Global terrorism and adaptive expectations in financial markets: Evidence from Japanese equity market

Global terrorism and adaptive expectations in financial markets: Evidence from Japanese equity market

Research in International Business and Finance 26 (2012) 97–119 Contents lists available at ScienceDirect Research in International Business and Fin...

461KB Sizes 0 Downloads 14 Views

Research in International Business and Finance 26 (2012) 97–119

Contents lists available at ScienceDirect

Research in International Business and Finance j o ur na l ho me pa ge : w w w . e l s e v i e r . c o m / l o c a t e / r i b a f

Global terrorism and adaptive expectations in financial markets: Evidence from Japanese equity market Michael A. Graham ∗, Vikash B. Ramiah School of Economics, Finance, and Marketing, Royal Melbourne Institute of Technology (RMIT) University, Melbourne, VIC 3001, Australia

a r t i c l e

i n f o

Article history: Received 16 March 2011 Received in revised form 7 July 2011 Accepted 8 July 2011 Available online 20 July 2011 JEL classification: D84 F20 F50 G1 G11 H56 Keywords: Terrorism Equity market Abnormal returns Systematic risk Adaptive expectation Non-parametric test Parametric test

a b s t r a c t The adaptive expectations model posits that economic agents’ expectations adjust by constant proportion of previous discrepancy and the forecast for the following period is the same for all the subsequent future periods, if the expectation is a permanent. We apply this hypothesis and event study methodology to examine the impact of five terrorist attacks (New York World Trade Centre, Bali, Madrid, London, and Mumbai) on Japanese industries. Being a watershed event, the negative impact of the attacks in the U.S. was apparent. Our evidence suggests an initial step-change in risk incorporated into expectations after the U.S., Bali and Madrid bombings. The two subsequent attacks had no effect on the market implying no the forecast error in risk expectation in Japan after the initial terrorist attacks. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Terrorism, a significant geopolitical risk that affects global financial markets, is not a recent phenomenon. The terrorist attack on the World Trade Centre twin towers in New York on September

∗ Corresponding author. E-mail addresses: [email protected] (M.A. Graham), [email protected] (V.B. Ramiah). 0275-5319/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.ribaf.2011.07.002

98

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

11, 2001 (9/11) has, however, proven to be a watershed event in global terrorism and fundamentally changed business operations and risk perception. Rosendorff and Sandler (2005) accentuate this point when they suggest that the event aptly underscored the susceptibility of modern societies to terrorist attacks as an everyday object could be utilised to wreck havoc on a gigantic scale. They also note that the magnitude of the attack and the resulting bloodshed were unprecedented in terms of terrorist attacks and that the financial losses resulting from the event (about USD 90 billion) were far greater than those associated with the terrorist attacks to date. Furthermore, they propose 9/11 set the standard for future attacks as terrorists try to exceed the scale of past attacks to capture and maintain media attention. This epochal event has intensified interest in the threat posed by modern-day terrorism and a new development on how terrorist risk is perceived. Combined with the globalisation trend in business, the post-9/11 environment has made a panoramic view of risk both essential and inevitable. Karolyi (2006) provide a review of the literature on terrorism and the financial markets and indicate that there is an additional psychological fear of terrorism on economic behaviour that may turn out to be a permanent. This fear implies economic agents may form expectations based on recently observed events. The adaptive expectations hypothesis suggests economic agents’ expectations adjust by constant proportion of previous discrepancy. The model suggests the expected target level is an exponentially weighted moving average of past observed averages, with recent values weighted more heavily (see Muth, 1960). Furthermore the forecast for the following period is the same for all the subsequent future periods, if the expectation is permanent and not transitory. However, if investors learn ex post that there is an error in their long term expectations of risk-return trade-off following new information in subsequent attacks, an adjustment in the permanent long term expectations would be made. Implicitly, being a turning point in global terrorism, if investors re-adjusted their risk-return expectation based on 9/11 terrorist event, then financial markets should not experience any significant abnormal returns and change in risk following subsequent terrorist events as these would have already been incorporated into investors’ permanent expectations. This paper examines the simple case of adaptive expectations hypothesis in financial markets following terrorist events and investigates the effect of dreadful attacks on the World Trade Centre in New York and four subsequent terrorist attacks (Bali, Madrid, London, and Mumbai) on all relevant Japanese industries. In a recent study, Brounrn and Derwall (2010) investigate the impact of 31 terrorist attacks on various stock markets and industries, including Japan. However, their analysis is limited to six industries (Airlines, Defense, Food, Hotels, Insurance, and Oil & Gas). We contribute to the literature by providing comprehensive analyses of all relevant industries in Japan. In addition, there is a nonnegligible time difference between Japan and the U.S., and the European markets which we take into consideration in our empirical analysis. That is, in our estimated models we explore asynchronicity between these markets by including a feedback effect from the U.S. and the European markets to the Japanese market. The deep economic ties as well as the flow of people between Japan on the one hand, and the Europe, Asia, and the U.S., on the other hand, suggests that terrorism poses a potent threat to the economic interests of Japan (Narayana and Narayan, 2010; Guzel and Ozdemir, 2011). This is evident from the speed with which Japan acted to push through counter-terrorism measures in their legislature following 9/11. Of financial markets, Japan was the first market to open following the initial September 11 and the subsequent four terrorist attacks (Bali, Madrid, London, and Mumbai). For example, unlike the U.S. market that opened six days after the attack, the Japanese market opened the day after the attack. Therefore, Japan offers a unique insight into the immediate reactions of domestic equity investors to international terrorism. Following the adaptive expectations model, we initially hypothesise that those subsequent terrorist events after the September 11, 2001 bombings in the U.S. would not have significant impact on abnormal returns and risk in Japanese equity markets as investors have already incorporated geopolitical risks resulting from terrorism in their expectations around the 9/11 attacks. Nikkinen et al. (2008) show that market response to terrorist attacks in the U.S. differs across regions. We extend this pattern of thought and examine the micro content of the Japanese market, industry portfolios, and further hypothesise differential industry effects of terrorist attacks. We recognise that

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

99

the some industry response may differ from the general market sentiment. In this study, we modify the methodologies used in the existing literature by excluding firm specific information, using regression analysis and non-parametric tests, to reinforce our findings. Chesney et al. (2011) propose non-parametric test to be the most appropriate method for analysing the impact of terrorism on financial markets. Most of the existing literature fails to exclude firm specific information and, thus, report results which contain both the impact of terrorist attacks and other non terrorist components. In earlier studies, Roll (1988) and Nikkinen et al. (2008) suggest that the timing and magnitude of changes in stock returns and volatility varies across markets around the world in response to major events. In response to 9/11 attacks, Chen and Siems (2004), Richman et al. (2005), and Brounrn and Derwall (2010) all document negative reactions of international markets. Richman et al. (2005) further show short and long term effects following the September 11 attack. Studies by Fernandez (2007, 2009), Hammoudeh and Li (2008) and Choi and Hammoudeh (2010) also show that financial markets in the Middle East react negatively to transnational terrorist attacks and geopolitical crises. Other studies have looked into the effects of the terrorist attack on industry portfolios. For instance, Drakos (2004) and Carter and Simkins (2004) investigate the effects of September 11 attacks on a set of airline stocks at various international stock markets and show an immediate impact on the world stock exchanges. Drakos (2004) documents an apparent shift in the riskiness of airlines stocks after the attack through an increase in systematic risk. Carter and Simkins (2004) find a negative reaction of both domestic and international airlines. Ito and Lee (2005), on the other hand, note a substitution effect between domestic and international travel in Japan. They show an upward spike of 6% in the Japanese domestic demand and a dramatic drop of 8.9% in international demand after the September 11 event. Cam (2008) further observes a differential impact on industry portfolios in the U.S. and shows that some industries were positively affected. Ramiah et al. (2010) shows significant negative abnormal returns and increase in systematic risk for the Australian market following September 11 bombing. Brounrn and Derwall (2010) studied 31 terrorist attacks on national and six industry portfolios and demonstrate that the September 11 terrorist event is the only one that had a long term effect on the international equity markets. Consistent with the literature, we find significant negative abnormal returns associated with September 11 for both the Japanese market index and industrial portfolios. The prior literature suggests inconsistent results in terms of changes in short term risk. Drakos (2004) suggests an increase in short term market risk in the airline industry whilst Richman et al. (2005) find the opposite for the market index. We find theoretical consistency with Drakos (2004) in that systematic risk for most of the Japanese industrial portfolios increases immediately after the September 11 attack. For the four terrorist attacks subsequent to September 11, our results generally suggest no impact on abnormal returns on the first trading day immediately after the incidents. In terms of the short term systematic risk, we also, generally, observe no change in most sectors following the terrorist attacks. However we note a step-change adjustment in long term risk following the Bali and Madrid bombings but no effect resulting from the subsequent two attacks in London and Mumbai, implying no the forecast error in risk expectation in Japan after the initial terrorist attacks. The rest of the paper is structured as follows: in Section 2, we present the data and methods used in this study. Section 3 presents the empirical findings and Section 4 concludes. 2. Data and methodology 2.1. Data We use daily stock return indexes, the 3 months treasury bills for Japan and the 10 year bond rate for the United States and Europe, for the period July 1999 to February 2007 for our empirical analysis. Our total sample comprises of 1859 stocks sourced from Datastream. Using the Global Industry Classification Standard, we categorise the individual stocks into 34 industrial groupings. Table 1 reports the descriptive statistics for each industry. It could be seen from Table 1 that the average daily return

100

Table 1 Descriptive statistics of daily returns for the Japanese industrial sectors from July 1999 to February 2007. Mean (%)

Stdev

Skewness

Excess kurt

Min (%)

Max (%)

Count

t-Test statistic

Aerospace Automobiles Banks Beverages Chemical Construction Electricity Electronics Equity and Non-Equity Investment Food Product Food & Drug Forest Gas Oil & Gas General Industrial General Financial Health Care Household Goods Industrial Transportation Leisure Goods Life & Non-Life Insurance Media Metal Personal Goods Pharmaceuticals Real Estate Retailers Software Support Service Technology Telecommunication Travel & Leisure Others All

0.0451 0.0146 0.0101 −0.0469 0.0118 0.0037 0.0477 −0.0266 0.0227 0.0072 −0.0058 0.0056 0.0158 0.0461 0.0077 0.0087 −0.0062 −0.0084 0.0193 −0.0452 −0.1413 −0.1228 0.0388 0.0167 −0.0037 0.0506 −0.0253 −0.1003 −0.0756 −0.0413 −0.0467 −0.041 0.0164 0.012

0.0004 0.0006 0.0004 0.0016 0.0008 0.0005 0.0004 0.001 0.0005 0.0004 0.0009 0.0002 0.0003 0.0005 0.0003 0.0023 0.0009 0.0009 0.0005 0.0016 0.0055 0.0025 0.0005 0.0005 0.0009 0.0014 0.0011 0.0022 0.0021 0.0019 0.001 0.0018 0.0004 0.0014

1.98 −2.91 0.91 −3.01 −4.52 0.23 2.83 −2.5 −0.6 −1.7 −1.97 −0.35 −1.15 1.27 −0.52 3.17 −0.11 −2.43 −1.4 −3.37 −3.17 −2.88 −0.5 0.71 −1.83 −1.5 −1.31 −2.3 −1.84 −0.46 −1.11 −2.66 −0.63 −3.11

3.97 14.46 16.15 9.58 28.38 2.06 8.72 7.88 2.53 6.75 5.64 0.89 1.85 1.57 1.52 17.33 1 9.32 3.68 13.43 10.24 10.02 2.68 2.94 3.66 5.98 3.74 6.78 3.7 10.58 0.86 8.3 −1.07 33.42

0.02 −0.35 −0.16 −0.53 −0.53 −0.13 0.02 −0.52 −0.09 −0.18 −0.36 −0.04 −0.07 −0.01 −0.09 −0.48 −0.24 −0.47 −0.16 −0.9 −1.77 −1.35 −0.12 −0.08 −0.34 −0.58 −0.45 −1.04 −0.83 −0.84 −0.27 −0.84 −0.05 −1.77

0.12 0.14 0.25 0.05 0.17 0.15 0.15 0.11 0.13 0.11 0.17 0.05 0.06 0.18 0.07 1.25 0.25 0.17 0.1 0.14 0.08 0.11 0.15 0.2 0.15 0.36 0.27 0.32 0.25 0.84 0.08 0.19 0.06 1.25

6 75 75 12 75 73 12 75 16 75 75 24 22 16 33 72 68 75 74 68 11 74 74 74 67 75 74 72 75 75 14 72 6 1859

2.73 2 2.01 −1.03 1.26 0.69 4.75 −2.21 1.76 1.53 −0.57 1.51 2.27 3.67 1.33 0.32 −0.57 −0.81 3.64 −2.29 −0.86 −4.28 7.45 3.1 −0.32 3.19 −1.96 −3.95 −3.17 −1.89 −1.81 −1.98 0.95 −3.7

JB-statistic 5 115 21 22 289 1 19 81 1 38 50 1 5 4 2 133 0 77 25 136 23 106 3 7 38 30 22 66 43 7 3 88 0 90,641

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

Return

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

101

Table 2 The five terrorist attacks and their consequences. Terrorist attack

Date

Event

First trading day

Injuries

Fatalities

September 11 – United States

11/09/2001

12/09/2001

5000

3025

Bali bombings – Indonesia

12/10/2002

14/10/2002

300

202

Madrid – Spain

11/03/2004

12/03/2004

1824

191

London – United Kingdom

07/07/2005

08/07/2005

700

52

Mumbai – India

11/07/2006

Four U.S. planes hijacked by terrorists crashed into the World Trade Center, the Pentagon and a field in Pennsylvania killing nearly 3000 people in a matter of hours. Two deadly explosions, one detonated by a suicide bomber and the other one carried out by a large van, ripped through two popular nightclubs, Sari Club and Padi’s Bar, in Kuta on the island of Bali killing 202 people, most of them young Australians, and injured more than 300. Deadly bombings which consisted of a series of coordinated bombings against the commuter train system left 191 dead and 1824 injured in Madrid, Spain. At the peak of the rush hour, bombs were detonated in three crowded subway trains and aboard a London bus which killed at least 52 people with 700 injured. A series of seven bombs that took place over a period of 11 min on the suburban railways in Mumbai with 209 people losing their lives and over 714 injured in the attacks.

11/07/2006

714

209

Source: Adapted and adjusted from Cam (2008) and various media releases.

for ten sectors are significantly negative whilst eleven sectors show significantly positive returns for the period under consideration. Table 1 also includes the standard deviation, skewness, excess kurtosis, and the number of firms in each of the industry sectors. Details of the five terrorist attacks that occurred in the United States, Bali, Madrid, London and Mumbai, and the first trading day subsequent to the attacks are summarised in Table 2. 2.2. Methodology We employ event studies methodology in our empirical analysis and define daily return as: DRit = ln

 SRI  it SRIit−1

(1)

where DRit is the daily return for stock i, SRIit is the stock return index for stock i at time t, and SRIit−1 is the stock return index for stock i at time t − 1. Following Brown and Warner (1985), the ex post abnormal returns (ARit ) are calculated as the difference between observed returns of firm i at event day t, and the expected return, E(Rit ): ARit = DRit − E(Rit )

(2)

The expected return, E(Rit ), is estimated using a market model over the last 260 observed daily returns and Rmt is the market return: E(Rit ) = ˇ0 + ˇ1 Rmt

(3)

In this model, the expected return is a weighted moving average of past observed return over 260 days. The forecast changes if investors learn ex post that there is an error in their long term expectations

102

Table 3 Abnormal returns on Japanese industry indices following five terrorist attacks. Industry

Bali

Madrid

London

Mumbai

AR (%)

t-Stat

AR (%)

t-Stat

AR (%)

t-Stat

AR (%)

t-Stat

AR (%)

t-Stat

0.80 −6.10** −2.25** −6.20** −6.21** −4.65** −3.70** −5.57** −6.42** −9.67** −3.48** −3.39** −4.41** −2.71** −1.27 −6.38** −6.00 −5.53** −6.07** −4.24** −7.16** −8.38** −4.68** −7.36** −7.08** −6.88** −8.72** −4.91** −8.05** −6.38** −7.92** −6.56** −3.90** −4.01** −5.57**

0.55 −7.13 −2.67 −6.81 −6.99 −6.24 −3.91 −4.56 −6.38 −6.01 −6.69 −5.91 −4.76 −3.47 −0.78 −6.48 −0.48 −6.01 −8.37 −5.6 −5.81 −5.97 −3.39 −6.49 −7.69 −7.12 −6.38 −5.66 −4.33 −6.26 −4.95 −2.51 −5.81 −2.46 −6.82

0.10 0.02 0.05 0.11 0.05 0.16 0.07 0.09 0.09 0.06 0.09 0.03 0.05 0.19 0.40 0.07 0.19 0.11 0.05 0.04 0.07 0.10 0.18 0.12 0.10 0.05 0.38 0.04 0.26 0.16 0.13 0.17 0.30 0.27 0.11

0.05 0.03 0.04 0.15 0.06 0.2 0.08 0.09 0.10 0.04 0.19 0.06 0.06 0.27 0.31 0.06 0.15 0.15 0.07 0.06 0.06 0.06 0.17 0.10 0.11 0.06 0.35 0.05 0.21 0.20 0.10 0.08 0.53 0.15 0.15

0.45 −0.27 −1.68* −0.46 −0.38 −0.09 −0.49 −0.32 −0.18 −1.15 0.16 −0.26 −0.80 −0.71 −1.24 −0.46 −1.24 0.68 2.50 −0.87 −0.55 −2.77 0.03 −0.84 −0.24 −0.67 0.89 −0.24 −0.08 −0.67 −1.04 −1.20 −0.12 −2.04 −0.42

0.33 −0.37 −1.93 −0.78 −0.52 −0.13 −0.90 −0.30 −0.21 −0.97 0.36 −0.63 −1.17 −1.39 −1.19 −0.55 −0.82 0.91 0.37 −1.26 −0.54 −1.57 0.02 −0.64 −0.30 −0.98 0.76 −0.36 −0.05 −0.71 −0.83 −0.52 −0.21 −0.91 −0.59

−1.47 −0.11 0.32 −0.47 0 −0.23 −0.13 0.36 −0.32 −0.02 0.18 −0.02 −0.23 0.10 −0.78 0.49 0.11 0.38 −0.08 −0.15 −0.37 −0.26 −0.32 −0.10 −0.43 0.06 0.10 0.06 −1.10 −0.48 −0.03 −0.59 0.03 0.12 −0.10

−1.37 −0.19 0.45 −0.86 0 −0.39 −0.26 0.52 −0.48 −0.03 0.42 −0.05 −0.4 0.22 −0.77 0.72 0.12 0.56 −0.15 −0.23 −0.49 −0.21 −0.37 −0.10 −0.64 0.09 0.11 0.09 −1.00 −0.55 −0.04 −0.38 0.04 0.08 −0.16

−1.27 −0.53 −0.54 −0.65 −0.58 −0.59 −0.77 −0.76 −1.20 −0.32 −0.69 −0.56 −0.25 −0.40 −1.09 −0.60 −0.95 0.21 −0.12 −0.44 −1.19 −0.98 −1.49 −1.67 −0.63 −0.67 −1.88 −0.64 −2.57 −0.99 −1.14 −2.39 −0.42 −0.17 −0.85

−0.70 −0.52 −0.39 −0.88 −0.54 −0.56 −0.80 −0.61 −0.89 −0.29 −1.04 −0.75 −0.28 −0.51 −0.75 −0.56 −0.51 0.21 −0.11 −0.47 −0.90 −0.49 −0.88 −1.04 −0.54 −0.67 −1.21 −0.53 −1.25 −0.61 −0.77 −1.11 −0.41 −0.08 −0.74

This table presents abnormal returns and the parametric t-test results for 34 Japanese industries after September 11, Bali, Madrid, London and Mumbai terrorist attacks. * Significance at 10% level. **

Significance at 5% level.

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

Aerospace Automobile Banks Beverages Chemicals Construction Electricity Electronics Engineering Equity & Non-Equity Investment Food Producers Food & Drug Forest Gas Gas and Oil General Industrial General Financial Health Care Household Goods Industrial Transportation Leisure Goods Life & Non-Life Insurance Media Metal Personal Goods Pharmaceuticals Real Estate Retailers Software Support Service Technology Telecommunication Travel & Leisure Others All

11-Sep

Table 4 Cumulative abnormal returns for five days on Japanese industry indices following five terrorist attacks. Industry

11-Sep CAR5 (%) 6.27 −8.96** −2.08 −3.86* −4.98** −3.58* −1.04 −2.29 −5.40** −2.71 −2.98** −2.33 −3.12 −2.65 2.10 −4.12 −4.74 −1.85 −4.66** −5.06** −8.84** −10.31** −4.44 −5.16* −4.34* −3.36 −6.16* −4.25* −6.99 −6.00** −6.80 3.85 −5.33** −1.54 −4.48**

Madrid

London

Mumbai

CAR5 (%)

t-Stat

CAR5 (%)

t-Stat

CAR5 (%)

t-Stat

CAR5 (%)

t-Stat

2.09 −3.94 −1.14 −1.81 −2.10 −1.72 −0.52 −0.72 −2.08 −0.86 −2.32 −1.51 −1.39 −1.73 0.53 −1.57 −1.32 −0.72 −2.36 −2.51 −2.83 −3.21 −1.19 −1.68 −1.83 −1.41 −1.81 −1.71 −1.34 −2.20 −1.62 0.57 −3.13 −0.41 −2.03

4.24 4.89** 3.00 2.72 2.97 1.82 1.31 5.34* 4.11* 5.78 2.86** 0.80 0.78 0.24 5.56* 3.19 5.93* 4.38** 2.81 2.00 5.16 1.70 6.20** 4.77 3.58 4.99** 5.76* 3.88** 7.10** 4.59* 6.61* 10.89** 2.51* 5.13 3.94*

1.05 2.20 1.23 1.57 1.35 0.89 0.69 1.92 1.72 1.26 2.04 0.64 0.37 0.15 1.88 1.25 1.77 1.98 1.44 1.12 1.78 0.48 2.24 1.26 1.40 2.30 1.85 2.09 1.99 1.79 1.81 2.21 1.75 1.10 1.90

5.98 0.07 2.29 1.45 1.75 4.65** 0.52 1.27 2.74 1.43 1.83 2.24** 0.90 1.75 1.63 1.44 4.88 1.25 3.72** 2.69 0.20 4.17 0.12 2.26 2.64 −0.14 3.32 3.27 1.06 0.08 −0.27 −0.94 1.18 2.52 1.90

1.55 0.04 1.11 1.05 0.83 2.20 0.43 0.42 1.11 0.46 1.57 2.22 0.54 1.55 0.71 0.66 1.19 0.91 2.02 1.39 0.07 1.10 0.03 0.61 1.07 −0.08 1.02 1.76 0.22 0.03 −0.07 −0.16 0.72 0.49 0.91

−0.79 1.10 −0.24 0.34 0.49 −0.44 0.04 2.72 0.59 0.86 0.70 0.98 0.30 −0.15 −1.43 1.02 0.66 1.38 0.20 0.58 1.18 −0.13 1.56 −0.34 1.24 0.89 2.72 1.75 1.56 0.44 1.34 3.39 1.12 −0.32 0.87

−0.31 0.76 −0.15 0.25 0.32 −0.28 0.04 1.48 0.35 0.55 0.61 0.91 0.21 −1.39 −0.56 0.57 0.30 0.79 0.14 0.34 0.64 −0.05 0.67 −0.13 0.68 0.54 1.09 0.97 0.53 0.19 0.65 0.91 0.72 −0.09 0.57

−3.62 −3.89 −4.37 −3.31** −4.50* −4.61* −4.20** −5.51* −5.32 −3.99 −2.46 −2.41 −2.55 0.26 0.26 −4.32* −6.03 −3.45 −3.02 −3.69 −5.66* −7.36* −5.65 −5.66 −4.33 −2.89 −5.42 −4.97 −6.58 −4.91 −6.12 −6.53 −3.63 −4.31 −4.46

−0.88 −1.47 −1.28 −2.08 −1.65 −1.75 −2.07 −1.81 −1.60 −1.53 −1.56 −1.26 −1.22 −1.35 0.08 −1.66 −1.38 −1.36 −1.16 −1.58 −1.74 −1.70 −1.36 −1.44 −1.49 −1.27 −1.46 −1.63 −1.29 −1.21 −1.64 −1.29 −1.40 −0.90 −1.58

**

Significance at 5% level.

103

This table presents five day cumulative abnormal returns and the parametric t-test results for 34 Japanese industries after September 11, Bali, Madrid, London and Mumbai terrorist attacks. * Significance at 10% level.

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

Aerospace Automobile Banks Beverages Chemicals Construction Electricity Electronics Engineering Equity & Non-Equity Investment Food Producers Food & Drug Forest Gas Gas and Oil General Industrial General Financial Health Care Household Goods Industrial Transportation Leisure Goods Life & Non-Life Insurance Media Metal Personal Goods Pharmaceuticals Real Estate Retailers Software Support Service Technology Telecommunication Travel & Leisure Others All

**

Bali t-Stat

104

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

Others (Mining, Tobacco, Oil equipment) Travel & Leisure Telecommunication Technology Support Service Software Retailers Real Estate Pharmaceutical Personal Goods Metal Media Life & Non-Life Insurance Leisure Good Industrial Transportation Household Goods Healthcare General Financial General Gas and Oil Gas Forest Food & Drug Food Producers Equity & Non-Equity Investment Engineering Electronic Electricity Construction Chemicals Beverages Banks Automobile Aerospace

-0.15000

-0.10000

-0.05000

0.00000

0.05000

0.10000

A R and CAR5 AR

CAR5

Fig. 1. AR and CAR5 on Japanese industry indices following September 11.

of risk-return trade-off following new information in subsequent attacks. Following Karolyi (2006), we assume that this change is permanent and the expected persistence of this change is incorporated e in our estimation by adjusting the returns by a constant of the previous discrepancy, t+1 = e + e ), 0 ≤  ≤ 1, where  is the series of returns being predicted and  e is the expectation of (t − t+1 t t t formed at t − 1. The abnormal return for industry I at time t, ARIt , is obtained by averaging the abnormal return of each form within the industry: 1 ARit N N

ARIt =

i=1

(4)

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

105

Table 5 The impact of five terrorist attacks on Japanese industry indices – non-parametric results. Industry

11-Sep

Bali

Madrid

London

Mumbai

Aerospace Automobile Banks Beverages Chemicals Construction Electricity Electronics Engineering Equity & Non-Equity Investment Food Producers Food & Drug Forest Gas Gas and Oil General Industrial General Financial Health Care Household Goods Industrial Transportation Leisure Goods Life & Non-Life Insurance Media Metal Personal Goods Pharmaceuticals Real Estate Retailers Software Support Service Technology Telecommunication Travel & Leisure Others All

−0.74 −3.92** −2.22** −3.89** −3.98** −4.31** −2.27** −3.20** −4.12** −1.69* −4.34** −4.39** −2.72** −3.02** −0.87 −4.06** −3.05** −3.34** −4.21** −3.41** −3.64** −2.69** −3.22** −3.80** −4.75** −4.51** −3.86** −4.48** −3.27** −4.06** −3.46** −1.98** −4.60** −2.06** −4.41**

0.57 −0.14 0.02 0.50 0.30 0.24 0.37 0.16 0.56 −0.11 0.91 −0.24 0.48 0.51 −0.08 −0.46 0.81 0.43 0.01 0.55 0.21 −0.20 0.93 0.41 0.59 0.21 0.24 0.40 0.51 1.10 0.19 0.20 0.36 0.70 0.40

−0.13 −1.22 −2.09** −1.48 −1.16 −0.84 −0.97 −0.94 −1.28 −1.64 −0.55 −1.68* −1.82* −2.48** −1.87* −1.10 −1.48 −1.35 −0.82 −1.87 −0.97 −2.07** −1.24 −1.38 −1.28 −1.70 −0.47 −1.46 −0.94 −2.06** −1.41 −1.09 −1.03 −2.12** −1.63

−1.07 −0.52 0.80 −0.65 −0.06 −0.01 −0.29 0.12 −0.34 −0.10 0.82 0.36 −0.25 0.46 −1.42 0.68 0.22 −0.08 −0.20 0.55 −1.09 −0.36 −0.79 −0.05 −0.68 0.15 −0.51 −0.22 −0.76 −0.87 −0.51 −0.69 −0.71 0.45 −0.24

−0.86 −0.51 −0.20 −1.09 −0.49 −0.29 −0.89 −0.89 −1.10 −0.57 −0.76 −0.47 0.22 −0.59 −0.50 −0.65 −0.69 0.10 −0.05 −0.34 −0.85 −0.42 −0.89 −1.01 −0.35 −0.15 −1.76 −0.94 −1.41 −0.94 −1.15 −1.55 −0.67 −0.20 −0.76

This table presents the non-parametric t-test results for 34 Japanese industries after September 11, Bali, Madrid, London and Mumbai terrorist attacks. * Significance at 10% level. ** Significance at 5% level.

2.3. Parametric tests The parametric tests used in this study rely on the important assumption that the industry abnormal returns and cumulative abnormal returns are normally distributed. The standard t-statistic for the abnormal return is given by: tARIt =

ARIt SD(ARIt )

(5)

where SD(ARIt ) is an estimate of the standard deviation of the abnormal returns. By cumulating the periodic abnormal return for each industry over five days, we obtain the five day cumulative abnormal return, CAR5It : CAR5It =

5  t=1

ARIt

(6)

106

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

Table 6 The short run impact of September 11 attacks on the systematic risk of Japanese industries. Industry

r˜It − r˜ft = I + ˇI1 [˜rIt − r˜ft ] + ˇI2 [˜rmt − r˜ft ] × D + ε˜ it I

Aerospace t-Statistics Automobile t-Statistics Banks t-Statistics Beverages t-Statistics Chemicals t-Statistics Construction t-Statistics Electricity t-Statistics Electronics t-Statistics Engineering t-Statistics Equity and Non-Equity Investment t-Statistics Food & Drug t-Statistics Food Producers t-Statistics Forest t-Statistics Gas t-Statistics Gas & Oil t-Statistics General Industrial t-Statistics General Financial t-Statistics Health Care t-Statistics Household Goods t-Statistics Industry Transportation t-Statistics Leisure Goods t-Statistics Life & Non-Life Insurance t-Statistics Media t-Statistics Metal t-Statistics Personal Goods t-Statistics Pharmaceuticals t-Statistics Real Estate t-Statistics Retailers t-Statistics Software t-Statistics Support Service t-Statistics

−0.05 −31.27 −0.05 −41.84 −0.05 −39.43 −0.05 −39.45 −0.05 −40.53 −0.05 −41.35 −0.05 −38.08 −0.05 −34.69 −0.05 −37.97 −0.05 −29.76 −0.05 −37.97 −0.05 −44.74 −0.05 −39.27 −0.05 −40.68 −0.05 −30.87 −0.05 −39.69 −0.05 −32.28 −0.05 −38.99 −0.05 −42.00 −0.05 −41.43 −0.05 −35.54 −0.05 −33.43 −0.05 −31.38 −0.05 −37.13 −0.05 −39.39 −0.05 −38.67 −0.05 −32.28 −0.05 −39.43 −0.05 −26.58 −0.05 −38.25

ˇI1 0.08 1.17 0.15 3.13 0.13 2.51 0.12 2.25 0.15 2.91 0.14 2.86 0.12 2.26 0.17 2.86 0.14 2.52 0.20 2.89 0.14 2.52 0.16 3.50 0.16 2.92 0.14 2.83 0.14 2.16 0.15 2.79 0.19 3.02 0.15 2.88 0.16 3.25 0.13 2.61 0.15 2.55 0.11 1.81 0.20 3.16 0.19 3.40 0.15 2.96 0.16 2.95 0.22 3.51 0.13 2.49 0.22 2.84 0.10 1.90

ˇI2 −0.77 −1.15 1.48** 3.02 0.21 0.40 1.55** 2.96 1.51** 2.97 1.01** 2.03 0.71 1.31 1.27** 2.11 1.58** 2.89 2.65** 3.91 1.58** 2.89 0.62 1.36 0.91* 1.74 0.36 0.72 −0.12 −0.17 1.58** 3.05 1.45** 2.33 1.31** 2.50 1.48** 3.05 0.89* 1.79 1.84** 3.18 2.28** 3.72 1.00 1.57 1.87** 3.40 1.84** 3.58 1.73** 3.26 2.32** 3.74 1.13** 2.19 2.10** 2.78 1.63** 3.04

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

107

Table 6 (Continued) r˜It − r˜ft = I + ˇI1 [˜rIt − r˜ft ] + ˇI2 [˜rmt − r˜ft ] × D + ε˜ it

Industry

Technology t-Statistics Telecommunication t-Statistics Travel & Leisure t-Statistics Others t-Statistics All t-Statistics

I

ˇI1

ˇI2

−0.05 −30.32 −0.05 −20.65 −0.05 −41.43 −0.05 −29.62 −0.01 −10.28

0.15 2.18 0.21 2.15 0.13 2.61 0.17 2.42 −0.02 −0.30

2.08** 3.04 1.63** 1.69 0.89* 1.79 0.79 1.15 −0.36 −0.71

This table presents the regression analysis results for 34 Japanese industries after September 11 terrorist attack (see Eq. (12)). The multiplicative dummy variable equation illustrates the impact on systematic risk. * Significance at 10% level. ** Significance at 5% level.

The t-statistic for the five day cumulative abnormal return is obtained by dividing CAR5it by the standard deviation of the five day cumulative abnormal return, SD(CAR5It ): tCAR5Iit =

CAR5It SD(CAR5It )

(7)

2.4. Non-parametric tests Corrado and Truong (2008) argue in favour of robustness test in Asia-Pacific event studies as abnormal returns are not normally distributed. Specifically, the distribution of the abnormal returns tends to exhibit fat tails and positive skewness. Under these circumstances, parametric tests tend to reject the null too often when testing for positive abnormal performance and too seldom when testing for negative abnormal returns. As a robustness test, we turn to an alternative non-parametric test developed by Corrado (1989). Chesney et al. (2011) suggest that non-parametric test is the most appropriate method for analysing the impact of terrorism on financial markets. This test is more powerful at detecting the false null hypothesis of no abnormal returns. We transform each firm’s abnormal returns, ARit into ranks, Ki , over the combined period, Ti , of 260 days. This is denoted as: Ki = rank(ARit )

(8)

The period is broken up into the 244 days prior to the event, the event day and 15 days after the event. The ranks in the event period for each firm are then compared with the expected average rank, K¯ i , under the null hypothesis of no abnormal returns. This is given by: K¯ i = 0.5 +

Ti 2

(9)

The non-parametric t-statistic, tnp , for the null hypothesis of no abnormal returns for each industry is given by: tnp =

(1/N)

N i=1

(Ki − K¯ i )

¯ SD(K)

(10)

¯ is the standard deviation of the average rank, and is denoted by: where SD(K)

  T 1 1  2 ¯ = (Kit − K¯ i ) SD(K) 2 T

t=1

N

(11)

108

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

Table 7 The long run impact of September 11 attacks on the systematic risk of Japanese industry indices. Industry

Aerospace t-Statistics Automobile t-Statistics Banks t-Statistics Beverages t-Statistics Chemicals t-Statistics Construction t-Statistics Electricity t-Statistics Electronics t-Statistics Engineering t-Statistics Equity and Non-Equity Investment t-Statistics Food & Drug t-Statistics Food Products t-Statistics Forest t-Statistics Gas t-Statistics Gas & Oil t-Statistics General Industrial t-Statistics General Finance t-Statistics Health Care t-Statistics Household Goods t-Statistics Industry Transportation t-Statistics Leisure Goods t-Statistics Life & Non-Life Insurance t-Statistics Media t-Statistics Metal t-Statistics Personal Goods t-Statistics Pharmaceuticals t-Statistics Real Estate t-Statistics Retailer t-Statistics Software t-Statistics Support Service

r˜It − r˜ft = ı0 + ı1I [˜rmt − r˜ft ] + ı2I [˜rmt − r˜ft ] × SD + ı3I (SD) ı0

ı1

ı2

ı3

−0.05 −32.74 −0.05 −42.8 −0.05 −42.31 −0.05 −41.48 −0.05 −42.24 −0.05 −44.14 −0.05 −39.86 −0.05 −36.31 −0.05 −39.63 −0.05 −31.23 −0.05 −39.63 −0.05 −47.43 −0.05 −40.73 −0.05 −42.46 −0.05 −31.6 −0.05 −41.13 −0.05 −33.29 −0.05 −41.4 −0.05 −44.37 −0.05 −43.71 −0.05 −36.5 −0.05 −34.03 −0.05 −33.14 −0.05 −39.17 −0.05 −41.11 −17.14 −40.98 −0.05 −32.58 −0.05 −41.17 −0.05 −27.41 −0.05

0.03 0.44 0.10 2.01 0.07 1.31 0.06 1.06 0.09 1.68 0.08 1.61 0.08 1.5 0.11 1.87 0.07 1.27 0.10 1.41 0.07 1.27 0.11 2.51 0.10 1.86 0.10 1.91 0.08 1.15 0.08 1.57 0.12 1.93 0.10 1.83 0.09 1.82 0.07 1.33 0.08 1.30 0.03 0.51 0.12 1.90 0.11 1.90 0.08 1.60 −2.10 −2.16 0.15 2.34 0.07 1.38 0.13 1.72 0.04

−0.19 −0.74 0.33* 1.72 0.22 1.11 0.32 1.64 0.38* 1.95 0.30 1.64 −0.03 −0.15 0.20 0.89 0.47 2.25** 0.91** 3.49 0.47** 2.25 0.13 0.79 0.29 1.44 0.11 0.59 0.35 1.37 0.43** 2.21 0.45 1.87 0.20 1.04 0.52** 2.85 0.40** 2.15 0.59** 2.64 0.77** 3.24 0.54** 2.27 0.69** 3.33 0.52** 2.69 12.02** 3.31 0.60** 2.46 0.30 1.56 0.64** 2.19 0.39*

0.03** 5.52 0.02** 5.45 0.03** 6.67 0.03** 6.39 0.02** 6.05 0.02** 6.74 0.02** 5.84 0.03** 5.81 0.02** 5.89 0.03** 5.56 0.02** 5.89 0.02** 6.54 0.02** 5.51 0.02** 5.74 0.02** 4.38 0.02** 5.77 0.02** 4.99 0.03** 6.63 0.02** 6.45 0.02** 6.15 0.02** 5.10 0.02** 4.60 0.03** 5.57 0.03** 6.12 0.02** 6.14 0.58** 8.26 0.02** 4.51 0.02** 5.88 0.03** 4.62 0.02**

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

109

Table 7 (Continued) Industry

t-Statistics Technology t-Statistics Telecommunication t-Statistics Travel & Leisure t-Statistics Others t-Statistics All t-Statistics

r˜It − r˜ft = ı0 + ı1I [˜rmt − r˜ft ] + ı2I [˜rmt − r˜ft ] × SD + ı3I (SD) ı0

ı1

−39.01 −0.05 −31.23 −0.05 −21.46 −0.05 −44.64 −0.05 −31.01 −0.01 −10.78

0.82 0.07 1.06 0.11 1.14 0.05 1.10 0.09 1.25 −0.05 −0.94

ı2 1.89 0.55** 2.10 0.69* 1.87 0.33* 1.80 0.52** 2.00 0.14 0.71

ı3 5.14 0.03** 5.00 0.03** 4.11 0.02** 5.79 0.03** 5.10 0.01** 2.99

This table presents the regression analysis results for 34 Japanese industries after September 11 terrorist attack. The first multiplicative dummy variable equation illustrates the impact on systematic risk and the second additive dummy variable equation shows the impact on the intercept (see Eq. (14)). * Significance at 10% level. ** Significance at 5% level.

2.5. Regression analysis Using the CAPM, we test if terrorist attacks have had an impact on the systematic risk of Japanese industries on the days of the attack. We include a multiplicative dummy variable in the standard CAPM to test this possibility. The model we estimate is therefore: r˜It − r˜ft = I + ˇI1 [˜rmt − r˜ft ] + ˇI2 [˜rmt − r˜ft ] × D + ε˜ it

(12)

where r˜It is industry I’s return at time t, r˜ft is risk free return at time t, r˜mt is return on the market at time t and D is a dummy variable that takes the value of 1 on the day of the event, and 0 otherwise. This variable captures the effect of terrorist attacks on the systematic risk. The inclusion of an additive dummy variable in Eq. (12), results in a near singular variance-covariance matrix. As a result, we estimate a separate equation to test if the intercept was affected by the attacks: r˜It − r˜ft = ϕI + ˛1I [˜rmt − r˜ft ] + ˛2I D + ε˜ it

(13)

We use the returns for each industry 244 days prior to the event, and 15 days after the event. Standard tests and residual diagnostics revealed no major concerns with the two econometric models in Eqs. (12) and (13). Using a Wald test, we check whether the dummy variables are redundant. To deal with the problem of exact multicollinearity in the estimation of Eq. (12) with an additive dummy variable that we specified above, we construct a variation of Eq. (12). This gives rise to the following Eq. (12.1): r˜It − r˜ft = I + ˇI1 [˜rmt − r˜ft ] +

15 

2 ˇI,j [˜rmt − r˜ft ] × WDj +

j=2

15 

3 ˇI,j WDj + ε˜ it

(12.1)

j=2

where WDj represents the a window dummy variable starting on the first day of trading1 (j = 2 shows a window of 2 days, 3 for a 3-day window, . . ., 15 for a 15-day window). We use Eq. (12.1) as a robustness check. Further, we consider the long term impact of the terrorist events on the market. The test determines whether the level of risk, specifically captured by structural changes, is altered after the event day: r˜it = ϕI + ı1I [˜rmt − r˜ft ] + ı2I [˜rmt − r˜ft ] × SD + ı3I SD + ε˜ it

1

Similar to Nanda and Hammoudeh (2007).

(14)

110

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

Table 8 The short run impact of Bali bombings on the systematic risk of Japanese industries. Industry

Aerospace t-Statistics Automobile t-Statistics Banks t-Statistics Beverages t-Statistics Chemicals t-Statistics Construction t-Statistics Electricity t-Statistics Electronics t-Statistics Engineering t-Statistics Equity and Non-Equity Investment t-Statistics Food & Drug t-Statistics Food Producers t-Statistics Forest t-Statistics Gas t-Statistics Gas & Oil t-Statistics General Industrial t-Statistics General Financial t-Statistics Health Care t-Statistics Household Goods t-Statistics Industry Transportation t-Statistics Leisure Goods t-Statistics Life & Non-Life Insurance t-Statistics Media t-Statistics Metal t-Statistics Personal Goods t-Statistics Pharmaceuticals t-Statistics Real Estate t-Statistics Retailers

r˜It − r˜ft = I + ˇI1 [˜rmt − r˜ft ] + ˇI2 [˜rmt − r˜ft ] × D + ε˜ it

r˜It − r˜ft = ϕI + ˛1I [˜rmt − r˜ft ] + ˛2I D + ε˜ it



ˇI1

ˇI2

ϕ

˛1I

˛2I

−0.01 −7.75 −0.01 −17.94 −0.01 −11.06 −0.01 −16.91 −0.01 −13.69 −0.01 −14.3 −0.01 −16.63 −0.01 −11.75 −0.01 −12.62 −0.01 −12.48 −0.01 −22.23 −0.01 −20.23 −0.01 −15.23 −0.01 −7.96 −0.01 −18.26 −0.01 −10.36 −0.01 −12.67 −0.01 −12.74 −0.01 −15.96 −0.01 −13.91 −0.01 −11.13 −0.01 −7.37 −0.01 −9.39 −0.01 −8.29 −0.01 −13.69 −0.01 −13.39 −0.01 −10.01 −0.01

0.10 1.28 0.05 1.93 0.08 1.49 −0.01 −0.13 0.02 0.53 0.02 0.41 0.07 1.77 0.08 1.51 −0.01 −0.12 0.07 1.38 0.01 0.44 0.02 0.70 0.02 0.43 0.11 1.45 0.05 1.59 0.03 0.52 0.01 0.19 0.00 0.09 0.01 0.33 0.00 −0.05 0.04 0.76 0.09 0.94 0.02 0.24 0.05 0.66 0.01 0.24 0.01 0.18 0.02 0.26 −0.03

2.41 1.54 1.89** 6.89 −0.20 −0.19 0.93 1.14 0.29 0.32 0.63 0.73 0.48 0.67 −0.26 −0.25 0.80 0.82 0.31 0.32 0.29 0.48 −0.01 −0.02 0.59 0.69 1.39 0.93 0.14 0.20 0.07 0.06 −0.42 −0.42 −0.25 −0.25 0.41 0.52 0.49 0.51 0.83 0.76 0.54 0.29 0.79 0.60 0.43 0.29 0.87 0.90 0.18 0.19 0.08 0.06 0.18

−0.01 −7.75 −0.01 −17.94 −0.01 −11.06 −0.01 −16.91 −0.01 −13.69 −0.01 −14.3 −0.01 −16.63 −0.01 −11.75 −0.01 −12.62 −0.01 −12.48 −0.01 −22.23 −0.01 −20.23 −0.01 −15.23 −0.01 −7.96 −0.01 −18.26 −0.01 −10.36 −0.01 −12.67 −0.01 −12.74 −0.01 −15.96 −0.01 −13.91 −0.01 −11.13 −0.01 −7.37 −0.01 −9.39 −0.01 −8.29 −0.01 −13.69 −0.01 −13.39 −0.01 −10.01 −0.01

0.10 1.28 0.05 1.93 0.08 1.49 −0.01 −0.13 0.02 0.53 0.02 0.41 0.07 1.77 0.08 1.51 −0.01 −0.12 0.07 1.38 0.01 0.44 0.02 0.70 0.02 0.43 0.11 1.45 0.05 1.59 0.03 0.52 0.01 0.19 0.00 0.09 0.01 0.33 0.00 −0.05 0.04 0.76 0.09 0.94 0.02 0.24 0.05 0.66 0.01 0.24 0.01 0.18 0.02 0.26 −0.03

−0.02 −1.54 −0.06** −6.89 0.00 0.19 −0.01 −1.14 0.00 −0.32 0.00 −0.73 0.00 −0.67 0.00 0.25 −0.01 −0.82 0.00 −0.32 0.00 −0.48 0.00 0.02 0.00 −0.69 −0.01 −0.93 0.00 −0.20 0.00 −0.06 0.00 0.42 0.00 0.25 0.00 −0.52 0.00 −0.51 −0.01 −0.76 0.00 −0.29 −0.01 −0.60 0.00 −0.29 −0.01 −0.90 0.00 −0.19 0.00 −0.06 0.00

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

111

Table 8 (Continued) Industry

t-Statistics Software t-Statistics Support Service t-Statistics Technology t-Statistics Telecommunication t-Statistics Travel & Leisure t-Statistics Others t-Statistics All t-Statistics

r˜It − r˜ft = I + ˇI1 [˜rmt − r˜ft ] + ˇI2 [˜rmt − r˜ft ] × D + ε˜ it

r˜It − r˜ft = ϕI + ˛1I [˜rmt − r˜ft ] + ˛2I D + ε˜ it



ˇI1

ϕ

˛1I

˛2I

−14.09 −0.01 −8.01 −0.01 −9.72 −0.01 −10.81 −0.01 −5.16 −0.01 −16.45 −0.01 −5.26 0.00 −1.02

−0.58 0.00 0.03 0.04 0.58 0.03 0.44 0.03 0.27 −0.02 −0.40 0.13 1.11 0.34 1.48

−14.09 −0.01 −8.01 −0.01 −9.72 −0.01 −10.81 −0.01 −5.16 −0.01 −16.45 −0.01 −5.26 0.00 −1.02

−0.58 0.00 0.03 0.04 0.58 0.03 0.44 0.03 0.27 −0.02 −0.40 0.13 1.11 0.34 1.48

−0.19 −0.01 −1.17 −0.01 −0.77 0.00 −0.26 −0.01 −0.51 0.00 −0.23 0.00 −0.03 0.00 −0.13

ˇI2 0.19 1.90 1.17 0.99 0.77 0.31 0.26 1.19 0.51 0.19 0.23 0.06 0.03 0.60 0.13

This table presents the regression analysis results for 34 Japanese industries after Bali bombings (see Eqs. (12) and (13)). The first multiplicative dummy variable equation illustrates the impact on systematic risk and the second additive dummy variable equation shows the impact on the intercept. *Significance at 10% level. ** Significance at 5% level.

where SD is a structural dummy variable that takes the value of 0 prior to the event, and 1 after the day of the event. This variable captures the structural changes and influence of terrorist attacks on the systematic risk over a long term horizon. To control for the natural time zone difference-asynchronicity between Japan and the U.S., and the European markets, Eq. (12) is modified. In our effort to capture the effects of the US market and the European markets, our existing asset pricing models are fitted with two interactive terms namely the two market risk premia from these two regions. For example the adjusted equation (12) takes the following form: Europe

r˜It − r˜ft = I + ˇI1 [˜rmt − r˜ft ] + ˇI2 [˜rmt − r˜ft ] × D + ˇI3 [˜rmt

Europe

− r˜ft

US ] + ˇI4 [˜rmt − r˜ftUS ] + ε˜ it (12.2)

We run Eq. (12.2) for all the sectors to check the robustness of our original findings. Similarly, Eqs. (13) and (14) are also fitted with these two risk premia. 3. Empirical findings We report the results of the impact of the five different terrorist attacks detailed in Table 2 on the Japanese Stock Exchange in this section. Using parametric and non-parametric tests we investigate whether the returns and systematic risk of 34 Japanese industries were affected by these five events. The results of the study support strong negative impact on returns for most industries and a general increase in systematic risk of some industries on the first day of trading following the U.S. September 11 attacks. We further document a general increase in the long term systematic risk for some Japanese industrial sectors. Generally, these suggest a new development, a turning point, in terrorist risk perception characterised by an international outreach and an immediate effect on markets physically distant from the terrorist targets. Interestingly, we do not find similar evidence for the subsequent four attacks in Bali, Madrid, London, and Mumbai. We generally interpret that to mean investors have already incorporated geopolitical terrorist risks in their future short-term expectations around the 9/11 event. However, the market risk increases in the long term for September 11, Bali bombings and Madrid bombings. This suggests a discovery ex post of a significant forecast error in the subsequent terrorist attacks leading

112

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

Table 9 The long run impact of Bali bombings on the systematic risk of Japanese industry indices. Industry

Aerospace t-Statistics Automobile t-Statistics Banks t-Statistics Beverages t-Statistics Chemicals t-Statistics Construction t-Statistics Electricity t-Statistics Electronics t-Statistics Engineering t-Statistics E&N Investment t-Statistics Food & Drug t-Statistics Food Products t-Statistics Forest t-Statistics Gas t-Statistics Gas & Oil t-Statistics General Industrial t-Statistics General Finance t-Statistics Health Care t-Statistics Household Goods t-Statistics Industry Transportation t-Statistics Leisure Goods t-Statistics Life & Non-Life Insurance t-Statistics Media t-Statistics Metal t-Statistics Personal Goods t-Statistics Pharmaceuticals t-Statistics Real Estate t-Statistics Retailer t-Statistics Software t-Statistics Support Service

r˜It − r˜ft = ı0 + ı1I [˜rmt − r˜ft ] + ı2I [˜rmt − r˜ft ] × SD + ı3I (SD) ı0

ı1I

−0.01 −5.61 0.00 −2.31 0.00 −6.21 −0.01 −9.82 0.00 −7.73 0.00 0.00 −0.01 −10.54 0.00 −6.14 −0.01 −7.04 −0.01 −8.86 0.00 −10.42 0.00 −11.25 −0.01 −8.68 −0.01 −10.31 −0.01 −6.21 −0.01 −7.81 −0.01 −5.51 −0.01 −8.40 0.00 −8.31 −0.01 −8.99 −0.01 −6.51 0.00 −3.97 −0.01 −6.57 −0.01 −5.42 −0.01 −7.89 −0.01 −8.31 −0.01 −6.12 0.00 −7.33 −0.01 −5.74 −0.01

−0.02 −0.35 −0.02 −0.61 −0.03 −0.61 −0.01 −0.43 0.00 −0.10 0.00 0.91 0.01 0.21 0.00 0.11 −0.02 −0.47 0.00 0.08 0.00 0.17 −0.01 −0.51 −0.04 −1.13 −0.01 −0.48 −0.01 −0.21 −0.02 −0.48 −0.03 −0.55 −0.03 −0.70 −0.01 −0.40 −0.02 −0.58 −0.02 −0.37 −0.04 −0.63 −0.01 −0.10 −0.03 −0.48 −0.01 −0.30 −0.02 −0.43 0.00 0.02 0.00 0.09 0.01 0.23 0.00

ı2I

ı3I **

0.15 2.06 0.13** 2.21 0.13** 2.52 0.13** 3.65 0.13** 3.03 0.14 0.00 0.09** 2.59 0.14** 2.70 0.13** 2.76 0.11** 2.06 0.11** 3.63 0.12** 4.17 0.14** 3.53 0.13** 3.97 0.13** 2.29 0.14** 3.02 0.17** 2.40 0.14** 3.30 0.14** 3.51 0.12** 3.13 0.14** 2.58 0.16** 2.01 0.13** 2.20 0.16** 2.49 0.14** 2.97 0.12** 3.00 0.1** 1.58 0.11** 2.45 0.11** 1.50 0.12**

0.00 0.01 0.00 −1.61 0.00 −1.58 0.00 −1.28 0.00 −1.69 0.00 0.05 0.00 −0.92 0.00 −1.43 0.00 −1.20 0.00 1.28 0.00 −1.02 0.00 −1.50 0.00 −1.62 0.00 −1.34 0.00 −0.74 0.00 −0.87 0.00 −0.14 0.00 −0.97 0.00 −1.23 0.00 −1.54 0.00 −0.99 0.00 −1.20 0.00 0.11 0.00 −0.73 0.00 −0.85 0.00 −1.47 0.00 −0.39 0.00 −1.12 0.00 0.38 0.00

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

113

Table 9 (Continued) Industry

r˜It − r˜ft = ı0 + ı1I [˜rmt − r˜ft ] + ı2I [˜rmt − r˜ft ] × SD + ı3I (SD) ı0

t-Statistics Technology t-Statistics Telecommunication t-Statistics Travel & Leisure t-Statistics Others t-Statistics All t-Statistics

−6.59 0.00 −5.54 −0.01 −3.71 −0.01 −8.80 −0.01 −4.75 −0.01 −5.61

ı1I

ı2I

ı3I

0.02 −0.01 −0.18 −0.01 −0.14 0.01 0.23 0.02 0.32 −0.02 −0.35

2.16 0.15** 2.47 0.17* 1.85 0.11** 2.73 0.13 1.42 0.15** 2.06

−0.45 0.00 −1.04 0.00 −0.53 0.00 −1.61 0.00 0.44 0.00 0.01

This table presents the regression analysis results for 34 Japanese industries after Bali bombings. The first multiplicative dummy variable equation illustrates the impact on systematic risk and the second additive dummy variable equation shows the impact on the intercept (see Eq. (14)). * Significance at 10% level. ** Significance at 5% level.

to a decision that the permanent component of expected risk has undergone a “step-change” (see Lawson, 1980). We discuss the results of the individual terrorist events below. 3.1. United States attacks – September 11, 2001 Tables 3 and 4 summarise the parametric empirical results for September 11 for the different Japanese industrial sectors. We report the abnormal return on the first day of trading following the attacks and the five day cumulative abnormal return as well as their respective t-statistics for the 34 different industries. It should be noted that, unlike the U.S. market that opened 6 days after the attack, the Japanese market opened the day after the attack. In other the words, we are assessing the performance of the Japanese stock market on the 12th of September of 2001. The results reported in Tables 3 and 4 show a consistent negative effect on equities listed in the Japanese Stock Exchange following the September 11 attack. Fig. 1 supports this hypothesis, except for the Telecommunication, Gas and Oil, and Aerospace industries. Columns 2 and 3 of Table 3 report the abnormal returns and the parametric t-statistics for the various sectors. With the exception of Aerospace, Gas and Oil, and General Financials, all the other industries exhibited a negative and statistically significant abnormal return on the first trading day after the September 11 attacks. In other words, 29 out of 34 sectors were affected by the event. The Equity and Non-Equity Investment sector, which fell by 9.67%, was the industry most affected in the short term. However, our results show a rebound for this sector after 5 trading days (see Table 4). Chen and Siems (2004) assess the short term effect of September 11 on the global capital markets and the banking sectors. They document a fall of 6.2% and 6.3% for the market and banking sector, respectively, after the attacks. Richman et al. (2005), supporting Chen and Siems (2004), find a negative impact of 6.5% for the Japanese stock market. Consistent with Chen and Siems (2004) and Richman et al. (2005), we observe a decrease of 5.5% on the market index. It should be noted that we exclude firms with firm specific information surrounding the terrorist events. This may account for the reduced value that we report relative to the results of Chen and Siems (2004) and Richman et al. (2005). Furthermore, the Japanese banking sector, after excluding firm specific information, shows a drop of 2.25%. Carter and Simkins (2004) and Drakos (2004) demonstrate a 10–16% decline in the Japanese Airline businesses immediately after the attack. Our result for the Travel and Leisure industry confirms this drop, albeit, a drop of only 3.9%. Generally, our findings on the effect of the terrorist attacks on Japanese industrial portfolios show a clear and consistent fall in various. The cumulative abnormal returns suggest that 17 industrial groupings experience negative returns whilst one sector, Aerospace, reports positive cumulative abnormal returns over the following five trading days after September 11 (see Table 4). The second column of Table 4 shows that the Life and

114

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

Table 10 The short run impact of Madrid bombings on the systematic risk of Japanese industries. Industry

Aerospace t-Statistics Automobile t-Statistics Banks t-Statistics Beverages t-Statistics Chemicals t-Statistics Construction t-Statistics Electricity t-Statistics Electronics t-Statistics Engineering t-Statistics Equity & Non-Equity Investment t-Statistics Food & Drug t-Statistics Food Producers t-Statistics Forest t-Statistics Gas t-Statistics Gas & Oil t-Statistics General Industrial t-Statistics General Financial t-Statistics Health Care t-Statistics Household Goods t-Statistics Industry Transportation t-Statistics Leisure Goods t-Statistics Life & Non-Life Insurance t-Statistics Media t-Statistics Metal t-Statistics Personal Goods t-Statistics Pharmaceuticals t-Statistics Real Estate t-Statistics Retailers t-Statistics Software t-Statistics

r˜It − r˜ft = I + ˇI1 [˜rmt − r˜ft ] + ˇI2 [˜rmt − r˜ft ] × D + ε˜ it

r˜It − r˜ft = ϕI + ˛1I [˜rmt − r˜ft ] + ˛2I D + ε˜ it



ˇI1

ˇI2

ϕ

˛1I

˛2I

0.00 −1.47 0.00 −4.31 0.00 −3.92 0.00 −4.69 0.00 −3.73 0.00 −2.81 0.00 −7.49 0.00 −1.57 0.00 −2.20 0.00 −1.41 0.00 −6.97 0.00 −6.63 0.00 −4.16 0.00 −6.98 0.00 −2.98 0.00 −2.79 0.00 −0.02 0.00 −3.04 0.00 −2.73 0.00 −3.49 0.00 −2.10 0.00 −0.74 0.00 −0.65 0.00 −1.41 0.00 −2.73 0.00 −4.82 0.00 −0.56 0.00 −3.04 0.00 0.04

0.02 0.35 0.03 0.93 −0.01 −0.15 0.02 0.60 0.01 0.41 0.03 1.03 −0.03 −1.37 0.04 0.94 0.00 0.08 −0.01 −0.11 0.02 0.89 −0.01 −0.73 −0.01 −0.27 −0.02 −0.71 −0.06 −1.26 −0.02 −0.50 0.01 0.18 −0.03 −0.79 0.02 0.86 −0.02 −0.60 0.02 0.37 −0.09 −1.13 0.04 0.71 0.01 0.14 0.02 0.55 0.00 −0.01 0.02 0.41 0.02 0.64 −0.01 −0.12

−0.17 −0.26 0.12 0.36 0.79** 1.99 0.23 0.88 0.20 0.59 0.07 0.21 0.26 1.03 0.17 0.36 0.14 0.36 0.60 1.10 0.14 0.75 −0.03 −0.12 0.41 1.31 0.35 1.53 0.65 1.37 0.28 0.74 0.67 0.98 −0.23 −0.68 −0.07 −0.24 0.46 1.48 0.29 0.62 1.43* 1.78 0.05 0.08 0.44 0.74 0.15 0.39 0.33 1.05 −0.33 −0.62 0.15 0.51 0.16 0.23

0.00 −1.47 0.00 −4.31 0.00 −3.92 0.00 −4.69 0.00 −3.73 0.00 −2.81 0.00 −7.49 0.00 −1.57 0.00 −2.20 0.00 −1.41 0.00 −6.97 0.00 −6.63 0.00 −4.16 0.00 −6.98 0.00 −2.98 0.00 −2.79 0.00 −0.02 0.00 −3.04 0.00 −2.73 0.00 −3.49 0.00 −2.10 0.00 −0.74 0.00 −0.65 0.00 −1.41 0.00 −2.73 0.00 −4.82 0.00 −0.56 0.00 −3.04 0.00 0.04

0.02 0.35 0.03 0.93 −0.01 −0.15 0.02 0.60 0.01 0.41 0.03 1.03 −0.03 −1.37 0.04 0.94 0.00 0.08 −0.01 −0.11 0.02 0.89 −0.01 −0.73 −0.01 −0.27 −0.02 −0.71 −0.06 −1.26 −0.02 −0.50 0.01 0.18 −0.03 −0.79 0.02 0.86 −0.02 −0.60 0.02 0.37 −0.09 −1.13 0.04 0.71 0.01 0.14 0.02 0.55 0.00 −0.01 0.02 0.41 0.02 0.64 −0.01 −0.12

0.00 0.26 0.00 −0.36 −0.02** −1.99 −0.01 −0.88 0.00 −0.59 0.00 −0.21 −0.01 −1.03 0.00 −0.36 0.00 −0.36 −0.01 −1.10 0.00 −0.75 0.00 0.12 −0.01 −1.31 −0.01 −1.53 −0.01 −1.37 −0.01 −0.74 −0.01 −0.98 0.01 0.68 0.00 0.24 −0.01 −1.48 −0.01 −0.62 −0.03* −1.78 0.00 −0.08 −0.01 −0.74 0.00 −0.39 −0.01 −1.05 0.01 0.62 0.00 −0.51 0.00 −0.23

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

115

Table 10 (Continued) Industry

Support Service t-Statistics Technology t-Statistics Telecommunication t-Statistics Travel & Leisure t-Statistics Others t-Statistics All t-Statistics

r˜It − r˜ft = I + ˇI1 [˜rmt − r˜ft ] + ˇI2 [˜rmt − r˜ft ] × D + ε˜ it

r˜It − r˜ft = ϕI + ˛1I [˜rmt − r˜ft ] + ˛2I D + ε˜ it



ˇI1

ϕ

˛1I

˛2I

0.00 −1.00 0.00 −1.44 0.00 −0.55 0.00 −3.76 0.00 −0.84 0.00 −0.51

0.02 0.53 0.03 0.51 0.14 1.44 0.01 0.28 0.01 0.10 −0.07 −0.31

0.00 −1.00 0.00 −1.44 0.00 −0.55 0.00 −3.76 0.00 −0.84 0.00 −0.51

0.02 0.53 0.03 0.51 0.14 1.44 0.01 0.28 0.01 0.10 −0.07 −0.31

−0.01 −0.88 −0.01 −0.89 −0.01 −0.46 0.00 −0.39 −0.02 −0.96 −0.02 −0.30

ˇI2 0.37 0.88 0.51 0.89 0.48 0.46 0.10 0.39 0.99 0.96 0.73 0.30

This table presents the regression analysis results for 34 Japanese industries after Madrid bombings (see Eqs. (12) and (13)). The first multiplicative dummy variable equation illustrates the impact on systematic risk and the second additive dummy variable equation shows the impact on the intercept. * Significance at 10% level. ** Significance at 5% level.

Non-Life Insurance sector was the worst performing sector with −10.31% as CAR5. Over the five day period, the Automobile, Leisure Goods, Life and Non-Life Insurance, and Travel and Leisure deteriorated whilst the remaining sectors show a rebound. Chen and Siems (2004) find that both the market and the banking sector rebound after six days. We document a 5 day rebound for the banking sector and observe a −4.48% CAR5 for the entire Japanese market (see Table 4). Table 5 presents results of the non-parametric tests which we conduct as a robustness check. Analogous to the parametric tests, the negative impact of the events of September 11 on Japanese industries is also detected by the non-parametric tests. This is represented by the negative non-parametric tstatistics. The general conclusion that can be drawn from the above discussion is that of an adverse impact on industrial sectors in Japan following the September 11 attacks. A general assumption is that, following a terrorist attack, returns of equities fall as a result of an increase in systematic risk. We, thus, conduct an empirical test to determine if the industries negatively affected by the events of September 11 experience a general increase in their systematic risk on the first trading day following the terrorist attacks using the multiplicative regression analysis in Eq. (12). Columns 2–4 of Table 6 report the results of the multiplicative dummy variable model (Eq. (12)). A positive (negative) coefficient of the multiplicative dummy variable (ˇI2 ) reflects an increase (decrease) in systematic risk on the first trading day. The sign of the coefficient (ˇI2 ) is positive for all of the industries included in the study with exception of Aerospace and Gas and Oil. When the coefficient of the multiplicative dummy variable is statistically different from zero, it implies a significant statistical change in the systematic risk of the industry. The t-statistics results from column 4 of Table 6 show that systematic risk statistically increased in 26 sectors.2 Excluding General Financial, the remaining 25 sectors recorded a statistical decrease in abnormal returns on the first trading day. For example the systematic risk of Pharmaceuticals was 0.16 (see column 3 of Table 6) prior to the attack and increased to 1.86 (see column 4 of Table 6) immediately after the attack. The general increase in systematic risk of 1.733 can be attributed to the terrorist risk. This may explain the fall in abnormal return of the stock. In estimating equations (12), we only show the short term impact of the September 11 attacks on the Japanese industrial sectors. Applying Eq (14), we can observe the long term impact of this terrorist 2 Automobile, Beverages, Chemicals, Construction, Electronics, Engineering, E & N Investments, Food and Drug, Forest, General Financial, General Industrial, Health Care, Household Goods, Industrial Transportation, Leisure Goods, Life and Non-Life Insurance, Metal, Personal Goods, Pharmaceuticals, Real Estate, Retailers, Software, Support Service, Technology, Telecommunication, and Travel and Leisure sectors. 3 Wald test was carried out for all of the industries to test if the coefficient of the multiplicative variable (ˇI2 ) is redundant. The results, which is not reported, show that the terrorist systematic risk (ˇI2 ) is a significant factor.

116

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

Table 11 The long run impact of Madrid bombings on the systematic risk of Japanese industry indices. Industry

Aerospace t-Statistics Automobile t-Statistics Banks t-Statistics Beverages t-Statistics Chemicals t-Statistics Construction t-Statistics Electricity t-Statistics Electronics t-Statistics Engineering t-Statistics E&N Investment t-Statistics Food & Drug t-Statistics Food Products t-Statistics Forest t-Statistics Gas t-Statistics Gas & Oil t-Statistics General Industrial t-Statistics General Finance t-Statistics Health Care t-Statistics Household Goods t-Statistics Industry Transportation t-Statistics Leisure Goods t-Statistics Life & Non-Life Insurance t-Statistics Media t-Statistics Metal t-Statistics Personal Goods t-Statistics Pharmaceuticals t-Statistics Real Estate t-Statistics Retailer t-Statistics Software t-Statistics Support Service

r˜It − r˜ft = ı0 + ı1I [˜rmt − r˜ft ] + ı2I [˜rmt − r˜ft ] × SD + ı3I (SD) ı0

ı1I

0.00 −1.58 0.00 −3.62 0.00 −3.12 0.00 −4.02 0.00 −3.30 0.00 −2.65 0.00 −5.20 0.00 −1.68 0.00 −2.07 0.00 −1.63 0.00 −4.61 0.00 −4.85 0.00 −3.73 0.00 −5.69 0.00 −2.75 0.00 −2.62 0.00 −0.21 0.00 −2.88 0.00 −2.51 0.00 −3.15 0.00 −2.04 0.00 −0.94 0.00 −0.77 0.00 −1.65 0.00 −2.69 0.00 −3.89 0.00 −0.61 0.00 −2.42 0.00 −0.36 0.00

0.02 0.27 0.03 0.81 0.00 −0.03 0.01 0.43 0.01 0.31 0.02 0.43 −0.04 −1.00 0.04 0.88 0.00 0.05 0.00 −0.03 0.01 0.36 −0.01 −0.54 0.00 −0.02 −0.02 −0.71 −0.06 −1.16 −0.02 −0.47 0.00 0.01 −0.02 −0.50 0.01 0.38 −0.03 −0.86 0.02 0.43 −0.08 −1.07 0.04 0.72 0.01 0.19 0.02 0.36 −0.01 −0.18 0.01 0.14 0.00 0.06 −0.01 −0.13 0.02

ı2I 0.08 1.05 0.05 1.03 0.08 1.26 0.07* 1.88 0.09* 1.78 0.09* 1.95 0.11** 2.70 0.07 1.10 0.08 1.38 0.1* 1.68 0.08** 2.24 0.11** 3.27 0.07 1.55 0.13** 3.54 0.2** 2.96 0.12** 2.29 0.08 0.98 0.11** 2.15 0.09* 1.83 0.12** 2.58 0.07 1.19 0.20 2.14** 0.03 0.35 0.09 1.20 0.08 1.42 0.1** 2.06 0.03 0.35 0.08 1.43 0.08 0.89 0.06

ı3I −0.01** −6.20 −0.01** −9.41 −0.01** −6.95 −0.01** −11.63 −0.01** −9.09 −0.01** −9.92 −0.01** −9.01 −0.01** −8.36 −0.01** −8.61 −0.01** −8.89 −0.01** −12.58 −0.01** −13.37 −0.01** −10.59 −0.01** −10.57 −0.01** −6.21 −0.01** −9.20 −0.01** −7.14 −0.01** −9.02 −0.01** −10.35 −0.01** −10.49 −0.01** −7.85 −0.01** −5.40 −0.01** −7.43 −0.01** −6.38 −0.01** −8.88 −0.01** −9.13 −0.01** −7.73 −0.01** −9.38 −0.01** −6.33 −0.01**

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

117

Table 11 (Continued) Industry

t-Statistics Technology t-Statistics Telecommunication t-Statistics Travel & Leisure t-Statistics Others t-Statistics All t-Statistics

r˜It − r˜ft = ı0 + ı1I [˜rmt − r˜ft ] + ı2I [˜rmt − r˜ft ] × SD + ı3I (SD) ı0

ı1I

ı2I

−1.04 0.00 −1.54 0.00 −0.65 0.00 −2.85 0.00 −1.26 0.00 −0.74

0.34 0.03 0.58 0.15 1.63 0.00 −0.01 0.02 0.21 −0.05 −0.26

0.90 0.06 0.81 −0.07 −0.64 0.08* 1.83 0.07 0.68 0.30 1.40

ı3I −7.55 −0.01** −7.37 −0.01** −5.17 −0.01** −10.59 −0.01** −4.53 −0.01* −1.66

This table presents the regression analysis results for 34 Japanese industries after Madrid bombings. The first multiplicative dummy variable equation illustrates the impact on systematic risk and the second additive dummy variable equation shows the impact on the intercept (see Eq. (14)). * Significance at 10% level. ** Significance at 5% level.

attack. That is, we test if the increase in systematic risk observed on the first trading day after the event persists in the long term. The results, presented in Table 7 column 4, show that seventy percent of the industries exhibit an increase in systematic risk in the long run. For example, the systematic of the Chemicals industry increased by 0.38 after the September attacks. The robustness tests conducted are supportive of the above conclusions. The results from estimating equation (12.1) show that the highest impact is captured within the shortest window. Eq. (12) is the shortest window as the dummy variable is equal to one on the first day of trading and zero otherwise. In estimating equation (12), we find that the Automobile industry experience an increase in systematic risk of 1.48 (see Table 6). When Eq. (12.1) is estimated, it gives rise to larger window intervals, and for windows of 5-day, 10-day, 15-day, the change in systematic risk are 0.93, 0.46 and 0.31 respectively. Clearly, the robustness check also suggests a general increase in systematic risk as a result of a terrorist attack, in this case September 11. Eq. (12.2) controls for the natural asynchronicity between the markets. Estimating this model shows an increase of 1.45 in systematic risk for the Automobile industry, whereas Eq. (12) shows that it has increased by 1.48. Although there is a small difference in the estimates, the general conclusion about the increase in systematic risk remains the same. The findings of the other robustness tests are similar and in that sense the conclusions of our study are quite robust. For brevity purposes, we do not report these outputs. 3.2. Bali bombings – 12th October 2002 Tables 3 and 8 show the effect of Bali bombings on abnormal return and systematic risk, respectively, for Japanese industries on the first day of trading. The actual attacks occurred on the Saturday 12th of October 2002 and the first day of trading in Japan was on the 14th of October 2002. Columns 3 and 4 of Table 3 report the abnormal returns and the parametric t-statistics for the various sectors. In contrast to the September 11 results, this event did not have any pervasive and immediate effect on neither the abnormal returns nor the systematic risks of the Japanese equity industries and the general index. The results of the non-parametric robustness tests in Table 5 also support these findings.4 We also observe a slow response of the Japanese market to this event. Five trading days after the bombings, 17 industries experience positive returns as shown in the cumulative abnormal return Table 4. As could be seen in columns 4 and 5 of Table 4, the telecommunication industry registered

4 Our analysis on how terrorist risk changes systematic risk on the first day of trading reveals no effect except for the Automobile industry that recorded an increase in risk (see Table 8).

118

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

the highest positive CAR5 (10.89%) whilst the Food Producers industry recorded the lowest positive CAR5 of 2.86%. We are not the first to identify such reaction as Ito and Lee (2005) detect a shift from international travels to domestic travel in Japan and Canada following September 11. We observe an increase in the long term systematic risk of 31 sectors in Japan following the Bali attack. For example, as shown in Table 9, the Automobile and Engineering industries systematic risk increase by 0.33 and 0.47, respectively. This suggests an error in their long term expectations of risk following new information in Bali attacks. 3.3. Madrid – 11th March 2004 The Madrid bombings occurred on Thursday 11th March 2004. We examine the Japanese industry reactions both immediately, and five day following the event. The results of the parametric test immediately after the attacks and five day after the attacks are shown in columns 6 and 7 of Table 3 and Table 4, respectively. We find weak evidence of a negative impact of Madrid bombings on the Japanese industries in the first trading day where only the Banking industry fell by 1.68%. Rosendorff and Sandler (2005) suggest that the Madrid bombing was another defining moment given that it made it clear that terrorist attacks could occur anywhere and that terrorist could cynically respond to security upgrades by finding less secure avenues. Unsurprisingly, we observe an increase in systematic risk following the Madrid attacks for the Japanese Banking and Insurance industries on the first trading day (see Table 10). In the long term, we also show another adjustment in the change in risk and document an increase in systematic risk for 16 industries in Table 11. This shows that although counter-terrorism policy may be determined independently by countries, their outcome of their decision is interdependent and divergent (Enders and Sandler, 1993; Sandler and Enders, 2004). 3.4. London – 7th July 2005 On Thursday 7th July 2005, London was subjected to terrorist attacks. Surprisingly, the Japanese stock market’s response to the attack is insignificant on both returns and systematic risks in Japan following the attacks.5 The abnormal returns on the first trading day and cumulative abnormal returns are not statistically different from zero implying that Japanese industries were insensitive to the London attacks. The non-parametric t-statistic also supports these findings. 3.5. Mumbai – 11 July 2006 Analogous to the Bali bombings, we find that the Japanese industrial segments reacted cautiously to the Mumbai bombings. Whereas there is no effect detected on the first trading day for all industries in both returns and systematic risks, we show a negative CAR5 for 8 industries (see Table 4). Similar to the response engendered by the US terrorist attacks, this negative effect was more pronounced in the Life and Non-Life Insurance. We also find no change in the long term systematic risk of the Japanese industries. 4. Conclusion The literature generally notes that terrorism and military attacks increase the cost of doing business because of added security and increased risks. Whilst the market reaction as a whole has been extensively studied, only a handful of papers have looked into the specific sectors and the impacts of terrorism on foreign countries on domestic securities. Even then, only the banking and airline industries have been extensively studied. Additionally, the extant literature shows a dearth of studies on the terrorist attacks after September 11. In this paper, we utilise the adaptive expectations hypothesis and event study methodology to assess the impact of these terrorist attacks all Japanese industrial sectors.

5

These results, though available, are not reported for the sake of brevity.

M.A. Graham, V.B. Ramiah / Research in International Business and Finance 26 (2012) 97–119

119

Of the five recent terrorist attacks studies, our results show that the events of September 11 had the greatest effect on the Japanese market. The majority of the industries were down on the first trading day, and fifty percent of the industries were still negatively affected five days after the event. We also demonstrate an increase in systematic risk of these sectors in both the short run and long run following the attacks on the twin towers in New York. On the other hand, we find that the Japanese sectors were insensitive to the later terrorist attacks in Mumbai and London bombings. We interpret that to mean no new information on terrorist risk from these terrible events. There was an initial step-change in risk which was incorporated into expectations after the U.S., Bali and Madrid bombings. The evidence from the subsequent attacks imply there was no the forecast error in risk expectation in Japan after these three attacks. References Brounrn, D., Derwall, J., 2010. The impact of terrorist attacks on international stock markets. European Financial Management 16, 585–598. Brown, S.J., Warner, J.B., 1985. Using daily stock returns: the case of event studies. Journal of Financial Economics 14, 3–31. Cam, M., 2008. The impact of terrorism on United States industries. Economic Papers (Economic Society of Australia) 27 (2), 115–135. Carter, D.A., Simkins, B.J., 2004. The market’s reaction to unexpected, catastrophic events: the case of airline stock returns and the September 11th attacks. The Quarterly Review of Economics and Finance 44, 539–558. Chen, A.H., Siems, T.F., 2004. The effects of terrorism on global capital markets. European Journal of Political Economy 20, 349–366. Chesney, M., Reshetar, G., Karaman, M., 2011. The impact of terrorism on financial markets: an empirical study. Journal of Banking & Finance 35, 253–267. Choi, K., Hammoudeh, S., 2010. Volatility behavior of oil, industrial commodity and stock markets in a regime-switching environment. Energy Policy 38, 4388–4399. Corrado, C.J., 1989. A non parametric test for abnormal security price performance in event studies. Journal of Financial Economics 23, 385–395. Corrado, C.J., Truong, C., 2008. Conducting event studies with Asia-Pacific security market data. Pacific-Basin Finance Journal 16, 493–521. Drakos, K., 2004. Terrorism-induced structural shifts in financial risk: airline stocks in the aftermath of the September 11th terror attacks. European Journal of Political Economy 20, 435–446. Enders, W., Sandler, T., 1993. The effectiveness of anti-terrorism politics: vector-autoregression-intervention analysis. American Political Science Review 87, 829–844. Fernandez, V., 2007. Stock markets turmoil: worldwide effects of Middle East conflicts. Emerging Markets Finance and Trade 43 (3), 61–105. Fernandez, V., 2009. The behavior of stock returns in the mining industry following the Iraq war. Research in International Business and Finance 23, 274–292. Guzel, A., Ozdemir, Z.A., 2011. The Feldstein–Horioka puzzle in the presence of structural shifts: the case of Japan versus the USA. Research in International Business and Finance 25, 195–202. Hammoudeh, S., Li, H., 2008. Sudden changes in volatility in emerging markets: the case of Gulf Arab stock markets. International Review of Financial Analysis 17 (1), 47–63. Ito, H., Lee, D., 2005. Comparing the impact of the September 11th terrorist attacks on international airline demand. International Journal of the Economics of Business 12, 225–249. Karolyi, G.A., 2006. What do we know about terrorism and financial markets. Canadian Investment Review Summer. Lawson, T., 1980. Adaptive expectations and uncertainty. Review of Economic Studies 47, 305–320. Muth, J.F., 1960. Optimal properties of exponentially weighted forecast. Journal of American Statistical Association 55, 299–306. Nanda, M., Hammoudeh, S., 2007. Systematic risk, and oil price and exchange rate sensitivities in Asia-Pacific stock markets. Research in International Business and Finance 21, 326–341. Narayana, P.K, Narayan, S., 2010. Testing for capital mobility: new evidence from a panel of G7 countries. Research in International Business and Finance 24, 15–23. Nikkinen, J., Omran, M.M., Sahltrom, P., Aijo, J., 2008. Stock returns and volatility following the September 11 attacks: evidence from 53 equity markets. International Review of Financial Analysis 17, 27–46. Ramiah, V., Cam, M.-A., Calabro, M., Maher, D., Ghafouri, S., 2010. Changes in equity returns and volatility across different Australian industries following the recent terrorist attacks. Pacific-Basin Finance Journal 18 (1), 64–76. Richman, V., Santos, M.R., Barkoulas, J.T., 2005. Short- and long-term effects of the September 11 event: the international evidence. International Journal of Theoretical and Applied Finance 8, 947–958. Roll, R., 1988. The International Crash of October 1987. In: Kamphius, R., Kormendi, R., Waston, J. (Eds.), Black Monday and the Future of Financial Markets. Mid-American Institute. Rosendorff, B.P., Sandler, T., 2005. The political economy of transnational terrorism. Journal of Conflict Resolution 49, 171–182. Sandler, T., Enders, W., 2004. An economic perspective on transnational terrorism. European Journal of Political Economy 20, 301–316.