −1 0 1 2 3 4 5

0.63% −0.14% 0.27% 0.06% 0.20% −1.11% 0.05% 0.14% −0.29% −0.11% 0.14%

0.18% −0.35% 0.13% −0.04% −0.09% −1.03% −0.51% −0.20% −0.36% −0.24% −0.36%

2.40⁎⁎ −0.51 0.91 0.25 0.81 −3.95⁎⁎⁎

1.49 −1.00 0.34 −0.03 −0.22 −4.21⁎⁎⁎ −1.07 −0.59 −1.46 −0.62 −0.36

0.16 0.43 −1.11 −0.46 0.49

Panel B: cumulative abnormal return Window

Mean

Median

T-test

Wilcoxon signed-rank test

(−5 to 5) (−1 to 1) (−1 to 0)

−0.16% −0.86% −0.91%

−0.22% −0.81% −0.73%

−0.20 −1.98⁎⁎ −2.61⁎⁎

−0.40 −2.31⁎⁎ −2.42⁎⁎

⁎⁎ Indicates significance at 5% level. ⁎⁎⁎ Indicates significance at 1% level.

inconsistent with Bown and Crowley's (2007) findings that the temporary trade barrier distorts a foreign country's exports to third markets. A possible explanation is that the data used to measure these two variables cannot make a clear distinction between sales to the country imposing antidumping or countervailing duties and sales to other countries. Hypothesis 5 predicts that more government assistance will positively affect an investigated firm's abnormal stock returns associated with antidumping and countervailing investigations. The coefficient for government subsidy is significantly positive (b = 0.352, p b 0.01) in model 5. This result indicates that the market may view government subsidy as a signal of effective help to investigated firms. This result is qualitatively consistent with those of Finger et al. (1982) and Baldwin and Moore (1991), who suggest that temporary trade investigations are susceptible to political pressure. Therefore, hypothesis 5 is supported. Column 6 reports the result of a model with all of the predictor variables added. The establishment of a plant overseas and government subsidy are significantly and positively associated with abnormal returns surrounding the announcement of antidumping and countervailing investigations. This result suggests that government assistance is as important as suitable strategic restructuring when firms undergo antidumping and countervailing investigations. For robustness, we use the Fama–French three-factor model to estimate abnormal return and cumulative abnormal return in a sensitivity analysis. Tables 5 and 6 provide the abnormal returns, cumulative average abnormal returns, and multivariate regression results. When the three-factor model is used, the average abnormal return is still significant and negative on day 0. The investigated firms lose 0.88% of their firm value within a two-day window and 0.79% within a three-day window. Similarly, the regression results show that a manufacturing plant in a non-named country and government assistance are positively related to abnormal returns surrounding antidumping and countervailing announcements. Overall, our results remain qualitatively the same when we use the Fama–French three-factor model to estimate the abnormal return. We also alter the event window to consider the effects of other periods in our unreported results, calculating a longer 3-day window (day −1 to day 1) and an 11-day window (day −5 to day 5). The results for the 3-day window are generally consistent with those of our 2-day window in both direction and significance. The results from our 11-day window are consistent in direction but less significant. 4.3. Further tests: moderating effects All of the producers' exported merchandise in the scope of antidumping or countervailing investigations is affected by the trade remedy actions. As disclosure of the information regarding these investigations is voluntary, some companies announced the antidumping or countervailing investigations, but others did not. Therefore, it is important to incorporate this moderating effect in the analysis. If a company produced or exported merchandise included in the scope of the investigation but did not announce it, we define it as a non-announcement

102

W. Li et al. / International Review of Financial Analysis 36 (2014) 97–105

Table 4 Regression results for cumulative average abnormal return. This table reports multivariate regression results with the cumulative abnormal return over two-day event window as the dependent variables. CAR (−1, 0) is the cumulative abnormal return over day −1 to day 0 (estimated from Eq. (2)). Plant abroad is equal to 1 if the investigated firms have established at least one manufacturing plant in non-named country. Exports ratio is the proportion of exports in revenues. Investigated product ratio is calculated as the proportion of investigated products in revenues. Government subsidy is calculated as government subsidies scaled by revenues. Bad news is a dummy variable which is equal to 1 if the information mentioned in announcement is unfavorable. All of the firm characteristics are based on the data at beginning of the year. To control fixed effect, year and industry variables are also included. The sample consists of 97 events from 2006 to 2012. R-square in models 2 to 6 are compared with model 1, and the differences report an increase in r-square. Robust standard errors are reported in parentheses. Variables

(1)

(2)

(3)

(4)

(5)

(6)

Constant

−0.046⁎⁎⁎ (0.016)

−0.072⁎⁎⁎ (0.015)

−0.046⁎⁎⁎ (0.016)

−0.046⁎⁎⁎ (0.015)

−0.051⁎⁎⁎ (0.015)

−0.076⁎⁎⁎ (0.016)

0.352⁎⁎⁎ (0.103)

0.025⁎⁎ (0.010) −0.005 (0.020) 0.009 (0.022) 0.358⁎⁎⁎ (0.115)

Explanatory variables Plant abroad

0.026⁎⁎⁎ (0.009)

Exports ratio

0.004 (0.017)

Investigated product ratio

0.015 (0.018)

Government subsidy

Control variables Bad news

Determination

−0.021⁎⁎ (0.010) 0.008 (0.012) 0.018⁎

−0.024⁎⁎ (0.010) 0.012 (0.011) 0.022⁎⁎

−0.021⁎⁎ (0.010) 0.008 (0.012) 0.018⁎

−0.021⁎⁎ (0.011) 0.007 (0.012) 0.018⁎

−0.022⁎⁎ (0.010) 0.011 (0.011) 0.020⁎⁎

−0.025⁎⁎ (0.010) 0.015 (0.011) 0.024⁎⁎⁎

Industry Year N Adjusted r-squared R-squared Increase in r-square

(0.010) Yes Yes 97 0.134 0.278 –

(0.009) Yes Yes 97 0.220 0.358 0.080⁎⁎⁎

(0.010) Yes Yes 97 0.123 0.278 0.000

(0.010) Yes Yes 97 0.134 0.288 0.010

(0.009) Yes Yes 97 0.172 0.319 0.041⁎⁎⁎

(0.009) Yes Yes 97 0.242 0.400 0.122⁎⁎⁎

Initiation

⁎ Indicates significance at 10% level. ⁎⁎ Indicates significance at 5% level. ⁎⁎⁎ Indicates significance at 1% level.

investigated firm and include it in our new sample. After excluding firms with missing data or those confounded by any other information announcements, the final sample consists of 442 observations.

Table 5 Robustness test for Fama–French three-factor model—AR and CAR. This table reports the abnormal returns and cumulative abnormal returns for windows surrounding the event day when Fama–French three-factor model is used to estimate AR and CAR. Abnormal returns are estimated from Eq. (3). Cumulative abnormal returns are estimated from Eq. (2). The sample consists of 97 events from 2006 to 2012. Panel A: abnormal return Event day

Mean

Median

T-test

Wilcoxon signed-rank test

−5 −4 −3 −2 −1 0 1 2 3 4 5

0.48% −0.21% 0.33% −0.11% 0.26% −1.14% 0.08% −0.01% −0.33% −0.18% −0.02%

−0.09% −0.36% −0.08% −0.37% −0.26% −1.27% −0.49% −0.56% −0.50% −0.28% −0.16%

1.82⁎ −0.75 1.15 −0.49 1.06 −4.18⁎⁎⁎

0.62 −1.63 0.32 −0.90 −0.42 −4.50⁎⁎⁎ −1.05 −1.39 −2.42⁎⁎

0.27 −0.04 −1.40 −0.82 −0.07

−1.20 −0.93

Panel B: cumulative average abnormal return Window

Mean

Median

T-test

Wilcoxon signed-rank test

(−5 to 5) (−1 to 1) (−1 to 0)

−0.85% −0.79% −0.88%

−1.38% −1.09% −0.68%

−1.02 −1.89⁎ −2.69⁎⁎⁎

−1.83⁎ −2.40⁎⁎ −2.81⁎⁎⁎

⁎ Indicates significance at 10%level. ⁎⁎ Indicates significance at 5% level. ⁎⁎⁎ Indicates significance at 1% level.

We conduct a univariate analysis to examine whether there is any difference in the cumulative abnormal return between investigated firms with and without announcements. As reported in Table 7, panel A, we find that the average cumulative abnormal return is significantly lower in the investigated firms with announcement (−0.97%) than in the investigated firms without announcement (0.03%). It appears that the announcements highlight the negative effect of antidumping or countervailing investigations. Therefore, the market reaction is stronger for firms with announcements than for firms without. In addition, we also examine whether the establishment of a plant in a non-subject country and government assistance affect the cumulative abnormal return in these two groups. Panel B in Table 7 provides the mean test results between the different groups, firms with (without) announcement and firms with (without) a plant abroad. Consistent with the results in Table 4, in the groups of firms reporting an announcement, the average cumulative abnormal return is significantly higher in firms with a plant in a non-subject country (0.33%) than in firms without (− 1.27%) (t-statistic is −1.944). For the firms without a plant abroad, the announcement of antidumping or countervailing investigations led to a more negative market reaction (t-statistic is 3.345). Panel C in Table 7 provides the mean test results for firms with (without) announcement and firms with more (less) government subsidy. We define a firm with more government subsidy as one that receives more government subsidy than the median. The average cumulative abnormal return is −0.12% in firms that reported an announcement and received more government subsidy and −1.71% in firms that reported an announcement but received less government subsidy. The difference is significant at the 5% level (t-statistic is −2.343). We also find that in the groups with less government subsidy, the average cumulative abnormal return is significantly higher if the firms did not announce (t-statistic is 3.373). Taken together, our mean test results suggest that the announcement of antidumping and countervailing investigations

W. Li et al. / International Review of Financial Analysis 36 (2014) 97–105

103

Table 6 Robustness test for Fama–French three-factor model—regression results. This table reports multivariate regression results with the cumulative abnormal return over two-day event window as the dependent variables when Fama–French three-factor model is used to estimate AR and CAR. CAR (−1,0) is the cumulative abnormal return over day −1 to day 0 (estimated from Eq. (2)). Plant abroad is equal to 1 if the investigated firms have established at least one manufacturing plant in non-named country. Exports ratio is the proportion of exports in revenues. Investigated product ratio is calculated as the proportion of investigated products in revenues. Government subsidy is calculated as government subsidies scaled by revenues. Bad news is a dummy variable which is equal to 1 if the information mentioned in the announcement is unfavorable. All of the firm characteristics are based on the data at beginning of the year. To control fixed effect, year and industry variables are also included. The sample consists of 97 events from 2006 to 2012. R-square in models 2 to 6 are compared with model 1, and the differences report an increase in r-square. Robust standard errors are reported in parentheses. Variables

(1)

(2)

(3)

(4)

(5)

(6)

Constant

−0.022 (0.019)

−0.044⁎⁎ (0.018)

−0.022⁎ (0.019)

−0.022 (0.018)

−0.025 (0.019)

−0.046⁎⁎ (0.017)

0.232⁎⁎ (0.104)

0.021⁎⁎ (0.009) 0.003 (0.016) 0.007 (0.016) 0.233⁎⁎ (0.113)

Explanatory variables Plant abroad

0.022⁎⁎ (0.009)

Exports ratio

0.010 (0.015)

Investigated product ratio

0.015 (0.014)

Government subsidy

Control variables Bad news Initiation Determination Industry Year N Adjusted r-squared R-squared Increase in r-square

−0.019⁎⁎ (0.009) −0.001 (0.012) 0.013 (0.009) Yes Yes 97 0.165 0.304 –

−0.021⁎⁎ (0.009) 0.004 (0.011) 0.017⁎ (0.009) Yes Yes 97 0.233 0.368 0.064⁎⁎

−0.019⁎⁎ (0.009) 0.000 (0.012) 0.013 (0.009) Yes Yes 97 0.159 0.308 0.004

−0.019⁎⁎ (0.009) −0.001 (0.012) 0.013 (0.009) Yes Yes 97 0.168 0.316 0.012

−0.019⁎⁎⁎ (0.009) 0.002 (0.012) 0.015 (0.009) Yes Yes 97 0.179 0.324 0.020⁎⁎

−0.022⁎⁎ (0.009) 0.005 (0.011) 0.018⁎ (0.009) Yes Yes 97 0.231 0.391 0.087⁎⁎

⁎ Indicates significance at 10% level. ⁎⁎ Indicates significance at 5% level. ⁎⁎⁎ Indicates significance at 1% level.

gives a stronger signal to the market. When the firms reported the announcement, the market responded more favorably if they established at least one manufacturing plant in a non-named country or received more government assistance. Next we ran multivariate regressions and Table 8 reports the results. Non-announcement is a dummy variable that is equal to 1 if a company produced or exported merchandise included in the scope of an investigation but did not announce. Column 1 adds the interaction of a plant abroad and non-announcement. When the holding levels of the other variables are fixed, we find that when the firms announced, the cumulative abnormal return is significantly higher in firms with a plant in a nonsubject country than in firms without (b = 0.019, p b 0.05). When the firms did not establish a plant abroad, a firm without announcement has a predicted 1.3% cumulative abnormal return higher than a firm that announced the information about antidumping or countervailing investigations. The differential between those that did not announce but have a plant abroad, relative to those that did not have a plant abroad but announced, is about 1.8%, with the same levels for the other variables. Column 2 adds the interaction of government subsidy and non-announcement. Similarly, when the firms announced, the market reaction was more positive if the firms obtained more government assistance (b = 0.216, p b 0.05). The intercept for firms with announcement is below that for firms without announcement, but the gap narrows as government subsidy increases (coefficient for interaction term is − 0.202). We add all of the interaction variables together in column 3 and the results are consistent with those in column 1 and column 2. Our results indicate that the market reacts more negatively to firms that report the announcement of antidumping or countervailing investigations. The negative effect of announcement on abnormal returns is significantly weaker when the firms established a plant in a non-subject country or receive more government assistance.

5. Conclusion Building on previous studies that found that antidumping and countervailing duties result in a significant change in the business operation of exporting firms, we examined the effect of international business strategy and government assistance on how the stock market reacts to antidumping and countervailing investigations. Our results show that the exporting firm subject to the antidumping and countervailing investigations is associated with a negative stock market reaction. As it is hard to clearly ascertain the effect on earning prospects of the investigated firms, investors utilize effective strategy responses to temporary trade barriers as a signal to determine the prospect of investigated firms. Consistent with the past finding that tariff-jumping FDI is an effective way to bypass temporary trade barriers (Belderbos et al., 2004; Peng et al., 2008), we find that investigated firms that established at least one manufacturing plant in a non-named country enjoyed better abnormal stock returns associated with antidumping and countervailing investigations. As extra duties will not be imposed on investigated firms when the products are manufactured in these plants, investors lend more credence to these firms' earnings prospects. The results of this study also show that government assistance positively affects the abnormal stock returns associated with antidumping and countervailing investigations. This finding suggests that if the government of a named country offers financial support to boost investigated firms, the earning prospects are more positive for investors. Therefore, the stock market responds more positively when an investigated firm is associated with more government assistance. Overall, our results indicate that in addition to a timely adjustment of international business strategy, government assistance is an important way to offset the negative influence of antidumping and countervailing investigations. This study contributes to the empirical literature examining the effect of antidumping and countervailing investigations on named firms.

104

W. Li et al. / International Review of Financial Analysis 36 (2014) 97–105

Table 7 Cumulative average abnormal return in different groups. This table reports the cumulative abnormal return over two-day event window in different groups. Cumulative abnormal returns are estimated from Eq. (2). Plant abroad is equal to 1 if the investigated firms have established at least one manufacturing plant in a non-named country. Government subsidy is calculated as government subsidies scaled by revenues. All of the firm characteristics are based on the data at beginning of the year. If a company produced or exported merchandise included in the scope of the investigation but did not announce it, we define it as a non-announcement investigated firm. We define a firm with more government subsidies as one that receives more government subsidies than the median. Panel A: announcement Group

N

Mean

Announced Not announced

97 345

−0.91% 0.03%

S.D. 0.034 0.029

T-test −2.716⁎⁎⁎

Table 8 The impact of announcement on cumulative abnormal return. This table reports multivariate regression results with the cumulative abnormal return over two-day event window as the dependent variables. CAR (−1, 0) is the cumulative abnormal return over day −1 to day 0 (estimated from Eq. (2)). Plant abroad is equal to 1 if the investigated firms have established at least one manufacturing plant in a non-named country. Government subsidy is calculated as government subsidies scaled by revenues. Non-announcement is a dummy variable that is equal to 1 if a company produced or exported merchandise included in the scope of an investigation but did not announce. Bad news is a dummy variable which is equal to 1 if the information mentioned in the announcement is unfavorable. All of the firm characteristics are based on the data at beginning of the year. To control fixed effect, year and industry variables are also included. The sample consists of 442 events from 2006 to 2012. Robust standard errors are reported in parentheses. Variables

(1)

(2)

(3)

Constant

−0.034⁎⁎ (0.014)

−0.036⁎⁎ (0.014)

−0.039⁎⁎⁎ (0.014)

0.013⁎⁎⁎ (0.004) 0.019⁎⁎ (0.008) −0.014 (0.009)

0.014⁎⁎⁎ (0.005)

0.018⁎⁎⁎ (0.005) 0.020⁎⁎⁎ (0.008) −0.015⁎

Panel B: plant abroad Group Announced

Not announced

With plant abroad Mean S.D. N Mean S.D. N

T-test

0.33% 0.032 22 −0.07% 0.021 31 −0.547

Without plant abroad

T-test

−1.27% 0.034 75 0.04% 0.029 314 3.345⁎⁎⁎

−1.944⁎⁎

Plant abroad 0.203

Announced

Not announced

T-test

Mean S.D. N Mean S.D. N

Non-announcement ∗ plant abroad Government subsidy Non-announcement ∗ government subsidy

Panel C: government subsidy Group

Explanatory variables Non-announcement

More government subsidy

Less government subsidy

T-test

−0.12% 0.032 49 0.07% 0.022 173 0.411

−1.71% 0.035 48 0.01% 0.030 172 3.373⁎⁎⁎

−2.343⁎⁎

−0.278

⁎⁎ Indicates significance at 5% level. ⁎⁎⁎ Indicates significance at 1% level.

While previous studies have mainly focused on the export prices of products (Avsar, 2013; Konings & Vandenbussche, 2005), trade diversion (Malhotra & Malhotra, 2008; Prusa, 2001), and technology adoption decisions (Crowley, 2006; Gao & Miyagiwa, 2005), this study directly assesses investors' expectations of changes in firms' future value caused by antidumping and countervailing investigations. Using an event study methodology to measure the abnormal response in stock returns surrounding the antidumping and countervailing events, the possible earnings management problem of investigated firms (Jones, 1991; Magnan, Nadeau, & Cormier, 1999) can be avoided in analysis. This study also contributes to the research focusing on response strategies to trade remedy investigations. Previous studies have shown that the strategic restructuring of investigated firms, such as tariffjumping FDI and trade deflection, is useful in responding to trade barriers (Belderbos, 1997; Blonigen, 2002; Bown & Crowley, 2007). Our findings indicate that the effect of government assistance on investigated firms subject to antidumping and countervailing investigations is also significantly positive. We acknowledge a few limitations of our study, which also present some interesting directions for future research. First, while our study focuses on the stock market reactions surrounding antidumping and countervailing announcements, it might also be interesting to compare investor reactions from firms with different extra duties imposed in the same case. Second, as our results show that providing government subsidies is an effective way to offset the negative effect of trade remedy investigations, future research could examine the effect of other forms of government assistance on antidumping and countervailing actions (e.g., diplomatic effort). In conclusion, we analyzed the effect of international business strategy and government assistance on how the stock market responds to

Control variables Bad news Initiation Determination Industry Year N Adjusted r-squared

−0.013⁎⁎⁎ (0.004) 0.010⁎ (0.006) 0.007 (0.005) Yes Yes 442 0.096

0.216⁎⁎ (0.095) −0.202⁎ (0.107)

(0.009) 0.236⁎⁎ (0.093) −0.224⁎⁎ (0.104)

−0.013⁎⁎⁎ (0.004) 0.010⁎ (0.006) 0.008 (0.005) Yes Yes 442 0.085

−0.013⁎⁎⁎ (0.004) 0.010⁎ (0.006) 0.008 (0.005) Yes Yes 442 0.099

⁎ Indicates significance at 10% level. ⁎⁎ Indicates significance at 5% level. ⁎⁎⁎ Indicates significance at 1% level.

antidumping and countervailing investigations. We propose and find that a plant in a non-named country and government assistance can serve as market signals of the earning prospects of investigated firms, and thus affect the stock market response to antidumping and countervailing investigations. References Avsar, V. (2013). Antidumping, retaliation threats, and export prices. World Bank Economic Review, 27, 133–148. Baldwin, R., & Moore, M. (1991). Political aspects of the administration of the trade remedy laws. Washington, DC: Brookings Institution. Belderbos, R. A. (1997). Antidumping and tariff jumping: Japanese firms' DFI in the European Union and the United States. Weltwirtschaftliches Archiv—Review of World Economics, 133, 419–457. Belderbos, R., & Sleuwaegen, L. (1998). Tariff jumping DFI and export substitution: Japanese electronics firms in Europe. International Journal of Industrial Organization, 16, 601–638. Belderbos, R., Vandenbussche, H., & Veugelers, R. (2004). Antidumping duties, undertakings, and foreign direct investment in the EU. European Economic Review, 48, 429–453. Blonigen, B. A. (2002). Tariff-jumping antidumping duties. Journal of International Economics, 57, 31–49. Blonigen, B. A., Tomlin, K., & Wilson, W. W. (2004). Tariff-jumping FDI and domestic firms' profits. Canadian Journal of Economics, 37, 656–677. Bown, C. P. (2011). taking stock of antidumping, safeguards and countervailing duties, 1990–2009. World Economy, 34, 1955–1998. Bown, C. P., & Crowley, M. A. (2007). Trade deflection and trade depression. Journal of International Economics, 72, 176–201. Brander, J. A., & Spencer, B. J. (1981). Tariffs and the extraction of foreign monopoly rents under potential entry. Canadian Journal of Economics, 14, 371–389. Brenton, P. (2001). Anti-dumping policies in the EU and trade diversion. European Journal of Political Economy, 17, 593–607. Crowley, M. A. (2006). Do safeguard tariffs and antidumping duties open or close technology gaps? Journal of International Economics, 68, 469–484.

W. Li et al. / International Review of Financial Analysis 36 (2014) 97–105 Fama, E. F., & French, K. R. (1993). Common risk-factors in the returns on stocks and bonds. Journal of Financial Economics, 33, 3–56. Feinberg, R. M., & Reynolds, K. M. (2008). Friendly fire? The impact of US antidumping enforcement on US exporters. Review of World Economics, 144, 366–378. Finger, M., Hall, K., & Nelson, D. (1982). The political economy of administered protection. American Economic Review, 72, 452–466. Gallaway, M. P., Blonigen, B. A., & Flynn, J. E. (1999). Welfare costs of the US antidumping and countervailing duty laws. Journal of International Economics, 49, 211–244. Ganguli, B. (2008). The trade effects of Indian antidumping actions. Review of International Economics, 16, 930–941. Gao, X. W., & Miyagiwa, K. (2005). Antidumping protection and R&D competition. Canadian Journal of Economics—Revue Canadienne D Economique, 38, 211–227. Gayle, P. G., & Puttitanun, T. (2009). Has the Byrd amendment affected US imports? World Economy, 32, 629–642. Hendricks, K. B., & Singhal, V. R. (2008). The effect of supply chain disruptions on shareholder value. Total Quality Management and Business Excellence, 19, 777–791. Im, K. S., Dow, K. E., & Grover, V. (2001). Research report: A reexamination of IT investment and the market value of the firm—An event study methodology. Information Systems Research, 12, 103–117. Jacobs, B. W., Singhal, V. R., & Subramanian, R. (2010). An empirical investigation of environmental performance and the market value of the firm. Journal of Operations Management, 28, 430–441. Jones, J. J. (1991). Earnings management during import relief investigations. Journal of Accounting Research, 29, 193–228. Jung, Y. J., & Lee, S. H. (2003). The legacy of the Byrd Amendment controversies: Rethinking the principle of good faith. Journal of World Trade, 37, 921–958. Knetter, M. M., & Prusa, T. J. (2003). Macroeconomic factors and antidumping filings: Evidence from four countries. Journal of International Economics, 61, 1–17. Konings, J., & Vandenbussche, H. (2005). Antidumping protection and markups of domestic firms. Journal of International Economics, 65, 151–165. Lee, S. H., & Baik, Y. S. (2010). Corporate lobbying in antidumping cases: Looking into the Continued Dumping and Subsidy Offset Act. Journal of Business Ethics, 96, 467–478.

105

Leidy, M. P. (1997). Macroeconomic conditions and pressures for protection under antidumping and countervailing duty laws: Empirical evidence from the United States. International Monetary Fund Staff Papers, 44, 132–144. Magnan, M., Nadeau, C., & Cormier, D. (1999). Earnings management during antidumping investigations: Analysis and implications. Canadian Journal of Administrative Sciences— Revue Canadienne Des Sciences De L Administration, 16, 149–162. Malhotra, N., & Malhotra, S. (2008). Liberalization and protection: Antidumping duties in the Indian pharmaceutical industry. Journal of Economic Policy Reform, 11, 115–122. Marsh, S. J. (1998). Creating barriers for foreign competitors: A study of the impact of anti-dumping actions on the performance of US firms. Strategic Management Journal, 19, 25–37. Nieberding, J. F. (1999). The effect of US antidumping law on firms' market power: An empirical test. Review of Industrial Organization, 14, 65–84. Niels, G., & Francois, J. (2006). Business cycles, the exchange rate, and demand for antidumping protection in Mexico. Review of Development Economics, 10, 388–399. Peng, M. W., Wang, D. Y. L., & Jiang, Y. (2008). An institution-based view of international business strategy: A focus on emerging economies. Journal of International Business Studies, 39, 920–936. Prusa, T. A. (2001). On the spread and impact of antidumping. Canadian Journal of Economics, 591–611. Reynolds, K. M. (2006). Subsidizing rent-seeking: Antidumping protection and the Byrd amendment. Journal of International Economics, 70, 490–502. Schuler, D. A., Rehbein, K., & Cramer, R. D. (2002). Pursuing strategic advantage through political means: A multivariate approach. Academy of Management Journal, 45, 659–672. Takacs, W. E. (1981). Pressures for protectionism: An empirical analysis. Economic Inquiry, 19, 687–693. Wang, Y. T. (2010). The profits of firms and countervailing duties in vertically related markets. Singapore Economic Review, 55, 749–756. Yang, L. (2013). Xinhua Insight: New PV stimulus package to offset EU barriers. China Daily. Zeng, K., & Liang, W. (2010). US antidumping actions against China: The impact of China's entry into the World Trade Organization. Review of International Political Economy, 17, 562–588.