Ripple Effects of The 2011 Japan Earthquake On International Stock Markets PDF

Research in International Business and Finance 41 (2017) 556–576
Contents lists available at ScienceDirect
Research in International Business and Finance

journal homepage: www.elsevier.com/locate/ribaf
Ripple effects of the 2011 Japan earthquake on international stock MARK

markets
⁎
Pourya Valizadeh, Berna Karali , Susana Ferreira
Department of Agricultural and Applied Economics, The University of Georgia, Athens, GA, 30602, USA
AR TI CLE I NF O AB S T R A CT
JEL classification: This paper provides a comprehensive analysis of the impacts of Japan’s 2011 earthquake on 19
D8 stock market sector returns in Japan and its trading partners both in the short and long run. Using
F36 an event study methodology, we find that the impact of this event was not limited to Japan or
F65 industries directly hit by the earthquake. Our short-run analysis indicates that all sector indices in
G14
Japan and many in its trading partners were affected by the earthquake. The direction of the
G15
Q54
impact on trading partners, however, was not the same for all sectors; while the earthquake
adversely affected the majority of the sectors analyzed, some sectors benefited. Further, we find
Keywords:
that the magnitude of the abnormal returns did not systematically vary across trading partners
Abnormal returns
according to their shares in Japan’s trade flow. The long-run analysis reveals how the
BHAR
Event study consequences of the earthquake unfolded beyond the event date.
Japan’s earthquake
Stock market
1. Introduction
The March 11, 2011 Tohoku-Oki earthquake and tsunami in Eastern Japan and the resulting Fukushima Daiichi nuclear accident
had a significant negative impact on the economy of Japan (Nanto et al., 2011). Direct damages were estimated at $211 billion
(Kajitani et al., 2013) making it one of the largest disasters on record (CRED, 2011). The earthquake hit the home of the automobile
and semiconductor manufacturing industries, destroying not only factories but also disrupting the supply of raw materials, leading to
a suspension of the workflow in many large-scale automobile manufacturers, such as Toyota, Nissan, and Honda (Khazai and Daniell,
2011). Japan’s manufacturing sector faced an additional challenge due to power shortages, resulting in disruption and temporary
closings of manufacturing plants, even those that were not directly damaged by the earthquake (Kachi and Takahashi, 2011).
In this age of just-in-time production processes, even a small disruption in the supply of a single component can cause turmoil in
the entire production line (Nanto et al., 2011). This is particularly important in high-tech industries such as automobiles,
telecommunications, and consumer electronics. Considering that Japan is a central hub for international supply chains (JETRO,
2011), interruptions in the Japanese industrial activity can have adverse impacts on global supply chains. Thus, the 2011 Japan’s
earthquake might have affected different industries in other countries either negatively due to supply chain distortions or positively
because of a potential rise in the market share for domestic firms as a result of Japan’s reduced competitiveness.
These possible impacts on both Japanese and global industries can be explored through an analysis of stock market behavior
around the date of the earthquake (i.e., an event study analysis) as stock prices of publicly traded companies indicate how investors
value firms’ current performance and future earnings potential, provided that stock markets are relatively efficient. Specifically, we
⁎
Corresponding author.
E-mail address: bkarali@uga.edu (B. Karali).
http://dx.doi.org/10.1016/j.ribaf.2017.05.002
Received 16 March 2017; Received in revised form 28 April 2017; Accepted 2 May 2017
Available online 11 May 2017
0275-5319/ © 2017 Elsevier B.V. All rights reserved.
P. Valizadeh et al. Research in International Business and Finance 41 (2017) 556–576
seek to answer the following questions. First, which specific sectors in the stock markets of Japan and its trading partners were
affected by the 2011 earthquake and how strong were the effects and in which direction? Second, were the stock markets of Japan’s
major trading partners more affected than those of minor trading partners?
Earlier studies examining the reaction of stock markets to natural disasters have focused on stock returns of companies aggregated
at the state level (e.g., Bourdeau-Brien and Kryzanowski, 2017), entire stock market indices (e.g., Worthington and Valadkhani, 2004;
Worthington, 2008; Ferreira and Karali, 2015), and single sector indices, such as insurance (e.g., Shelor et al., 1992; Aiuppa et al.,
1993; Lamb, 1995; Cagle, 1996; Angbazo and Narayanan, 1996; et al., 2002Yamori and Kobayashi, 2002; Krämer and Schich, 2008;
Wang and Kutan, 2013), and real estate (e.g., Shelor et al., 1990). We know of only one study, by Worthington and Valadkhani
(2005), that investigates the impacts of several natural disasters, industrial accidents, and terrorist attacks on different market sector
indices, such as energy, healthcare, financial, technology, telecommunication services, and utilities, of the Australian capital market.
Perhaps not surprisingly, results from this study indicate that the effects of these events depend on the business sector in question.
Another strand of the literature examines the existence of contagion in international financial markets following financial crises
and natural disasters. However, the empirical research has been more concentrated on financial crises, such as the Mexican Peso crisis
of 1994 (e.g., Ahlgren and Antell, 2010), the Asian crisis of 1997 (e.g., Rodriguez, 2007), the Russian bond market crash of 1998 (e.g.,
Ahlgren and Antell, 2010), and the financial crisis of 2007–2008 (e.g., Boubaker et al., 2015; De Angelis and Gardini, 2015;
Lehkonen, 2015), because stock markets have been the first to react in every financial crisis (Kontonikas et al., 2013; Chevapatralul
and Tee, 2014; Jayech, 2016). Applied to natural disasters, there are only a limited number of studies. For instance, Lee et al. (2007)
employ a sample of 26 international exchange rates and stock indices and find that although the foreign exchange markets of some
countries suffered from contagion in the aftermath of the 2004 Southeastern Asia tsunami, no single country’s stock market reacted.
On the contrary, Asongu (2012) finds that while the foreign exchange markets in 33 countries did not reveal contagion effects in the
immediate aftermath of the Japan’s 2011 earthquake, the stock markets of several countries (Taiwan, Bahrain, Saudi Arabia, and
South Africa) suffered from a contagion effect. Ferreira and Karali (2015), analyzing the impact of major earthquakes on the return
and volatility of aggregate stock market indices in 35 financial markets between 1994 and 2014, conclude that global financial
markets were resilient to earthquake shocks.
Evidently, the results from these earlier studies are not conclusive about the impacts of natural disasters (and specifically
earthquakes) on international stock markets. One possible explanation could be due to differences in the severity and the
geographical location of the disasters they consider. For instance, the Southeastern Asia tsunami occurred in the western part of
Indonesia. Although it also affected other countries, none of these countries are as important and influential as Japan in the world
economy. Therefore, disruptions in these economies probably have fewer consequences compared to disruptions in highly
industrialized economies such as Japan. Further, the analysis of the response of aggregated market indices, such as in Ferreira
and Karali (2015), cannot shed light on which specific sectors are driving the results, and may mask very negative abnormal returns
in specific sectors if these are compensated with positive returns in others.
Regarding the 2011 Japanese earthquake, except for Asongu (2012), the majority of earlier studies have investigated the
implications of the Fukushima accident for the stock returns of conventional and alternative utilities and nuclear energy both at
domestic and international levels (e.g., Kawashima and Takeda, 2012; Lopatta and Kaspereit, 2014; Ferstl et al., 2012; Betzer et al.,
2013; Mama and Bassen, 2013). We are aware of only one study, by Takao et al. (2013), that analyzes the impact of the earthquake
on other stock prices (specifically those of non-life insurance companies in Japan). However, the impact of the earthquake was not
restricted just to the insurance and energy sectors; automobile, semiconductor, electronics, food, and even tourism industries were
also affected (Nanto et al., 2011). These impacts could be transferred to industries in other countries, as well. For instance, the US
banned imports of certain types of food from several regions in Japan following the Fukushima incident; and the European Union,
China, and many other countries increased the surveillance of food products from Japan (Nanto et al., 2011). Thus, the food industry
in the US and other countries could be influenced. As another example, damages to Japanese automobile industry could have
consequences for companies in other countries that rely on Japan for critical components required in the production line.
This motivates us to provide the first comprehensive analysis of the impacts of the 2011 Japanese earthquake on all business
sectors in stock markets of Japan as well as some of its trading partners. Our analysis seeks to open the “black box” of aggregate
market indices to help policymakers and investors thoroughly understand the economic implications of natural disasters. Because
these implications are potentially different in different sectors and countries, a disaggregated analysis can assist in better preparing
for and responding to disasters. Our study contributes to the existing literature in several ways. First, we use sector-level indices
instead of aggregate stock market prices in contrast to previous studies. This increases the power of our study to detect reactions of
stock markets to natural disasters; by disaggregating the stock market’s broad price index, we can detect possible asymmetries in the
effects of the disasters on different sectors. If natural disasters affect different sectors in different directions then when these impacts
are aggregated they may offset each other and the entire stock market index may not react.1 Second, using a sample of Japan’s major
and minor trading partners, we analyze possible links between trade linkages and contagion effects across international stock
markets. We do this because international trade linkages are a channel through which country-specific crises can spread over
international financial markets (Forbes, 2002). Third, our approach allows us to compare the magnitudes of the impacts on different
1
It should be noted a similar aggregation issue might arise from using sector-level indices. While some of the companies included in a given sector index might
exhibit positive abnormal returns in their stocks, the others might have negative abnormal returns. These positive and negative impacts might offset each other, and
therefore a reaction in the entire sector index might not be observed. A commonly used alternative is to conduct the analysis at firm-level stock prices. However, a firm-
level analysis for all the sectors and all the countries is beyond the scope of this paper. Therefore, we take an intermediate approach between aggregate stock market
indices and firm-level prices and focus on sector-level indices to shed light on the dynamics across sectors and countries.
557
industries across countries. Lastly, we examine the long-run effects of the earthquake, whereas the existing literature mostly focuses
on the short-run effects.2
We find that the earthquake affected all sector indices in the stock markets of Japan and its trading partners both in the short and
long run. Our findings indicate that while all sectors in the Japanese stock market were adversely affected, some sectors in other
countries benefited from the earthquake. However, the magnitude of abnormal returns in the stock markets of trading partners was
small. While our results are consistent with Asongu (2012), they differ from the findings of Lee et al. (2007) and Ferreira and Karali
(2015), which used aggregate stock market indices. This contrast suggests that negative and positive impacts on different sectors
(found in our study) counterbalance one another and create a zero impact on aggregate indices (found in those earlier studies). We
also find that countries’ shares in Japan’s trade flow cannot explain the magnitude of the abnormal returns. Put differently, Japan’s
major trading partners were not necessarily more affected than its minor trading partners by the disruptions in Japan’s supply chain.
2. Methodology
2.1. Short-run analysis
The event study methodology of Fama et al. (1969) has become the standard approach for measuring stock price reactions to
company-related announcements or unexpected events. This methodology is based on the efficient market hypothesis (EMH) that
states financial markets are informationally efficient; that is, stock prices always incorporate and reflect all the relevant information
and immediately adjust to reflect the effects of unanticipated events (Fama, 1970; MacKinlay, 1997). Assessment of an event impact
requires a measure of abnormal returns (ARs), for which two different approaches are commonly used in the literature. In the first
method, ARs are computed as prediction errors from a benchmark model of normal returns. The market model in which company
stock returns are regressed on the returns of a market portfolio, such as S & P 500, is generally used as the benchmark. One
disadvantage of this approach is that if events occur over the same calendar period for firms that are in the same or related industries
then the assumption of independent ARs is violated, and therefore statistical inferences are invalid (Binder, 1985).
The second approach involves using dummy variables corresponding to the event periods to model ARs as regression coefficients.
In other words, in this method, a dummy variable taking the value of one during the event period (and zero otherwise) is added to the
market model. One advantage of this approach is that it can solve the problem of dependent ARs through estimating a seemingly
unrelated regression (SUR) model. Another advantage of this approach in the SUR framework is that it allows hypothesis testing
because it takes into account any heteroskedasticity across equations as well as any contemporaneous correlation among the
disturbances (Binder, 1985), making this approach suitable for our study.
We begin by parameterizing the ARs in the sector-level return equations by including event-day dummy variables in the market
model:
t1
Rit = αi + βi Rmt + ∑ γit dt + εit ,
t =t0 (1)
where Rit is the continuously compounded daily return on stock market price index for sector i (Pit), computed
asRit = 100 × (lnPit − lnPi, t −1), αi and βi are the market alpha and beta, respectively, Rmt is calculated in a similar manner to represent
daily return on the market portfolio price index (Pmt). Including the market return in the regression removes the portion of the sector’s
return that is related to variation in the overall stock market and this, in turn, reduces the variance of the AR. Thus, the ability of the
model to detect ARs increases (MacKinlay, 1997). We define t0 as the event day and t1 as the last day of the event window. The
variable dt is a dummy variable that takes the value of one if day t is in the event window [t0, t1] and zero otherwise. Therefore, the
parameter γit represents the AR on sector i’s index on day t during the event window. Finally, εit is the zero-mean disturbance term.
We extend the basic model presented in Eq. (1) in several ways. First, we include five lags of the dependent variable and market
return as regressors to account for serial correlation found in daily stock returns using the Breusch-Godfrey test (Breusch, 1978;
Godfrey, 1978) test. Second, by including weekday dummy variables (Wl) in the model, we control for any day-of-the-week effects.
Third, if market model parameters α and β shift in the post-earthquake period, the usual interpretation of the event dummy
coefficients as ARs will no longer hold (Binder, 1985; Mama and Bassen, 2013). To address this issue, we define a dummy variable
(Dt) taking the value of one during the post-event period (including the event day) and zero otherwise. We then include Dt as well as
its interaction with the market return (i.e., Rmt) to control for possible intercept and slope shifts, respectively. Therefore, our
empirical model is specified as follows:
t1 4 5 5
Rit = αi + αiD Dt + βi Rmt + βiD Dt Rmt + ∑ γit dt + ∑ δil Wlt + ∑ θiτ Ri, t − τ + ∑ ϑi, τ Rm, t − τ + εit .
t =t0 l =1 τ =1 τ =1 (2)
The next step is to define the event and estimation windows. The event window is the period over which the impact of the
earthquake is measured. It is typical to define the length of the event window to be longer than one day even if the event being
studied happened only in one day (MacKinlay, 1997). Typically, the event day is denoted by t0 = 0, which in this study refers to
2
To the best of our knowledge, there is only one study, by Kawashima and Takeda (2012), which investigates the long-term impact of the 2011 earthquake on stock
prices of electric power utilities in Japan.
558
Friday, March 11, 2011. For the short-run analysis, we define a 3-day event window such that [t0 = 0, t1 = 2]. Defining a narrow
event window reduces the impacts of confounding events and increases the validity and reliability of the event study (De Jong and
Naumovska, 2015). However, to capture any delayed reaction in the stock markets in the short-run analysis, we also examine the
behavior of post-event ARs by defining two relatively longer event windows of six (t1 = 5) and 16 days (t1 = 15). We estimate Eq. (2)
for a given country in a system of equations via SUR, each equation representing one sector.3 This allows us to conduct hypothesis
tests regarding the impact of the earthquake on different business sectors in a given country. Moreover, we estimate Eq. (2) for a
given sector in a system of equations, each equation corresponding to one of the countries included in our study, to perform
hypothesis tests regarding the effects of the earthquake on a given sector across countries.
To measure the overall reaction in stock markets following the earthquake, ARs obtained after SUR estimation are aggregated to
obtain cumulative abnormal returns (CARs) over the event window for each sector i:
t1
CARi [t0, t1] = ∑ γît.
t =t0 (3)
Under the null hypothesis that the event has no effect on stock market returns, ARs and CARs will be asymptotically normally
distributed as the length of the estimation window increases (MacKinlay, 1997). Therefore, Student’s t-test can be used to test the
statistical significance of the CARs. The sign of the estimated CAR for each sector determines whether it has been negatively or
positively affected by the earthquake.4
It is customary to start the estimation window (i.e., the sample period used to estimate model parameters) about one year before
the event date and to include event-window observations in the regression-based event study approach. Earlier studies also
recommend using post-event window days as part of the estimation window. This is especially important when the market model is
modified to allow for parameters to change after the event (Binder, 1998). Moreover, Binder (1985) argues that with only 250 daily
return observations asymptotic test statistics are biased. Thus, the estimation window in our study starts at 250 days before the event
(i.e., on March 26, 2010) and ends at 104 days after the event (i.e., on August 3, 2011). We select this period to avoid possible
contamination of our results by the stock market crash of August 2011.5
2.2. Long-run analysis
Event-study methods discussed so far are commonly used for the short-run analysis. To examine the stock market impact of the
2011 earthquake in the long run, we apply the buy-and-hold abnormal return (BHAR) approach. Although there is no consensus on
the best way to measure the ARs in the long run, the BHAR method has been advocated by Barber and Lyon (1997) and Lyon et al.
(1999) for its ability to match investors’ behavior. The buy-and-hold approach is a passive investment strategy in which a stock is
purchased at the market closing price on the day before the event and held for a long period (T days). Investors employing this
investment strategy are generally not concerned with short-term price movements; institutional investors, for instance, fall in this
category (Carpentier and Suret, 2015). The BHAR for a firm is then defined as the difference between its realized buy-and-hold return
after the event and the normal (expected) buy-and-hold return in the absence of the event. We again choose T = 104 days and
calculate BHARs for all sectors using the returns from day t0 to T. BHAR for sector i is given by
⎡ T ⎤ ⎡ T ⎤
BHARiT = ⎢ ∏ (1 + Rit ) − 1⎥ − ⎢ ∏ (1 + Rmt ) − 1⎥
⎢⎣ t = t ⎥⎦ ⎢⎣ t = t ⎥⎦
0 0 (4)
where Rit and Rmt are defined as before. The average BHAR over the period [t0, T] for sector i is calculated as:
T
1
BHARi [t0, T ] = ∑ BHARit .
(T − t 0 ) t =t0 (5)
Significance tests of the average BHARs are then performed to test the null hypothesis that the earthquake had no impact in the
long run. Since the long-run BHARs are found to be positively skewed (Barber and Lyon, 1997; Kothari and Warner, 1997) the
statistical significance tests are based on the Johnson skewness-adjusted t-test (Johnson, 1978).
3
The efficiency gain from using an SUR framework instead of an ordinary least squares (OLS) method depends on how different the regressors are from one sector’s
equation to the others. For a given country, the regressors across industries in our model differ only by the lagged dependent variables, and therefore the efficiency
gain might be small.
4
Aggregating ARs of each day during the event window helps us to deal with the lack of synchronism in stock market trading hours as well. There is a difference of
12h between American and Asian countries and 5 or 6h between European and Asian countries. At the time of the earthquake, financial markets in Japan and China
were about to close whereas the stock markets in the US and European countries were still not open. Therefore, the market reaction on the event day was not likely
complete for Japan and China as these markets were open only about 15min and one hour, respectively, after the earthquake, and therefore probably did not have
enough time to completely incorporate the new information (our analysis of ARs shows that the earthquake imposed some impacts on market sectors in Japan and
China on the event day). Thus, CARs over the event window instead of the ARs on the event day are more appropriate to make comparisons across countries in
different time zones.
5
The August 2011 stock market crash was a sharp drop in stock prices in markets across the US, Europe, and Asia. Underlying reasons include fears of contagion of
the European sovereign debt crisis to Italy and Spain, concerns over the slow economic growth of the US, and France’s AAA bond credit ratings (Bremer and
Dmitracova, 2011).
559
Table 1
Country Shares (%) in Japan’s Trade Flow in 2010.
Source: World Bank’s World Integrated Trade Solution (WITS).
Japan’s Exports Japan’s Imports
China 19.41 22.07

USA 15.63 9.96
Germany 2.64 2.78
France 0.87 1.48
Italy 0.73 0.98
Spain 0.42 0.38
3. Data
Our sample consists of Japan and six other countries comprising its major trading partners (China and the US), and minor trading
partners (Germany, France, Italy, and Spain). These countries are chosen based on their shares in Japan’s trade flow in 2010 as well as
the availability of their financial market data. For instance, although South Korea is one of Japan’s major trading partners (with 8.1%
and 4.1% shares in Japan’s exports and imports, respectively), due to the lack of stock market data on some of the sectors included in
our comprehensive analysis, it is excluded from our sample. Table 1 reports the shares in Japan’s exports and imports of goods and
services corresponding to each of the six countries in 2010.
Thomson Reuters’ Datastream is used as our primary data source. Datastream breaks down the stock market data for a given
country into six levels.6 Level one is the aggregate stock market index (i.e., the broadest local index); level two provides stock indices
for 10 industries, and level three through level six subdivide level two industries into sector classifications in increasing detail. We
focus on level three “supersector” classification that provides adjusted price indices (Datastream’s data type PI) for 19 different
supersectors in each country. Although level four offers a more detailed classification, for several supersectors, such as automobiles
and parts, banks, chemicals, construction and materials, travel and leisure, and media, its indices are identical to level three. For the
remaining sectors, levels three and four are also similar and we expect analogous results from these two sector classifications.7
Consequently, we use level three supersector indices as they provide enough diversity to perform a sector-level analysis. For the ease
of discussion, we refer to level three supersectors as sectors hereafter. Table 2 lists these 19 sectors along with their major subsectors.
In multi-country event studies either broad local return indices such as Nikkei 225 for Japan (Campbell et al., 2010) or global
return indices such as EAFE index8 (Park, 2004) can be used as a benchmark. Datastream’s level one stock market index is available
for all countries and represents the broadest local index. However, instead of using this local index we use the adjusted price index
(Datastream’s data type PI) of the broadest global index available, the world index (Datastream’s mnemonics TOTMKWD),9 as the
benchmark return in our analysis for several reasons. First, financial markets are assumed to be globally integrated (Bekaert and
Harvey, 1995; Karolyi and Stulz, 1996; Chan et al., 1997). Second, broad local indices, especially for Japan, were highly affected by
the earthquake. Fig. 1 compares the impact of the earthquake on two selected sector returns (automobiles and utilities) and the
broadest market return for Japan from 40 days before the event day (i.e., day zero) to 40 days after. It is obvious from this figure that
the market index is highly affected by the earthquake, and thus if used as the benchmark in our analysis it might obscure the impact
of the earthquake on the sectoral indices. Third, since we compare the impact of the earthquake across different countries, using the
same benchmark return for all provides a better assessment. Both the world index and the sector return indices in each selected
country are calculated using price indices in US dollars. It should be noted that changes in currency values are one potential channel
through which the March 2011 earthquake affected the relation between the return series of different countries. Because we are
interested in the overall effect on returns, the fluctuations in foreign exchange rates do not need to be controlled for in our model.
When the BHAR analysis is performed at the firm level, different approaches, such as matching by firm, matching by portfolio, or
matching by industry, can be used to choose the benchmark return. However, since our analysis is at the sector level, the usual
approaches are not applicable, and therefore we again use the world index as the benchmark return supported by the global
integration of financial markets. This also facilitates the comparison between short- and long-run analyses.
6
An advantage of using Datastream sector indices as opposed to self-constructing sector indices using firm-level data is that they are calculated using the same
methods for all the countries included in the analysis.
7
For example, the level three category of “food and beverage” is divided into level four subsectors as “beverages” and “food producers.” A more detailed analysis of
each sector could be performed using Datastream’s level six classification. However, the availability of data greatly varies across countries in that case. From 114
subsectors defined in level six, the total number of sectors in each country with available price data is as follows: 95 for Japan, 68 for China, 108 for US, 75 for
Germany, 83 for France, 66 for Italy, and 51 for Spain.
8
The EAFE (Europe, Australia, and Far East) index measures the stock market performance of developed markets outside of the US and Canada.
9
The world index contains 66 countries and 6982 active stocks.
560
Table 2
Datastream’s Level Three Sector Classification of Stock Markets.
Source: Thomson Reuters’ Datastream.
Sector Major subsectors
Automobiles and Parts Automobiles

Auto Parts
Tires
Banks Banks
Basic Resources Forestry and Paper
Industrial Metals and Mining
Mining
Chemicals Commodity Chemicals
Specialty Chemicals
Construction and Materials Building Materials and Fixtures
Heavy Construction
Financial Services Consumer Finance
Specialty Finance
Investment Services
Asset Managers
Mortgage Finance
Food and Beverage Beverages
Food Producers
Healthcare Healthcare Equipment and Services
Pharmaceuticals and Biotechnology
Industrial Goods and Services Aerospace and Defense
General Industrials
Electronic and Electric Equipment
Industrial Engineering
Industrial Transportation
Support Services
Insurance Nonlife insurance
Life Insurance
Media Media
Oil and Gas Oil and Gas Producers
Oil Equipment and Services
Alternative Energy
Personal and Household Goods Household Goods and Home Construction
Leisure Goods
Personal Goods
Tobacco
Real Estate Real Estate Investment and Services
Real Estate Investment Trusts
Retail Food and Drug Retailers
General Retailers
Technology Software and Computer Services
Technology Hardware and Equipment
Telecommunications Fixed Line Telecommunications
Mobile Telecommunications
Travel and Leisure Airlines
Hotels
Travel and Tourism
Utilities Conventional Electricity
Alternative Energy
Gas, Water and Multiutilities
4. Results
4.1. Short-run analysis
The estimated CARs from Eq. (3) for three different event windows are presented in Panels A to C of Table 3 (and also graphically
in the Appendix A Fig. A1–A7).10 Panel A shows that all sectors of the Japanese stock market were affected negatively over the 3-day
event window. Over the 6-day event window (Panel B), the construction sector reveals a small positive impact which could be due to
an expected increase in demand for reconstruction after the earthquake. Although the effect on this sector is again negative over the
16-day window (Panel C), it is statistically insignificant suggesting that the increase in demand could offset the adverse effects of the
earthquake to some extent.
10
Full results from estimating Equation (2) for each country are available from the authors upon request.
561
Fig. 1. Total market and selected sector indices for Japanese stock market.
After the earthquake, the closure of refineries and other industrial facilities in Japan, one of the world’s largest oil consumers, led
to a decrease in crude oil prices (Gibbons, 2011). This situation in Japan could have a positive effect on refined products which, in
turn, could counteract the negative effects of the earthquake on oil and gas sector to some degree. This is what we observe when
making a comparison between CARs for this sector.
In addition, comparing the CARs across different event windows reveals that while for some sectors, such as basic resources,
chemicals, healthcare, and industrial goods, the negative market reaction was concentrated during the first two days after the
earthquake, for other sectors, such as automobiles, banks, financial services, insurance, media, and utilities, the negative impact
increased as uncertainty about the damages was being resolved (i.e., 16-day CARs for these sectors are more negative).11 Further, if
we compare 3- and 16-day CARs across sectors, we see that the earthquake imposed the largest and smallest negative impact on the
utilities and the construction and materials sectors, respectively. Differences are statistically significant at the 1% significance level.
The very large negative effect on the utilities sector (with negative CAR of about a third 16 days into the crisis) is likely a consequence
of the Fukushima nuclear accident, one of the most disastrous aftermaths of the earthquake.
Several sectors in the Chinese stock market were also affected by the earthquake. However, the CARs are one order of magnitude
smaller than Japan’s. Moreover, Panel A of Table 3 shows that while some sectors (construction and materials, food and beverage,
insurance, real estate, retail, and utilities) were adversely affected by the earthquake, other sectors (healthcare and telecommunica-
tions) exhibited positive CARs. A similar pattern is observed in Panels B and C. Comparing the CARs over longer event windows,
especially the 16-day event window, to the 3-day CARs, we can see that for some sectors, such as chemicals, healthcare, real estate,
and telecommunications, the sign of the impact changes across event windows. This hints at the importance of additional information
that became available to the markets as the crisis unfolded and whose effects are captured in a longer-run analysis.
Results for the US show that in the immediate short-run (i.e., 3-day event window) the majority of the sectors benefited from the
earthquake and only three sectors (food and beverage, telecommunications, and utilities) showed statistically significant negative,
albeit small, CARs. Among the sectors with positive CARs, we find the automobiles, construction, and oil and gas sectors. While these
results for construction and oil sectors are expected, the positive reaction of automobiles sector might be more surprising as the US
relies on Japan and other countries12 for auto parts, such as semiconductors, fuel injection modules, air bags, and tires; and disruption
in the Japanese supply chain should have affected the US automobile industry adversely. However, these disruptions were mostly
related to supplies of made-in-Japan models like Toyota Prius and Honda Fit and to Japanese parts used in cars assembled in the US
(Edgerton, 2012). Moreover, the shortage of Japanese cars could create opportunities for American car makers to benefit from their
Japanese rivals’ troubles (The Economist, 2011). Therefore, the positive CAR for the automobiles and parts in the US could be an
indication of expected increase in demand for American car makers (Edgerton, 2012). However, we should notice that this positive
impact did not persist, and as can be seen, over the 16-day event window it is not statistically significant.
Panel B shows that over the 6-day event window more sectors in the US display significant negative impacts and the magnitudes
of these adverse impacts are larger. Moving to Panel C, we see that while some of the negative effects observed over shorter event
windows (e.g., technology, travel and leisure, and utilities) seem to be more prolonged, the adverse impact on other sectors, such as
personal goods, and telecommunications is short-term. Conversely, some of the sectors that showed positive CARs, such as banks,
media, and technology, during the first week after the earthquake, revealed negative impacts 16 days into the disaster. Although not
all the negative effects are statistically significant, they suggest that a longer-run analysis could be informative, particularly when
11
The immediate reactions in basic resources, healthcare, and industrial goods sectoral indices are self-explanatory in that the physical destruction of infrastructure
and large human costs were observable right after the earthquake. However, reactions in other sectors, such as insurance, banks, and financial services, might be
delayed as assessing the earthquake’s impact on those sectors requires a more detailed evaluation of the number of claims and specifics of the policies in the case of
insurance, the composition of the investment portfolios in the case of banking and financial institutions.
12
US imports about one-fourth of all auto parts required in the production lines of automobiles from Asia and Europe (Canis, 2011).
562
Table 3
Cumulative Abnormal Returns (CARs).
Sector Japan China US Germany France Italy Spain
Panel A: 3-day CARs

Automobiles −13.91*** 1.20 4.82*** 0.50 1.15* 3.52*** 2.69***
(0.000) (0.177) (0.000) (0.365) (0.070) (0.000) (0.007)
Banks −15.12*** −0.49 1.69*** 2.82*** 7.38*** 8.79*** 9.78***
(0.000) (0.358) (0.010) (0.000) (0.000) (0.000) (0.000)
Basic Resources −18.20*** −0.32 6.78*** 0.06 2.65*** 3.77*** 3.67***
(0.000) (0.712) (0.000) (0.888) (0.000) (0.000) (0.000)
Chemicals −19.25*** 1.07 3.25*** 1.45*** −0.76 3.70*** 0.79
(0.000) (0.218) (0.000) (0.004) (0.144) (0.000) (0.481)
Construction −6.17*** −1.73** 4.37*** 2.92*** 3.50*** 4.01*** 6.52***
(0.000) (0.012) (0.000) (0.000) (0.000) (0.000) (0.000)
Financial services −17.45*** −1.02 1.69*** −0.79 2.94*** 1.91*** 2.69***
(0.000) (0.254) (0.000) (0.173) (0.000) (0.002) (0.000)
Food and Beverage −15.25*** −1.43** −0.74** −4.29*** −0.86* 3.47*** 3.31***
(0.000) (0.033) (0.013) (0.000) (0.093) (0.000) (0.000)
Healthcare −12.08*** 2.56*** 0.55* −1.14*** −1.67*** 2.85*** 1.36**
(0.000) (0.001) (0.072) (0.005) (0.003) (0.000) (0.020)
Industrial Goods −15.04*** −0.87 1.99*** 1.43*** 1.54*** 2.70*** 2.69***
(0.000) (0.207) (0.000) (0.000) (0.000) (0.000) (0.000)
Insurance −14.92*** −1.94*** 0.27 −1.52*** 0.59 3.30*** 3.80***
(0.000) (0.010) (0.452) (0.005) (0.509) (0.000) (0.000)
Media −19.39*** −1.71 1.32*** 1.46* 1.18** 4.69*** 7.55***
(0.000) (0.108) (0.001) (0.081) (0.039) (0.000) (0.000)
Oil and Gas −14.02*** −0.84 4.09*** 24.45*** 1.83*** 3.04*** 2.64***
(0.000) (0.111) (0.000) (0.000) (0.001) (0.000) (0.000)
Personal Goods −13.07*** −0.14 −0.41 −1.07** −1.11** 0.31 0.42
(0.000) (0.865) (0.173) (0.014) (0.017) (0.757) (0.662)
Real Estate −17.16*** −2.47*** 1.77*** −0.79 1.59*** 3.10*** 0.24
(0.000) (0.003) (0.000) (0.111) (0.001) (0.000) (0.947)
Retail −18.15*** −2.26*** 0.75* −0.17 0.36 −0.49 2.45***
(0.000) (0.001) (0.069) (0.786) (0.487) (0.439) (0.001)
Technology −13.36*** −0.08 1.61*** −1.35*** 3.77*** 4.09*** 0.24
(0.000) (0.921) (0.000) (0.007) (0.000) (0.000) (0.709)
Telecommunications −12.83*** 2.23*** −0.85** 1.20* 1.15 2.83*** 1.00
(0.000) (0.004) (0.033) (0.085) (0.101) (0.010) (0.144)
Travel and Leisure −16.93*** −1.14 −0.99* −0.20 1.88*** 5.53*** 3.80***
(0.000) (0.284) (0.065) (0.779) (0.000) (0.000) (0.000)
Utilities −20.90*** −1.37* −1.31*** −1.78*** 0.88 3.57*** 4.35***
(0.000) (0.065) (0.000) (0.008) (0.155) (0.000) (0.000)
Panel B: 6-day CARs
Automobiles −9.44*** 2.11 1.76* −0.63 0.83 2.41* 4.16***
(0.000) (0.107) (0.059) (0.470) (0.380) (0.063) (0.005)
Banks −9.84*** −1.42* 1.77* 0.95 5.95*** 4.95*** 6.29***
(0.000) (0.088) (0.068) (0.359) (0.000) (0.007) (0.000)
Basic Resources −7.96*** 2.42** 5.21*** 2.41*** 4.14*** 3.71*** 5.20***
(0.000) (0.050) (0.000) (0.000) (0.000) (0.002) (0.000)
Chemicals −8.40*** 2.41* 1.83** 0.91 0.04 9.56*** 5.20***
(0.000) (0.059) (0.014) (0.233) (0.957) (0.000) (0.009)
Construction 1.66* −1.00 2.71*** 4.83*** 2.99*** 1.57 5.41***
(0.068) (0.329) (0.000) (0.000) (0.001) (0.185) (0.000)
Financial services −8.32*** −0.37 0.79 0.48 4.20*** 1.12 2.36**
(0.000) (0.797) (0.273) (0.580) (0.000) (0.214) (0.015)
Food and Beverage −9.71*** −0.50 −1.89*** −2.05** −1.40* 8.17*** 3.99***
(0.000) (0.633) (0.000) (0.013) (0.080) (0.000) (0.000)
Healthcare −5.38*** 0.47 −0.17 −2.06*** −2.11*** 2.51*** 1.25
(0.000) (0.724) (0.721) (0.002) (0.010) (0.004) (0.175)
Industrial Goods −6.26*** −0.56 0.75 1.72*** 2.31*** 2.69*** 2.68***
(0.000) (0.581) (0.186) (0.008) (0.000) (0.001) (0.005)
Insurance −13.19*** −2.64** −1.00* −1.58** 0.93 −0.12 1.22
(0.000) (0.018) (0.088) (0.042) (0.499) (0.924) (0.389)
Media −15.01*** −1.07 −0.80 7.21*** 1.04 2.42** 8.30***
(0.000) (0.519) (0.194) (0.000) (0.186) (0.050) (0.000)
Oil and Gas −0.02 0.10 3.31*** 17.64*** 1.80*** 1.81** 3.59***
(0.993) (0.902) (0.000) (0.000) (0.010) (0.050) (0.000)
Personal Goods −5.74*** −0.28 −2.13*** −3.23*** −2.93*** −0.17 0.77
(0.000) (0.816) (0.000) (0.000) (0.000) (0.891) (0.527)
Real Estate −7.02*** −2.12 0.44 1.44** 2.77*** 3.77*** 1.82
(0.000) (0.112) (0.590) (0.042) (0.000) (0.001) (0.689)
(continued on next page)
563
Table 3 (continued)
Retail −13.19*** −3.21*** −0.90 2.38*** 0.52 1.14 2.40**

(0.000) (0.003) (0.131) (0.004) (0.529) (0.328) (0.032)
Technology −9.58*** −0.97 −1.06* −0.31 1.77** 6.37*** 1.87*
(0.000) (0.453) (0.085) (0.680) (0.025) (0.000) (0.082)
Telecommunications −8.38*** 0.21 0.06 −0.52 0.72 −0.03 0.12
(0.000) (0.855) (0.921) (0.627) (0.471) (0.985) (0.922)
Travel and Leisure −8.57*** −0.60 −4.06*** −2.16** 1.28 4.28*** 6.72***
(0.000) (0.707) (0.000) (0.024) (0.103) (0.008) (0.000)
Utilities −13.97*** 0.88 −3.16*** −0.53 −0.06 2.58*** 2.56**
(0.000) (0.368) (0.000) (0.606) (0.949) (0.006) (0.014)
Panel C: 16-day CARs
Automobiles −17.91*** 0.48 2.29 −0.79 −3.74** −1.32 12.25***
(0.000) (0.861) (0.116) (0.647) (0.038) (0.571) (0.000)
Banks −16.26*** 4.44*** −0.74 −2.66 2.12 −6.62 2.74
(0.000) (0.009) (0.663) (0.254) (0.436) (0.125) (0.311)
Basic Resources −9.13*** 4.05* 7.85*** 0.14 5.94*** 9.66*** 4.66***
(0.000) (0.094) (0.000) (0.903) (0.000) (0.000) (0.000)
Chemicals −9.79*** −2.25 2.94** 1.06 −0.68 4.88** 1.91
(0.000) (0.398) (0.018) (0.499) (0.671) (0.041) (0.572)
Construction −1.97 0.44 5.79*** 4.87*** 2.95** −1.38 3.50*
(0.262) (0.846) (0.000) (0.001) (0.048) (0.510) (0.082)
Financial services −18.93*** −3.10 0.76 −4.36*** 4.44*** 1.62 −1.35
(0.000) (0.267) (0.549) (0.001) (0.000) (0.328) (0.478)
Food and Beverage −15.19*** −7.95*** 0.34 −4.00*** −0.05 1.63 6.00***
(0.000) (0.000) (0.660) (0.003) (0.976) (0.424) (0.000)
Healthcare −9.21*** −6.29*** 0.70 −4.72*** −3.41** −0.23 −2.81
(0.000) (0.005) (0.450) (0.000) (0.016) (0.878) (0.128)
Industrial Goods −8.67*** −2.59 2.83*** 4.06*** 3.11*** 1.85 3.42**
(0.000) (0.226) (0.006) (0.002) (0.004) (0.260) (0.047)
Insurance −22.82*** −0.11 −0.28 −3.37** −0.69 −3.91 1.43
(0.000) (0.962) (0.764) (0.017) (0.761) (0.136) (0.615)
Media −22.26*** −0.25 −0.84 4.10** 0.70 1.22 0.23
(0.000) (0.925) (0.373) (0.018) (0.609) (0.586) (0.924)
Oil and Gas 3.71 1.44 4.94*** 21.28*** 2.09* 0.49 2.43
(0.122) (0.315) (0.000) (0.000) (0.065) (0.711) (0.180)
Personal Goods −10.53*** −2.02 −1.05 −4.93*** −5.68*** −1.26 0.18
(0.000) (0.402) (0.242) (0.000) (0.000) (0.473) (0.932)
Real Estate −16.55*** 4.22* −0.37 −5.62*** 1.01 1.11 6.75
(0.000) (0.093) (0.762) (0.000) (0.420) (0.610) (0.275)
Retail −17.07*** −7.05*** −0.01 −2.96* −0.11 −0.79 8.39***
(0.000) (0.000) (0.993) (0.069) (0.937) (0.654) (0.000)
Technology −14.86*** −6.48*** −1.99* −0.55 2.47 7.30*** 2.77
(0.000) (0.008) (0.095) (0.776) (0.119) (0.001) (0.230)
Telecommunications −11.76*** −2.91 4.74*** 7.13*** 1.03 −3.92* −0.32
(0.000) (0.262) (0.000) (0.002) (0.540) (0.099) (0.881)
Travel and Leisure −15.75*** −3.41 −2.43* −2.72* −1.20 1.45 0.08
(0.000) (0.205) (0.054) (0.093) (0.335) (0.577) (0.973)
Utilities −32.57*** 1.08 −1.67* −3.43* −0.53 3.66** 2.49
(0.000) (0.599) (0.052) (0.072) (0.733) (0.029) (0.204)
Notes: This table presents cumulative abnormal returns (and their p-values in parentheses) calculated using the abnormal return estimates from a seemingly unrelated
regressions model with 19 equations, each representing one sector. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.
dealing with events whose consequences unfold beyond the event date.
Stock markets of Japan’s minor trading partners were also hit by the earthquake and varying effects across sectors are observed as
in the case with major trading partners. That is, while some sectors benefited from the earthquake, others were either negatively
affected or not affected at all.
Among European countries, Germany seems to be most adversely affected. This result is expected because Germany is the most
export-oriented country in Europe, and thus is more involved in global supply chains (Behravesh, 2015). It has larger shares in
Japan’s exports and imports than France, Italy, and Spain (see, Table 1). Accordingly, we can see from Panel A of Table 3 that several
sectors in Germany showed negative CARs. Among other European countries, some sectors of the French stock market, such as food
and beverage, healthcare, and personal goods, also displayed small negative impacts over the 3-day event window, whereas almost
all stock market sectors of Italy and Spain benefited from the event. Panel B displays a similar pattern and we can see that negative
impacts on some sectors in Germany and France were more pronounced compared to the shorter event window. Finally, Panel C
shows that not only more sectors in Germany and France were negatively affected, some sectors in Italy and Spain also revealed
negative responses over the 16-day event window, suggesting again that a longer-run analysis would be more enlightening.
564
The construction sector in all European countries reacted positively to the earthquake shock at least over the 3- and 6-day event
windows. Likewise, the CARs in the oil and gas sector were positive for these countries with Germany experiencing a larger gain. This
might indicate that the Germany’s stock market correctly anticipated the enormous backlash of the Fukushima nuclear accident on
nuclear power in Germany.13 Decreasing oil prices following a decline in Japan’s oil imports is expected to have positive impacts on
oil-importing countries. Our findings for this sector are consistent with this expectation. However, comparing the magnitudes of the
CARs between Germany and US as one of the largest oil importers in the world might indicate that not all the reaction of Germany’s
oil and gas sector is because of the decrease in oil prices. Indeed, the change in Germany’s nuclear energy policy might have also
contributed to this effect. Although the German government made the official announcement to phase out nuclear power plants
several weeks after the earthquake, the future of nuclear energy was unpromising even before the earthquake (see, Schneider et al.,
2011).
Overall, the short-run analysis indicates that the 2011 Japanese earthquake had implications for almost all sectors in the stock
markets of Japan’s trading partners. However, not surprisingly, the effects on these countries were smaller than impacts on Japan.
One reason could be that, at the time, Japan was going through a period of slow economic growth, which might have dampened
spillover effects for the world economy (Behravesh, 2015). Results also suggest that the earthquake’s impacts were not restricted to
the industries directly damaged by the earthquake in Japan. In addition, we find that business sectors across countries reacted
differently to the earthquake which could be explained by the characteristics of each country’s economy, such as their reliance on
Japanese exports, the size of their economies, and the competitiveness of their domestic industries. For some sectors, however, the
short-run analysis is noisy, displaying both positive and negative CARs. In longer horizons, the release of newer and more precise
information about the earthquake damages to the industrial supply chain, nuclear power plants, and trade flow could reveal a clearer
pattern. Thus, regarding the direction of the earthquake impact on those sectors a longer-run analysis might be more conclusive if
there are no other systemic confounding events.
To investigate the existence of a possible link between the magnitude of the earthquake impact on countries and their share in
Japan’s trade flow in the short run, we need to test the equality of CARs across trading partners. For each sector, we estimate a system
of seven equations, each representing one of the countries studied. The model is specified as in Eq. (2), but the returns for a given
sector are pooled together across seven countries. Accordingly, we can test the equality of the CARs between the two major trading
partners (China and US) and between each major and minor trading partner separately for the 3- and 6-day event windows (estimated
CARs from this system of equations are reported in the Appendix A Table A1).
We present the p-values for the equality tests of CARs in Table 4. The interpretation of the results is twofold. First, as can be seen,
the majority of the p-values indicate that the null hypothesis of the equality of CARs is rejected, implying that the earthquake effect is
country specific. However, comparing the magnitudes of CARs across these statistically different CARs does not reveal a clear
relationship between the degree of impact on a country’s stock market sectors and its share in Japan’s trade flow (as examples, see
Fig. A8 in the Appendix). This might be because we are looking at aggregate trade shares rather than trade shares by sector. In other
words, a more clear pattern in the results might be revealed with more detailed information on the sectoral composition of trade
flows between Japan and its partners.14
Second, for some sectors p-values indicate that the null hypotheses of equality of CARs between trading partners cannot be
rejected; that is, the earthquake impacted these countries equally. For example, the CARs for travel and leisure, and utilities sectors
were no different from each other in China, US, and Germany over the 3-day event window. Overall, these results suggest that there is
no direct link between a country’s share from Japan’s trade flow and the degree of the earthquake effect on its stock market sectors.
Put differently, a major trading partner is not necessarily more affected than a minor one.
4.2. Long-run analysis
Since the severity of disruptions in the supply chain due to the earthquake and the nuclear accident may be revealed with a delay,
stock market sectors, particularly those related to Japan’s exports and imports, might have a sluggish reaction. Therefore, we further
investigate the longer-run behavior of sector indices in the stock markets of Japan and its trading partners. We note, however, that
long-run analysis results should be interpreted cautiously as other confounding events that are partially or entirely unrelated to the
2011 earthquake might have affected sector-level ARs (Ferstl et al., 2012).
Table 5 presents estimated average BHARs using Equation (5) over the period [0,104] along with their p-values in parentheses.
These results show that, unlike the short-run analysis, all sectors in the Chinese stock market are negatively affected in the longer
horizon. This finding is expected because, at the time, China was Japan’s largest export market and a decrease in Japan’s exports
could be disruptive for China’s business sectors. However, some of the US and European sectors gained in the long run, as well. This is
most likely because of the new opportunities created by a reduction in Japan’s competitive pressure and a heightened perception of
the risks of nuclear energy. Examples include automobiles in Germany, France, Italy, and Spain; industrial goods and services, oil and
gas, and telecommunications in the US and Germany.
Similar to the short run, the longer-run analysis confirms that all sectors of the Japanese stock market and almost all sectors in
13
Although on October 28, 2010 the lifespan of nuclear power plants in Germany was extended (Mama and Bassen, 2013), in light of the Fukushima disaster,
Germany’s government decided to phase out all nuclear reactors by 2022 at the latest, starting in 2015 (Schneider, 2011). In 2011, nuclear energy accounted for almost
20% of Germany’s electricity supply.
14
Unfortunately, because the sectoral classification for trade data differs from Datastream’s, we were unable to perform this analysis.
565
Table 4
Equality Tests of Cumulative Abnormal Returns across Trading Partners.
Sector China vs. US vs.
US Germany France Italy Spain Germany France Italy Spain
Panel A: 3-day event window

Automobiles 0.001*** 0.546 0.917 0.064* 0.334 0.000*** 0.000*** 0.261 0.030**
Banks 0.008*** 0.001*** 0.000*** 0.000*** 0.000*** 0.370 0.000*** 0.000*** 0.000***
Basic Resources 0.000*** 0.831 0.045** 0.010*** 0.009*** 0.000*** 0.000*** 0.001*** 0.000***
Chemicals 0.015** 0.805 0.064* 0.029** 0.743 0.007*** 0.000*** 0.986 0.018**
Construction 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.036** 0.158 0.552 0.078*
Financial services 0.008*** 0.823 0.000*** 0.008*** 0.002*** 0.000*** 0.080* 0.750 0.372
Food and Beverage 0.440 0.003*** 0.551 0.000*** 0.000*** 0.000*** 0.930 0.000*** 0.000***
Healthcare 0.090* 0.001*** 0.000*** 0.526 0.476 0.001*** 0.001*** 0.001*** 0.268
Industrial Goods 0.002*** 0.028** 0.017** 0.001*** 0.003*** 0.223 0.388 0.416 0.522
Insurance 0.012** 0.740 0.051* 0.000*** 0.000*** 0.002*** 0.759 0.000*** 0.000***
Media 0.012** 0.023** 0.034** 0.000*** 0.000*** 0.938 0.862 0.002*** 0.000***
Oil and Gas 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.001*** 0.112 0.013**
Personal Goods 0.609 0.256 0.278 0.278 0.780 0.257 0.278 0.243 0.432
Real Estate 0.000*** 0.116 0.000*** 0.000*** 0.511 0.001*** 0.699 0.248 0.592
Retail 0.003*** 0.079* 0.015** 0.138 0.000*** 0.167 0.658 0.130 0.028**
Technology 0.037** 0.233 0.000*** 0.001*** 0.771 0.000*** 0.011** 0.015** 0.048**
Telecommunications 0.000*** 0.202 0.345 0.666 0.270 0.046** 0.018** 0.001*** 0.023**
Travel and Leisure 0.711 0.426 0.007*** 0.000*** 0.001*** 0.523 0.001*** 0.000*** 0.000***
Utilities 0.754 0.884 0.014** 0.000*** 0.000*** 0.628 0.004*** 0.000*** 0.000***
Panel B: 6-day event window
Automobiles 0.907 0.118 0.508 0.675 0.283 0.069* 0.529 0.526 0.171
Banks 0.005*** 0.105 0.000*** 0.001*** 0.000*** 0.224 0.062* 0.131 0.003***
Basic Resources 0.016** 0.826 0.345 0.458 0.176 0.002*** 0.131 0.137 0.243
Chemicals 0.931 0.382 0.138 0.000*** 0.290 0.312 0.082* 0.000*** 0.195
Construction 0.002*** 0.000*** 0.002*** 0.064* 0.000*** 0.100 0.952 0.463 0.156
Financial services 0.341 0.685 0.003*** 0.285 0.118 0.436 0.003*** 0.754 0.292
Food and Beverage 0.169 0.176 0.403 0.000*** 0.002*** 0.864 0.629 0.000*** 0.000***
Healthcare 0.614 0.078* 0.109 0.254 0.738 0.027** 0.054* 0.007*** 0.230
Industrial Goods 0.370 0.112 0.028** 0.025** 0.050** 0.294 0.090* 0.061* 0.125
Insurance 0.360 0.587 0.058* 0.126 0.039** 0.605 0.125 0.266 0.080*
Media 0.865 0.000*** 0.286 0.076* 0.000*** 0.000*** 0.087* 0.016** 0.000***
Oil and Gas 0.003*** 0.000*** 0.095* 0.143 0.007*** 0.000*** 0.165 0.264 0.995
Personal Goods 0.201 0.037** 0.066* 0.599 0.501 0.256 0.393 0.056* 0.028**
Real Estate 0.057* 0.018** 0.002*** 0.001*** 0.419 0.702 0.150 0.066* 0.872
Retail 0.137 0.000*** 0.012** 0.005*** 0.002*** 0.001*** 0.156 0.120 0.008***
Technology 0.957 0.571 0.068* 0.000*** 0.104 0.325 0.007*** 0.000*** 0.018**
Telecommunications 0.895 0.458 0.783 0.954 0.999 0.405 0.600 0.967 0.897
Travel and Leisure 0.054* 0.475 0.247 0.029** 0.001*** 0.083* 0.000*** 0.000*** 0.000***
Utilities 0.000*** 0.306 0.497 0.263 0.204 0.025** 0.004*** 0.000*** 0.000***
Notes: This table presents the p-values from equality tests of cumulative abnormal returns calculated using the abnormal returns estimates from a seemingly unrelated
regressions model with seven equations, each representing one country. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.
other countries reacted to the earthquake. Nonetheless, studying the behavior of industries that are directly hit by the earthquake
(e.g., automobiles, insurance, technology, and utilities) and the resilience of banking and financial sectors might be particularly
important. To this end, Fig. 2 displays the patterns of estimated BHARs during the [0,104] interval for these selected sectors.
Fig. 2(a) shows that the automobile sector in Japan and its major trading partners was adversely affected in this longer run 3-
month period, whereas minor trading partners overall benefited from the earthquake. For the US, this result suggests that although
the earthquake was perceived to create some opportunities for the American automakers in the immediate aftermath of the
earthquake because of the expected increase in demand, in the longer run the negative effects caused by the shortage of automobile
parts from Japan may have outweighed the positive impacts. However, the magnitude of the adverse effect is small, particularly when
it is compared to China. European countries, on the other hand, are not a major market for Japanese cars and the shortage of
Japanese automobiles could expand the opportunities for non-Japanese automobile industries in global markets. For instance,
Germany could benefit mainly from an increased share in international markets, as it is one of the main exporters of automobiles in
the world, while Italy and Spain could benefit mostly from their domestic markets when Japanese cars are in low supply. However,
the magnitude of the impact for Spain might be surprising. A detailed look at the Spain’s automobiles and parts sector index reveals
that it is comprised of only auto parts, while for other countries the index includes both automobiles and auto parts. This might
explain the large impact on the Spain’s automobiles and parts sector because vehicle parts are among Spain’s major exports (OEC,
2013).
It is obvious in Fig. 2(b) that almost all countries suffered from disruptions in the Japanese production of the high-tech products
566
Table 5
Average Buy-and-Hold Abnormal Returns.
Automobiles −7.47 ***

−8.27 ***
−1.83 ***
8.57***
2.71***
10.06 ***
20.78***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Banks −13.55*** 0.18 −9.74*** −6.17*** −4.67*** −13.15*** −1.79***
(0.000) (0.493) (0.000) (0.000) (0.000) (0.000) (0.004)
Basic Resources −9.91*** −2.10*** 1.17*** 1.67*** −2.40*** 2.67*** −0.28
(0.000) (0.000) (0.003) (0.000) (0.001) (0.000) (0.521)
Chemicals −6.37*** −5.29*** 3.78*** 7.41*** 8.92*** 10.41*** 6.53***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Construction −3.08*** −5.03*** −1.68*** −2.30*** 2.11*** −8.73*** −0.29
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.595)
Financial services −17.25*** −10.71*** −1.76*** −0.21 8.23*** 3.36*** −1.36***
(0.000) (0.000) (0.000) (0.299) (0.000) (0.000) (0.000)
Food and Beverage −4.48*** −1.84*** 3.17*** 7.76*** 8.59*** 7.70*** 2.22***
(0.000) (0.001) (0.000) (0.000) (0.000) (0.000) (0.000)
Healthcare −3.70*** −10.43*** 4.78*** 5.33*** 6.09*** 5.18*** 2.84***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Industrial Goods −5.29*** −10.58*** 0.62*** 4.74*** 4.85*** −1.87*** 3.46***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.001) (0.000)
Insurance −14.35*** −7.41*** −5.54*** −4.32*** −1.48*** −6.85*** 3.76***
(0.000) (0.000) (0.000) (0.000) (0.003) (0.000) (0.000)
Media −11.74*** −16.08*** 0.35** 4.95*** −2.93*** −10.47*** −13.18***
(0.000) (0.000) (0.019) (0.000) (0.000) (0.000) (0.000)
Oil and Gas −1.55*** −6.60*** 1.21*** 6.38*** −2.74*** −2.93*** −1.10***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.001)
Personal Goods −9.65*** −3.17*** 3.40*** 6.62*** 6.56*** 1.68*** 3.81***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Real Estate −10.92*** −2.26*** 2.95*** −1.09*** 3.28*** −1.88*** −9.52***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.002) (0.000)
Retail −3.97*** −10.84*** 2.88*** −7.21*** 0.84** 5.47*** 16.09***
(0.000) (0.000) (0.000) (0.000) (0.017) (0.000) (0.000)
Technology −9.36*** −14.20*** −1.20*** 0.98*** −0.32 0.16 1.84***
(0.000) (0.000) (0.000) (0.000) (0.434) (0.791) (0.002)
Telecommunications −4.85*** −5.14*** 3.08*** 6.32*** −1.10*** −12.49*** −4.41***
(0.000) (0.000) (0.000) (0.000) (0.009) (0.000) (0.000)
Travel and Leisure −11.03*** −8.69*** 0.78*** −3.05*** −0.30* 4.45*** 3.09***
(0.000) (0.000) (0.007) (0.000) (0.051) (0.000) (0.000)
Utilities −37.04*** −3.30*** 0.80*** −8.92*** −6.29*** 5.84*** 3.96***
(0.000) (0.000) (0.003) (0.000) (0.000) (0.000) (0.000)
Notes: This table presents average buy-and-hold abnormal returns calculated for [0,104] period and their p-values in parentheses. *, **
, and ***
indicate statistical
significance at the 10%, 5%, and 1% levels, respectively.
which may indicate that these products are not easy to substitute. Fig. 2(c) shows that although the banking sector in some countries
reacted positively in the short run, the average effects were negative in the longer run. The financial services sector in Fig. 2(d)
reveals a similar pattern, except for France. These results suggest that banking and financial systems in the selected countries were
not probably resilient to the earthquake. However, we should note that if financial institutions anticipated the stock market crisis of
August 2011, our results might be contaminated with these expectations, and therefore overstate the earthquake impacts on these
sectors.
Fig. 2(e) illustrates how the insurance sector is negatively affected in the longer run in all countries except for Spain. The impacts
of natural disasters on insurance have been studied the most (compared to other sectors) in the literature and our results are
consistent with the earlier studies. After natural disasters, insurance companies face two opposing effects. On the one hand, they raise
their premium rates in response to the increase in demand for coverage. On the other hand, they make payments for claims after
disasters, and thus experience capital outflows. For Spain these two effects might have offset each other resulting in no significant
impact, whereas for other countries the latter effect has probably outweighed the former. Alternatively, it could simply imply that
Spanish insurance companies did not have stakes in Japan’s insurance industry. Finally, Fig. 2(f) displays how utilities in Japan
suffered the most from the earthquake and nuclear accident with stock prices dropping by close to 40%. After the Fukushima nuclear
accident, negative attitudes toward nuclear energy in some countries, particularly Germany, intensified and the negative BHARs
there and in France, where nuclear power is a sizeable contributor to the energy mix, could be in part induced by this change.
5. Discussion
In an increasingly connected world, the economic repercussions from a natural disaster are not likely to be confined to the
567
Fig. 2. Estimated buy-and-hold abnormal returns, 0 to 104 0–104 days.
geographic area affected directly. In this paper we find that the occurrence of the 2011 Japan earthquake, a sizeable natural disaster,
affected not only Japan’s domestic stock market but also stock markets in other countries. Given the magnitude of this earthquake,
the catastrophic nature of the subsequent tsunami and nuclear meltdown, and the role of Japan in the world economy, some caution
is required in extending the results of this paper to other natural disasters. However, some lessons can be learned to build resilience in
the face of more intense and more frequent natural (especially hydrometeorological) disasters.
Finding evidence of contagion in financial markets following natural disasters could be of particular importance in investment
decisions. Over the past decade, through the popularity of different forms of international investment diversification, such as
exchange-traded funds (ETFs), the ability of domestic investors to own shares of companies based overseas has expanded
substantially. Natural disaster-induced contagion in stock markets could expose investors to higher risks, even as they try to gain
from diversification through investing in stock markets of other countries, if those countries are disaster-prone.
Natural disasters are often defined as the combination of a natural hazard with existing conditions of exposure and vulnerability.
While little can be done to prevent natural hazards, their financial risks could be reduced in a way that protects people and physical
assets. For example, governments have managed the financial risks of natural disasters through insurance programs; the government
of Japan established an earthquake insurance program for homeowners in 1966, and the US government provides residential flood
insurance through the National Flood Insurance Program since 1968. In this context, the large abnormal returns to Japan’s insurance
sector that we report in the paper are to be expected.
More broadly, the World Bank and the Global Facility for Disaster Reduction and Recovery (GFDRR) provide a comprehensive
approach for disaster risk management. According to their framework, “risk identification; risk reduction; preparedness; financial
protection; and planning for disaster recovery” are necessary actions for building resilience against natural disasters (World Bank
Group, 2014). Findings from our study suggest that policymakers and investors should be concerned about natural disaster risks as
they can have large impacts on stock markets but, importantly, risk identification trascends national borders—the occurrence of high-
profile natural disasters in an industrialized country could affect other stock markets, even if these are in developed countries with
strong economies. Therefore, policymakers should be prepared for catastrophic natural disasters in foreign countries, not only in
terms of providing disaster-relief aid, but, more selfishly, in terms of limiting the potential spillover effects to their own domestic
markets.
Our findings can also help investors reduce their portfolio risks. We find that the earthquake’s impact was not the same for all
sectors; some sectors were adversely affected, whereas others benefited. Thus, diversification across stock market sectors and
countries seems to be an effective way to reduce the increased risks due to contagion in financial markets. For example, owning the
shares of US automobile manufacturers could reduce/offset the risk of buying shares of Japanese car makers. As another example,
including in one’s portfolio shares of companies involved in the reconstruction after natural catastrophes seems to reduce the
financial risks associated with natural disasters.
568
6. Conclusions
This paper investigated how the impacts of one of the largest disasters ever recorded spread across the stock market of Japan and
of its trading partners. We examined the short-run and long-run impacts of the 2011 Japanese earthquake on 19 sector indices in the
stock markets of Japan and its major (China and US) and minor (Germany, France, Italy, and Spain) trading partners.
We estimated abnormal returns and cumulative abnormal returns in the days following the earthquake through a regression-based
event study methodology. Our findings indicated that, in the short run, all business sectors in the Japanese stock market were
negatively affected, whereas some sectors in other countries gained positive abnormal returns following the earthquake. In general,
the sign and magnitude of the abnormal returns varied across countries and sectors. This result is consistent with the work of
Worthington and Valadkhani (2005) who find varying reactions in the Australian stock market to natural disasters, industrial
catastrophes, and terrorist attacks. However, a few common patterns emerge from our analysis. Similar to Lamb (1995), Cagle
(1996), Angbazo and Narayanan (1996), Yamori and Kobayashi (2002), and Takao et al. (2013), we find negative impacts from the
earthquake on the insurance industry across all countries. In addition, our short-run results for the utilities sector are consistent with
the literature focusing on the implications of the Fukushima nuclear accident, such as Kawashima and Takeda (2012), Lopatta and
Kaspereit (2014), Betzer et al. (2013), and Mama and Bassen (2013).
We also compared the magnitude of the impacts on each sector across all countries by statistically testing the equality of the
cumulative abnormal returns across major and minor trading partners. Results implied that there is no direct link between the
magnitude of the impact and being a major or a minor trading partner of Japan.
The long-run analysis was conducted via a buy-and-hold abnormal return approach. We found that all the sectors in the Japanese
stock market were adversely affected in the long run, as were those in China, with industries directly hit by the earthquake, such as
technology, telecommunication, media, and insurance, more affected. Results also indicated that banks, financial services, and other
sectors involved in the trade of services experienced decreases in the value of their stock indices in the longer run. However, because
markets may have anticipated the sharp drop in stock prices in August 2011, not all the impacts on banks and financial services
should be attributed to the earthquake.
Overall, the short- and long-run results suggested that stock markets of Japan and its trading partners were not resilient to the
massive earthquake on March 11, 2011. Similar to Asongu (2012), we showed that international financial markets are affected by
high-profile natural disasters, which is an indication of spillover or contagion in these markets. Our analysis, however, allows us to
identify what parts of the real economy are responsible for the effects and in which direction. This finding could be of importance as
the research in this area is limited and inconclusive. In a world in which the frequency and intensity of natural disasters are on the
rise, this type of analysis should prove beneficial to investors and policymakers.
Appendix A
Table A1
Cumulative Abnormal Returns (CARs) from Cross-country SUR.
Panel A: 3-day CARs

Automobiles −14.28*** 1.17 4.98*** 0.51 1.29** 3.85*** 2.67***
(0.000) (0.195) (0.000) (0.354) (0.043) (0.000) (0.006)
Banks −15.92*** −0.24 2.16*** 2.91*** 7.53*** 8.46*** 10.69***
(0.000) (0.669) (0.002) (0.000) (0.000) (0.000) (0.000)
Basic Resources −19.09*** 0.19 7.43*** −0.06 2.66*** 3.66*** 3.42***
(0.000) (0.847) (0.000) (0.894) (0.000) (0.000) (0.000)
Chemicals −20.31*** 1.24 3.68*** 1.50*** −0.77 3.66*** 0.71
(0.000) (0.162) (0.000) (0.003) (0.134) (0.000) (0.530)
Construction −6.04*** −1.83** 4.61*** 3.02*** 3.50*** 4.07*** 6.32***
(0.000) (0.012) (0.000) (0.000) (0.000) (0.000) (0.000)
Financial services −18.25*** −1.14 1.90*** −0.86 3.15*** 2.12*** 2.59***
(0.000) (0.251) (0.000) (0.140) (0.000) (0.000) (0.000)
Food and Beverage −15.92*** −1.48** −0.87*** −4.25*** −0.92* 3.57*** 3.27***
(0.000) (0.038) (0.004) (0.000) (0.062) (0.000) (0.000)
Healthcare −12.51*** 2.15** 0.55* −1.12*** −1.59*** 2.87*** 1.34**
(0.000) (0.012) (0.082) (0.006) (0.004) (0.000) (0.024)
Industrial Goods −16.22*** −0.59 2.11*** 1.39*** 1.55*** 2.67*** 2.60***
(0.000) (0.416) (0.000) (0.000) (0.000) (0.000) (0.000)
Insurance −16.05*** −1.74** 0.48 −1.42** 0.79 3.47*** 3.88***
(0.000) (0.026) (0.205) (0.011) (0.393) (0.000) (0.000)
Media −20.20*** −1.69 1.30*** 1.36 1.16** 4.74*** 7.40***
(0.000) (0.125) (0.001) (0.109) (0.038) (0.000) (0.000)
Oil and Gas −15.22*** −0.97* 4.52*** 24.38*** 1.91*** 3.08*** 2.52***
(0.000) (0.077) (0.000) (0.000) (0.000) (0.000) (0.000)
Personal Goods −13.75*** 0.01 −0.46 −1.10** −1.08** 0.89 0.37
(continued on next page)
569
Table A1 (continued)
(0.000) (0.988) (0.132) (0.012) (0.020) (0.423) (0.698)

Real Estate −17.50*** −2.44*** 2.04*** −0.75 1.72*** 3.13*** 0.12
(0.000) (0.007) (0.000) (0.140) (0.001) (0.000) (0.974)
Retail −19.25*** −1.98** 0.77* −0.20 0.42 −0.54 2.44***
(0.000) (0.012) (0.071) (0.749) (0.418) (0.394) (0.001)
Technology −14.37*** −0.07 1.86*** −1.28*** 3.92*** 4.08*** 0.26
(0.000) (0.938) (0.000) (0.010) (0.000) (0.000) (0.686)
Telecommunications −13.82*** 2.27*** −0.91** 0.84 1.17* 2.90*** 1.01
(0.000) (0.003) (0.031) (0.257) (0.096) (0.009) (0.133)
Travel and Leisure −17.54*** −1.27 −0.84 −0.29 1.96*** 5.58*** 3.74***
(0.000) (0.252) (0.129) (0.680) (0.000) (0.000) (0.000)
Utilities −20.98*** −1.51** −1.25*** −1.66** 0.92 3.60*** 4.46***
(0.000) (0.048) (0.000) (0.014) (0.136) (0.000) (0.000)
Panel B: 6-day CARs
Automobiles −9.31*** 2.00 1.80* −0.53 0.91 2.81** 4.26***
(0.000) (0.126) (0.070) (0.541) (0.336) (0.030) (0.004)
Banks −11.45*** −1.41* 2.35** 0.77 5.87*** 5.18*** 8.17***
(0.000) (0.087) (0.018) (0.457) (0.000) (0.003) (0.000)
Basic Resources −9.34*** 2.50** 6.59*** 2.18*** 4.02*** 3.86*** 4.72***
(0.000) (0.043) (0.000) (0.001) (0.000) (0.001) (0.000)
Chemicals −8.58*** 2.36* 2.23*** 1.04 0.07 9.32*** 5.01***
(0.000) (0.064) (0.003) (0.171) (0.926) (0.000) (0.010)
Construction 2.16** −1.28 2.82*** 4.83*** 2.90*** 1.74 5.02***
(0.039) (0.218) (0.001) (0.000) (0.001) (0.156) (0.000)
Financial services −7.86*** −0.49 1.07 0.23 4.47*** 1.43 2.36**
(0.001) (0.732) (0.121) (0.793) (0.000) (0.112) (0.013)
Food and Beverage −10.36*** −0.30 −1.96*** −2.12*** −1.49* 8.50*** 4.08***
(0.000) (0.781) (0.000) (0.010) (0.058) (0.000) (0.000)
Healthcare −5.58*** 0.52 −0.24 −2.02*** −2.08** 2.54*** 1.10
(0.000) (0.707) (0.615) (0.002) (0.011) (0.004) (0.243)
Industrial Goods −6.38*** −0.46 0.65 1.65** 2.34*** 2.72*** 2.61***
(0.001) (0.659) (0.253) (0.012) (0.000) (0.001) (0.008)
Insurance −13.67*** −2.34** −1.09* −1.57** 1.27 0.40 1.66
(0.000) (0.038) (0.062) (0.039) (0.357) (0.751) (0.242)
Media −15.52*** −1.10 −0.80 6.80*** 1.04 2.76** 8.49***
(0.000) (0.517) (0.176) (0.000) (0.175) (0.022) (0.000)
Oil and Gas −0.88 0.15 3.52*** 16.03*** 1.88*** 2.00** 3.51***
(0.653) (0.853) (0.000) (0.000) (0.007) (0.033) (0.000)
Personal Goods −5.75*** −0.39 −2.17*** −3.15*** −3.02*** 0.59 0.81
(0.000) (0.743) (0.000) (0.000) (0.000) (0.660) (0.504)
Real Estate −6.36*** −2.18 1.06 1.53** 2.85*** 3.66*** 1.80
(0.000) (0.109) (0.198) (0.034) (0.000) (0.001) (0.691)
Retail −14.66*** −3.02*** −1.03* 2.39*** 0.60 1.16 2.34**
(0.000) (0.007) (0.082) (0.003) (0.466) (0.311) (0.039)
Technology −10.61*** −1.05 −1.13* −0.16 1.84** 6.35*** 1.90*
(0.000) (0.416) (0.058) (0.828) (0.021) (0.000) (0.079)
Telecommunications −10.12*** 0.26 0.08 −0.97 0.72 0.15 0.26
(0.000) (0.823) (0.883) (0.385) (0.480) (0.921) (0.831)
Travel and Leisure −9.65*** −0.80 −4.14*** −2.09** 1.32 4.63*** 6.42***
(0.000) (0.616) (0.000) (0.030) (0.100) (0.004) (0.000)
Utilities −13.38*** 0.92 −3.34*** −0.59 −0.04 2.56*** 2.88***
(0.000) (0.347) (0.000) (0.565) (0.970) (0.007) (0.005)
Notes: This table presents cumulative abnormal returns (and their p-values in parentheses) calculated using the abnormal return estimates from a seemingly unrelated
regressions model with seven equations, each representing one county. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.
570
Fig. A1. Cumulative abnormal returns (CARs) for Japan.
Fig. A2. Cumulative abnormal returns (CARs) for China.
571
Fig. A3. Cumulative abnormal returns (CARs) for US.
Fig. A4. Cumulative abnormal returns (CARs) for Germany.
572
Fig. A5. Cumulative abnormal returns (CARs) for France.
Fig. A6. Cumulative abnormal returns (CARs) for Italy.
573
Fig. A7. Cumulative abnormal returns (CARs) for Spain.
Fig. A8. Cumulative abnormal returns (CARs) on selected sectors across countries.
References
Ahlgren, N., Antell, J., 2010. Stock market linkages and financial contagion: a cobreaking analysis. Q. Rev. Econ. Finance 50 (2), 157–166.
574
Aiuppa, T.A., Carney, R.J., Krueger, T.M., 1993. An examination of insurance stock prices following the 1989 Loma Prieta Earthquake. J. Insur. 16 (1), 1–14.
Angbazo, L.A., Narayanan, R., 1996. Catastrophic shocks in the property-liability insurance industry: evidence on regulatory and contagion effects. J. Risk Insur. 63
(4), 619–637.
Asongu, S.A., 2012. The 2011 Japanese earthquake, tsunami and nuclear crisis: evidence of contagion from international financial markets. J. Financ. Econ. Policy 4
(4), 340–353.
Barber, B.M., Lyon, J.D., 1997. Detecting long-run abnormal stock returns: the empirical power and specification of test statistics. J. Financ. Econ. 43 (3), 341–372.
Behravesh, N., 2015. Global Economic Impact of the Japanese Earthquake, Tsunami, and Nuclear Disaster. Available at: https://www.ihs.com/country-industry-
forecasting.html?ID=1065929196.
Bekaert, G., Harvey, C.R., 1995. Time-varying world market integration. J. Finance 50 (2), 403–444.
Betzer, A., Doumet, M., Rinne, U., 2013. How policy changes affect shareholder wealth? The case of the Fukushima Dai-ichi nuclear disaster. Appl. Econ. Lett. 20 (8),
799–803.
Binder, J.J., 1985. On the use of the multivariate regression model in event studies. J. Account. 23 (1), 370–383.
Binder, J.J., 1998. The event study methodology since 1969. Rev. Quant. Finance Account. 11 (2), 111–137.
Boubaker, S., Jouini, J., Lahiani, A., 2015. Financial contagion between the US and selected developed and emerging countries: the case of the subprime crisis. Q. Rev.
Econ. Finance 61, 14–28.
Bourdeau-Brien, M., Kryzanowski, L., 2017. The impact of natural disasters on the stock returns and volatilities of local firms. Q. Rev. Econ. Finance 63, 259–270.
Bremer, C., Dmitracova, O., 2011. Analysis: France, Britain AAA-ratings Under Scrutiny. Reuters, Paris/London Available at: http://uk.reuters.com/article/2011/08/
08/uk-crisis-ratings-idUKTRE7773LW20110808.
Breusch, T.S., 1978. Testing for autocorrelation in dynamic linear models. Aust. Econ. Pap. 17 (31), 334–355.
CRED, 2011. Centre for Research on the Epidemiology of Disasters. CRED Crunch Newsletter Issue No. 25, September 2011. Available at: http://reliefweb.int/report/
world/cred-crunch-newsletter-issue-25-september-2011-2011-first-semester-balance.
Cagle, J.A.B., 1996. Natural disasters, insurer stock prices, and market discrimination: the case of hurricane Hugo. J. Insur. Issues 19 (1), 53–68.
Campbell, C.J., Cowan, A.R., Salotti, V., 2010. Multi-country event-study methods. J. Bank. Finance 34 (12), 3078–3090.
Canis, B., 2011. The Motor Vehicle Supply Chain: Effects of the Japanese Earthquake and Tsunami. Congressional Research Service, Washington D.C.
Carpentier, C., Suret, J.M., 2015. Stock market and deterrence effect: a mid-run analysis of major environmental and non-environmental accidents. J. Environ. Econ.
Manage. 71, 1–18.
Chan, K.C., Gup, B.E., Pan, M.S., 1997. International stock market efficiency and integration: a study of eighteen nations. J. Bus. Finance Account. 24 (6), 803–813.
Chevapatralul, T., Tee, K.-H., 2014. The effects of news events on market contagion: evidence from the 2007–2009 financial crisis. Res. Int. Bus. Finance 32, 83–105.
De Angelis, L., Gardini, A., 2015. Disequilibria and contagion in financial markets: evidence from a new test. J. Appl. Econ. 18 (2), 247–265.
De Jong, A., Naumovska, I., 2015. A note on event studies in finance and management research. Rev. Finance 20 (4), 1659–1672.
Edgerton, J., 2012. How the Japanese Tsunami Changed the Auto Industry. Available at: http://www.cbsnews.com/news/how-the-japanese-tsunami-changed-the-
auto-industry/.
Fama, E.F., Fisher, L., Jensen, M.C., Roll, R., 1969. The adjustment of stock prices to new information. Int. Econ. Rev. 10 (1), 1–21.
Fama, E.F., 1970. Efficient capital markets: a review of theory and empirical work. J. Finance 25 (2), 383–417.
Ferreira, S., Karali, B., 2015. Do earthquakes shake stock markets? PLoS One 10 (7), e0133319.
Ferstl, R., Utz, S., Wimmer, M., 2012. The effect of the Japan 2011 disaster on nuclear and alternative energy stocks worldwide: an event study. BuR-Bus. Res. 5 (1),
25–41.
Forbes, K.J., 2002. Are trade linkages important determinants of country vulnerability to crises? In: Edwards, S., Frankel, J. (Eds.), Preventing Currency Crises in
Emerging Markets. University of Chicago Press.
Gibbons, R., 2011. Japan's Earthquake Leads To Falling Oil Prices On Fears Of Decreased Demand. Thomson Reuters Available at: http://www.huffingtonpost.com/
2011/03/11/oil-prices-fall-in-early-_n_834404.html.
Godfrey, L.G., 1978. Testing against general autoregressive and moving average error models when the regressors include lagged dependent variables. Econometrica
46, 1293–1301.
JETRO, 2011. Japan External Trade Organization. Some Economic and Trade Impacts of the Earthquake, Tsunami and Nuclear Incident in Japan. Available at: http://
www.jetro.go.jp/australia/topics/20110603220-topics.html.
Jayech, S., 2016. The contagion channels of July-August-2011 stock market crash: a DAG-Copula based approach. Eur. J. Oper. Res. 249 (2), 631–646.
Johnson, N.J., 1978. Modified t-tests and confidence intervals for asymmetrical populations. J. Am. Stat. Assoc. 73 (363), 536–544.
Kachi, H., Takahashi, Y., 2011. Plant closures imperil global supplies. Wall Street J Available at: http://www.wsj.com/articles/
SB10001424052748704027504576198961775199034.
Kajitani, Y., Chang, S.E., Tatano, H., 2013. Economic impacts of the 2011 Tohoku-Oki earthquake and tsunami. Earthquake Spectra 29 (S1), S457–S478.
Karolyi, G.A., Stulz, R.M., 1996. Why do markets move together? An investigation of US-Japan stock return comovements. J. Finance 51 (3), 951–986.
Kawashima, S., Takeda, F., 2012. The effect of the Fukushima nuclear accident on stock prices of electric power utilities in Japan. Energy Econ. 34 (6), 2029–2038.
Khazai, B., Daniell, J.E., 2011. The March 2011 Japan Earthquake. analysis of losses, impacts, and implications for the understanding of risks posed by extreme events.
Technikfolgenabschätzung −Theorie und Parxis 20 (3), 22–33.
Kontonikas, A., MacDonald, R., Saggu, A., 2013. Stock market reaction to fed funds rate surprises: state dependence and the financial crisis. J. Bank. Finance 37 (11),
4025–4037.
Kothari, S.P., Warner, J.B., 1997. Measuring long-horizon security price performance. J. Financ. Econ. 43 (3), 301–339.
Krämer, W., Schich, S., 2008. Large-scale Disasters and the Insurance Industry. CESifo Working Paper No. 2243.
Lamb, R.P., 1995. An exposure-based analysis of property-liability insurer stock values around hurricane Andrew. J. Risk Insur. 62 (1), 111–123.
Lee, H., Wu, H., Wang, Y., 2007. Contagion effect in financial markets after the South-East Asia tsunami. Res. Int. Bus. Finance 21 (2), 281–296.
Lehkonen, H., 2015. Stock market integration and the global financial crisis. Rev. Finance 19 (5), 2039–2094.
Lopatta, K., Kaspereit, T., 2014. The cross-section of returns, benchmark model parameters, and idiosyncratic volatility of nuclear energy firms after Fukushima
Daiichi. Energy Econ. 41, 125–136.
Lyon, J.D., Barber, B.M., Tsai, C.L., 1999. Improved methods for tests of long-run abnormal stock returns. J. Finance 54 (1), 165–201.
MacKinlay, A.C., 1997. Event studies in economics and finance. J. Econ. Lit. 35 (1), 13–39.
Mama, H.B., Bassen, A., 2013. Contagion effects in the electric utility industry following the Fukushima nuclear accident. Appl. Econ. 45 (24), 3421–3430.
Nanto, D.K., Cooper, W.H., Donnelly, J.M., Johnson, R., 2011. Japan’s 2011 Earthquake and Tsunami: Economic Effects and Implications for the United States.
Congressional Research Service, Washington Available at: http://www.fas.org/sgp/crs/row/R41702.pdf.
OEC, 2013. The Observatory of Economic Complexity. Available at: http://atlas.media.mit.edu/en/profile/country/esp/.
Park, N.K., 2004. A guide to using event study methods in multi-country settings. Strateg. Manage. J. 25 (7), 655–668.
Rodriguez, J.C., 2007. Measuring financial contagion: a copula approach. J. Empir. Finance 14 (3), 401–423.
Schneider, M., Froggatt, A., Thomas, S., 2011. Nuclear Power in a Post-Fukushima World. The World Nuclear Industry Status Report 2010–2011. Worldwatch Institute.
Schneider, M., 2011. Fukushima Crisis: Can Japan Be at the Forefront of an Authentic Paradigm Shift? Bulletin of the Atomic Scientists. Available at: http://
thebulletin.org/fukushima-crisis-can-japan-be-forefront-authentic-paradigm-shift.
Shelor, R.M., Anderson, D.C., Cross, M.L., 1990. The impact of the California earthquake on real estate firms’ stock value. J. Real Estate Res. 5 (3), 335–340.
Shelor, R.M., Anderson, D.C., Cross, M.L., 1992. Gaining from loss: property-liability insurer stock values in the aftermath of the 1989 California earthquake. J. Risk
575
Insur. 59 (3), 476–488.

Takao, A., Yoshizawa, T., Hsu, S., Yamasaki, T., 2013. The effect of the great East Japan earthquake on the stock prices of non-life insurance companies. The Geneva
Papers on Risk and Insurance 38 (3), 449–468.
The Economist, 2011. The Global Car Industry After the Quake: Japanese Carmakers Are Shaken. That Helps Their Rivals. Available at: http://www.economist.com/
node/18712767.
Wang, L., Kutan, A.M., 2013. The impact of natural disasters on stock markets: evidence from Japan and the US. Compar. Econ. Stud. 55 (4), 672–686.
World Bank Group, 2014. Financial Protection Against Natural Disasters: An Operational Framework for Disaster Risk Financing and Insurance. World Bank,
Washington, DC Available at: https://openknowledge.worldbank.org/handle/10986/21725.
Worthington, A.C., Valadkhani, A., 2004. Measuring the impact of natural disasters on capital markets: an empirical application using intervention analysis. Appl.
Econ. 36 (19), 2177–2186.
Worthington, A.C., Valadkhani, A., 2005. Catastrophic shocks and capital markets: a comparative analysis by disaster and sector. Glob. Econ. Rev. 34 (3), 331–344.
Worthington, A.C., 2008. The impact of natural events and disasters on the Australian stock market: a GARCH-M analysis of storms, floods, cyclones, earthquakes and
bushfires. Glob. Bus. Econ. Rev. 10 (1), 1–10.
Yamori, N., Kobayashi, T., 2002. Do Japanese insurers benefit from a catastrophic event? Market reactions to the 1995 Hanshin-Awaji earthquake. J. Jpn. Int. Econ. 16
(1), 92–108.
576

Ripple Effects of The 2011 Japan Earthquake On International Stock Markets PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ripple Effects of The 2011 Japan Earthquake On International Stock Markets PDF

Uploaded by

Copyright:

Available Formats

Research in International Business and Finance 41 (2017) 556–576

Contents lists available at ScienceDirect

Research in International Business and Finance

Ripple eﬀects of the 2011 Japan earthquake on international stock MARK

2.1. Short-run analysis

2.2. Long-run analysis

Japan’s Exports Japan’s Imports

China 19.41 22.07

Sector Major subsectors

Automobiles and Parts Automobiles

4.1. Short-run analysis

Sector Japan China US Germany France Italy Spain

Panel A: 3-day CARs

Sector Japan China US Germany France Italy Spain

Retail −13.19*** −3.21*** −0.90 2.38*** 0.52 1.14 2.40**

4.2. Long-run analysis

Sector China vs. US vs.

US Germany France Italy Spain Germany France Italy Spain

Panel A: 3-day event window

Sector Japan China US Germany France Italy Spain

Automobiles −7.47 ***

Fig. 2. Estimated buy-and-hold abnormal returns, 0 to 104 0–104 days.

Sector Japan China US Germany France Italy Spain

Panel A: 3-day CARs

Sector Japan China US Germany France Italy Spain

(0.000) (0.988) (0.132) (0.012) (0.020) (0.423) (0.698)

Fig. A1. Cumulative abnormal returns (CARs) for Japan.

Fig. A2. Cumulative abnormal returns (CARs) for China.

Fig. A3. Cumulative abnormal returns (CARs) for US.

Fig. A4. Cumulative abnormal returns (CARs) for Germany.

Fig. A5. Cumulative abnormal returns (CARs) for France.

Fig. A6. Cumulative abnormal returns (CARs) for Italy.

Fig. A7. Cumulative abnormal returns (CARs) for Spain.

Insur. 59 (3), 476–488.

You might also like

Retail −13.19* −3.21* −0.90 2.38* 0.52 1.14 2.40