ABSTRACT
This study investigated the effects of terrorist attacks on the stock performance of tourism, travel, and leisure industries. The major tourist countries have been selected for this purpose. The novelty of this research is that not only it focuses on the relationship between terrorism and tourism stock performance and volatility but also uses an event study to examine this relationship. The results of this study revealed the significant effects of terrorist attacks on tourism firms’ performance and stock volatility in France, Spain, Thailand, Turkey, the United Kingdom, and the United States, while no significant effects were obtained for China and Germany. The overall panel event study analysis as well as the event study for individual countries illustrated the considerable adverse effects of terrorist attacks on firms’ performance in tourism, travel, and leisure industries.
Acknowledgements
The authors would like to thank Prof. Gulcay Tuna Payaslioglu (Department of Economics, Eastern Mediterranean University) for helpful comments in empirical analysis of this study.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Salih Katircioglu http://orcid.org/0000-0001-5108-9461
Cahit Adaoglu http://orcid.org/0000-0001-5771-9997
Notes
1 Russia is not included in the list of the countries due to a lack of tourism sector index. For France, the tourism stock index was active until February 2016. For more information on the sample countries, see https://www.e-unwto.org/doi/pdf/10.18111/9789284419029, accessed on April 13, 2018.
2 Italy is not included in the list, since there have been no major terrorist attacks in terms of human casualties after the introduction of the tourism sector index in the stock market.
3 For more information on the list and the criteria, see http://www.johnstonsarchive.net/terrorism/wrjp255b.html, accessed on March 30, 2018.
4 We also estimated both ARs and CARs, using market adjusted model and mean adjusted model of the event study. The findings of the performing market adjusted model reached the same conclusions as this study. Therefore, we believe it is unnecessary to report them here. However, performing mean return model abnormal returns on event day (0) varies for France terrorist incident (2), Spain terror incident (1), and Turkey’s examined incident. We will consider this difference in the following sections.
5 The best fit model in both mean and variance equation was found in the case of Spain, which followed ARMA (1, 0) and EGARCH (1, 2), respectively. For the UK, however, the best fit model followed ARMA (3, 1) and EGARCH (1, 2).
6 While estimating the abnormal returns by using the mean return model, the abnormal returns on event day (0) for France’s terrorist incident (2) is -1.46% and is statistically significant at 5% level.
7 While estimating the abnormal returns by using the mean return model, the abnormal returns on event day (0) for Madrid terrorist incident (1) is -2.67% and is statistically significant at 5% level.
8 While estimating the abnormal returns by using the mean return model, the abnormal returns on event day (0) for Turkey is -10.46% and is statistically significant at 1%.
9 The high negative reaction during period (CAR(0,+10)) is attributed to the other events rather than terrorist incident (2), as the pre-incident period abnormal return (CAR(-10,-1)) was -9.20% and was statistically significant at 1%.
10 Estimating Equations (3) and (4) and Equations (3) and (5) yield the same conclusion for all the variables except ψ and λ. Therefore, the reported parameters in Table 5 belong to the estimated results of Equations (3) and (4) except λ, which is from Equations (3) and (5).