Results
This part will provide all the results from the event study performed as explained in the methodology part. In table one the abnormal returns and the 5, 10 and the 30-day cumulative abnormal returns are described. The abnormal returns explain the effect on the day of the attack, the 5 day, 10 day and 30-day cumulative abnormal returns show later effects of the terrorist attacks because it could be possible that the terrorist effects are not displayed until after a couple of days. To find the significance of the tests, a t-test is used. The statistics of the t-test will also be described in table one. The effect that the t-test investigates the significance of is the effect of an attack on the AEX. This is also done in the paper of Chen and Siems (2004).
If the abnormal return is significant, it can be concluded that the stock market has been changing more than it normally does and then an attack has an abnormal effect on the AEX.
Terror Attack Date AR day 0 DayCar5 Daycar10 Daycar30
Madrid 11-03-2004 – 0.030659222 (-0.3699) – 0.056377366 (-7.6562)*** – 0.068525807
(-8.4355)*** – 0.029928284
(-13.2811)***
London 07-07-2005
– 0.017039809
(0.1545) 0.007062539
(-0.7209) 0.011170169
(1.0942) – 0.006268946
(3.8756)***
Paris 13-11-2015
(measured on 16-11-2015)
0.002243045 (-0.1077) 0.031530701
(5.2797)*** 0.041944771
(8.5792)*** – 0.007216499
(0.5927)
Brussels 22-03-2016
– 0.001013659
(0.1244) – 0.004198581
(-2.8812)** – 0.027364842
(-2.7460)** – 0.004172286
(1.9322)*
Table 1
T-values are between brackets
Significant at 0.1 *
Significant at 0.05 **
Significant at 0.01 ***
The hypothesis that was developed in section 2 of this thesis is: jihadi terrorist attacks in Europe have a negative effect on the Amsterdam Exchange Index. This hypothesis suggests that the effect of terrorist attacks is a significant negative one for the day itself and the thirty days after.
In the table, one can read that the impact on the event date is not significant for any of the attacks so the abnormal returns are not to be called abnormal. Differently, the cumulative abnormal return for day 5 (daycar5) is significant for Madrid (99% interval), Paris (99% interval) and Brussels (95% interval). This means that the returns are significantly abnormal and this means that terrorist attacks could have had an effect on the AEX there. The same goes for the cumulative abnormal return for day 10 (daycar10). The cumulative abnormal return for day 30 (daycar30) is significant for Madrid (99% interval), London (99% interval) and Brussels (90% interval). Noticeable is that the only significant positive cumulative abnormal returns are daycar5 and daycar10 in Paris. Also, Paris is the only one with a positive impact on the event date itself.
To describe the results more detailed subchapters will be used. In these subchapters, graphs will be shown and explained. The abnormal returns on the AEX from 100 days before the event an 30 days after the event are shown in the first graph of every subchapter. Every first graph has a red line on the event date which is the event date, except for figure 5, the red line here is a couple of days later due to the fact that the terrorist attack in Paris took place in the evening when the AEX was closed already. Next to this, a weekend was between the attack and the next opening, so the results could not be extracted the day after.
The second graph in every subchapter will give more detailed information about the cumulative abnormal returns. In every graph, the green line stands for the 5-day cumulative abnormal return, the red line stands for 10-day cumulative abnormal return and the orange line stands for 30-day cumulative abnormal return. On the y-axis is chosen for the 30-day car, which enables the option to put all the cumulative abnormal returns per event in one graph.
5.1 Madrid
The first attack that will be analyzed is the Madrid attack. This attack occurred on the 11th of March in 2003.
Figure 1 – Madrid (11-03-2003) – abnormal return
In figure one is the abnormal returns of the AEX and the event time are shown for the Madrid attack. The first thing that stands out in this graph are the fluctuations over time. This makes it difficult to say anything about the returns. On the event date, a negative return of -0.031 on the AEX can be seen, however, this is not a significant return, which makes it not an abnormal return.
Figure 2 – Madrid (11-03-2003) – cumulative abnormal return
In the next figure, the cumulative abnormal return can be observed for 5 days which has a negative impact of – 0.056, 10 days which has a negative impact of -0.069 and for 30 days which has a negative impact of -0.03. Also, all the observed points are significant at 0.01. This makes the cumulative returns, cumulative abnormal returns.
For the attack in Madrid, the hypothesis will be accepted. The cumulative abnormal returns are all significantly negative, which makes that the impact of the terrorist attack on the AEX is negative.
5.2 London
The second attack that is analyzed in this research is the attack in London in the 7th of July in 2005.
Figure 3 – London (07-07-2005) – abnormal return
In figure three the abnormal return of the AEX and the event time of the attack in London is shown. In this graph also, a lot of fluctuations are seen. In this graph, the results are similar to the results of the Madrid attack. The impact is -0.017 on the AEX, so it is a negative impact. However, it is not significant on the Event day itself.
Figure 4 – London (07-07-2005) – cumulative abnormal return
The 5 day and the 10-day cumulative abnormal returns for London both have a positive impact on the AEX; for the 5-day cumulative abnormal return this is 0.007 and for the 10-day cumulative abnormal return this is 0.01. Both of these cumulative abnormal returns are not significant. The hypothesis will be rejected for these results. The 30-day cumulative abnormal return is significant at 0.01 and the negative impact on the AEX is -0.0063, so for this result, the hypothesis will be accepted.
For the combined results it is difficult to draw one conclusion, but because only one of the three cumulative abnormal results is significantly negative the hypothesis will be rejected in this case.
5.3 Paris
The third attack that was investigated in this thesis is the attack in Paris on the 13th of November in 2015.
Figure 5 – Paris (13-11-2015) – abnormal return
In figure 5 the abnormal returns of the AEX and the event time of the Paris attack are shown. In this graph, the fluctuations are also present. The difference with other attacks is that the impact on the event date is positive 0.002, while other attacks have a negative impact on the event date. Still, this positive impact is not significant. Also, the attack in Paris took place after the AEX was closed, next to this a weekend came after the attack. This means that the day 0, in this case, is the Monday after the attack took place. This has been done because an effect cannot be measured if the AEX is closed.
Figure 6 – Paris (13-11-2015) – cumulative abnormal return
For Paris, the 5-day cumulative abnormal return has a positive impact of 0.032 on the AEX, the 10-day cumulative abnormal return has a positive impact of 0.042. Both of these returns are significant at 0.01. The 30-day cumulative abnormal return has a negative impact on the AEX of -0.007, however, this negative impact is not significant.
For the Paris attack, the hypothesis will be rejected, because the significant results did not have a negative impact and the results with a negative impact were not significant.
5.4 Brussels
The last attack that was investigated in this thesis is the Brussels attack that took place on the 22nd of March in 2016.
Figure 7 – Brussels (22-03-2016) – abnormal return
In figure 7 one can see the abnormal returns of the AEX and the event time of the Brussels attack. This graph again fluctuates a lot so drawing a conclusion by only looking at the graph is difficult. The abnormal return of the AEX on the event date is negative with 0.001. This is, however, again not significant.
Figure 8 – Brussels (22-03-2016) – cumulative abnormal return
For Brussels, the 5 cumulative abnormal return has a negative impact on the AEX of -0.001 and it is significant at 0.05. The 10-day cumulative abnormal return has a negative impact of -0.027 and is also significant at 0.05. The 30-day cumulative abnormal return has a negative impact of -0.004 and is significant at 0.01.
For this attack, the hypothesis will also be accepted, because of the cumulative abnormal returns that all show a significantly negative impact.
5.5 Summarizing results
The first thing that can be seen in every graph are the fluctuations over time. Although this is logical when studying a stock market, it is still something to mention. Second a negative impact on the event date is to be seen in every graph apart from the third figure which describes the AEX during the Paris attack which has a positive impact. Because of the many fluctuations, the impact of the attacks does not seem any bigger than normal. This is also proven by the t-test done by Stata, because of the fact that it is not significant for the event date.
To further evaluate the impact of news (of the terrorist attacks) on the AEX a measurement called cumulative abnormal return was used. These cumulative abnormal returns are significantly negative in some cases. Unfortunately, the cumulative abnormal return results are, unlike the abnormal return results, not pointing in one direction. Although there are more significant negative impacts, some of the positive impacts are also significant.
In short, this indicates that the European jihadi terrorist attacks do not have an effect on the event day itself, but the effect is significant a couple of days later. However, this effect can be both positive and negative. Also because of the fact that this model does not use the adjusted R squared or any control variables, it is still difficult to say if the cumulative abnormal returns that are measured by using the event – study are caused by the terrorist attacks, or by other variables that are not taken into account in this research.
For this the hypothesis cannot be rejected, nor can it fully be accepted. Still, looking into the results more detailed, the hypothesis: jihadi terrorist attacks in Europe have a negative effect on the Amsterdam Exchange Index, is accepted in the Madrid and Brussels attack. In the cases of London and Paris, the hypothesis is rejected.
Discussion
The effect of the jihadi terrorist attacks, that were found in the results section were not significantly negative on the event date. However, a significant result was expected after investigating the literature. The cumulative abnormal return, on the other hand, was significantly negative in more than half of the cases. Also, not all the impacts were negative, while this was expected. Madrid, London, and Brussels had a small negative return on the event date, while Paris had a small positive return on the first trading day after the event had happened. The Paris attack is the only attack with a different measured event day than the real event date. This is because of the fact that the event in Paris took place after the closing of the AEX on that day. Furthermore, a weekend (in which the AEX is closed) took place after it, so the impact on the AEX could only be measured a couple of days later. This could be a reason for the positive impact of the Paris attack. This will all be discussed in this part of the thesis.
The first reasons that comes to mind which could have led to the Paris attack triggering a positive return could be the so-called flight-to-safety effect (S.Narayan, T.-H.Le, & S.Sriananthakumar, 2018). This hypothesis holds that if an attack or another devastating event happens in one place, investors flee the area to go somewhere else and then invest in other markets. This results in positive returns in other markets.
Second, on the 16th of November 2015 (the day that is measured for the Paris attack) the gold and oil prices went up, this effect made Royal Dutch shell stocks go up too (Huijgevoort, 2015). The Royal Dutch Shell is the biggest company of the AEX, in 2016 the price of the AEX was for approximately 17% determined by Shell (Brandsema, 2016). This could have countered the negative impact of the attack in Paris. Apart from this, the AEX consists of 24 more companies. It could be that these companies differ so much from each other that some of these companies are affected by the attacks and others are not. Because of this, the effects on the AEX as a whole could be insignificant, or the negative impact could be countered.
Third, there may exist a smaller spillover effect than there has been described in the literature. This can have many reasons. It could be the case that the information spread is not as fast as it was thought to be, because of the fact that relaxation of control on movements of capital and foreign exchange transactions is stricter than before due to the economic crisis.
Fourth, which has much to do with the spillover effect. The integration between the European stock markets may be lower than suggested by the literature. This could mean that the stock markets in the countries where the attacks took place (Spain, England, France, and Belgium) are affected by the attacks, but the AEX is not. It could also be the case that the attacks do not have an effect in their own country too, this means that the integration between the European stock markets is as high as described in the literature, but because there is no effect on the stock market in the country, there is also no effect in other countries.
Another reason could be the distance, although this is accounted for by using Europe as a criterion. The attacks might still have been too far away to have any impact on the AEX at all.
Furthermore, it could also be the case that the Efficient market hypothesis did not hold for terrorist attacks, it might be that the returns of the AEX do not take the effects of terrorist attacks into account. Which makes terrorist attacks irrelevant information, or the AEX an inefficient market.
Eldor and Melnick (2004) did research on the effects of terror in the middle east on the stock markets in Israel, their conclusion was that terror had a big influence on the stock market in Israel. One of the differences with this thesis is that the Netherlands are in peace for a long time now and this might have as an influence on the AEX that the returns are more stable.
Next, many types of research use the 9/11 attack on New York’s Twin Towers to investigate an effect on stocks. However, this attack is bigger than any other attack which could be a factor that makes the impact of this attack much bigger.
Apart from these reasons, the methodology of this paper also has a couple of restrictions. An event-study does not have an adjusted R squared and neither does it have any control variables. This makes it difficult to say anything about the explanatory value of the results. While other methods do have those explanatory values, the event – study only uses the abnormal returns and the cumulative abnormal returns to say something about the data. Although the event – study still has these shortcomings; the literature used for this thesis all use the event study that is comparable to this method. This argues for the fact that this method is the best one for this research. However, it might be possible to test these shortcomings by using another method in new research. By doing this, a more useful and academic relevant conclusion could be drawn in future research.
Conclusion
Terrorism (especially jihadi terrorism) is getting more and more attention over the years. This is a good thing because as said in the introduction of this thesis, the number of attacks and the number of deaths per attack are rising. Especially in the last couple of years. Stock markets are being investigated for a long time already. The combination of those two factors (terrorism and stock markets), where this thesis is about, has been researched more and more in the last couple of years, this is among other things due to the fact that it is a relatively new threat. Because of this new risk for the economic world, research on this topic is economically and academically relevant. This also shows that this is a very contemporary subject. The main question of this thesis is: how have jihadi terrorist attacks in Europe affected the Amsterdam Exchange Index?
A lot of literature has been reviewed writing this thesis. Two theories have been described, namely the efficient market hypothesis and the spillover effects. Furthermore, different other papers have been described and were divided by country. Reviewing the literature and taking the two theories into account the following hypothesis has been formulated: jihadi terrorist attacks in Europe have a negative effect on the Amsterdam Exchange Index.
The terrorist attacks that are used to investigate in this thesis are Madrid, London, Paris, and Brussels. These were used because of the fact that they took place in Europe and because of the impact (the number of deaths) is high enough. The data used in this thesis is the returns of the AEX 100 days before until 30 days after the attack.
The formula 〖AR〗_JT = R_JT- R ̅_JT is used to calculate the abnormal returns. In this research, the cumulative abnormal returns are also used. The significance of these returns is calculated by using a t – test and in this way, it was measured if the terrorist attacks had a significant effect on the AEX.
Not all of the results were significant which means that not the whole hypothesis could be accepted. This is shown in the results part of the thesis. The abnormal returns on the event day were all insignificant which means that the European jihadi terrorist attacks do not have any influence on the AEX on the event date. The cumulative abnormal returns are significant in some cases, which means that in some attacks the AEX did have an abnormal return after a few days, however, this effect is both positive and negative. This concludes that the AEX is affected by European jihadi attacks in some cases. This results in the fact that the hypothesis cannot be accepted nor can it be fully rejected.
These results are discussed in the discussion part of this thesis. Also, the shortcomings of this research have been discussed in that part. The event-study does not have an R squared and neither does it have any control variables, this means that it is difficult to measure the explanatory value of this model. Finally, it might be possible to use another method to measure the explanatory value of this method, by doing this another conclusion which is more useful and academic relevant could be drawn from this and other new research.
Bibliography
Arin, K. P., Ciferri, D., & Spagnolo, N. (2008). The price of terror: The effects of terrorism on stock market returns and volatility. . Economics Letters, 101(3), 164-167.
Arshanapalli, B., & Doukas, J. (1993). International stock market linkages: Evidence from the pre- and post-October 1987 period* . Journal of Banking and Finance 17, 193-208.
Basu, S. (1977). Investment performance of common stocks in relation to their price‐earnings ratios: A test of the efficient market hypothesis. The journal of Finance, 32(3), 663-682.
Beechey, M., Gruen, D., & Vickery, J. (2000). The efficient market hypothesis: a survey. Sydney: Reserve Bank of Australia, Economic Research Department.
Brandsema, L. (2016, November 1). Waarom Shell zoveel invloed heeft op de koers van de AEX. Retrieved from finanzen.nl: https://www.finanzen.nl/aandelen/nieuws/Waarom-Shell-zoveel-invloed-heeft-op-de-koers-van-de-AEX-1001426641
Brounen, D., & Derwall, J. (2010). The Impact of Terrorist Attacks on International Stock Markets. European Financial Management 16(4), 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.
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, 539-558.
Chaudhry, N., Roubaud, D., Akhter, W., & Shahbaz, M. (2018). Impact of terrorism on stock markets: empirical evidence from the SAARC region. Finance Research Letters, -.
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, 253-267.
Corbet, S., Gurdgiev, C., & Meegan, A. (2017). Long-term stock market volatility and the influence of terrorist attacks in Europe. The Quarterly Review of Economics and Finance, 118-131.
Crenshaw, M. (1981). The causes of terrorism. Comparative politics, 13(4),, 379-399.
De Haan, J., Oosterloo, S., & Schoenmaker, D. (2015). Financial Markets and Institutions. Cambridge: Cambridge University Press.
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, 435-446.
Eckstein, Z., & Tsiddon, D. (2004). Macroeconomic consequences of terror: theory and the case of Israel. Journal of Monetary Economics, 51(5), 971-1002.
Eldor, R., & Melnick, R. (2004). Financial markets and terrorism. European Journal of Political Economy(20), 367-386.
Ender, W., & Sandler, T. (2011). The Political Economy of Terrorism. New York: Cambridge University Press.
Enders, W., Sandler, T., & Gaibulloev, K. (2011). Domestic versus transnational terrorism: Data, decomposition, and dynamics. Journal of Peace Research, 319-337.
Fama, E. F., Fisher, L., Jensen, M. C., & Roll, R. (1969). The adjustment of stock prices to new information. International Economic Review(10), 1-21.
Hardouvelis, G., Malliaropulos, D., & Priestley, R. (2006). EMU and European Stock Market Integration. The Journal of Business, Vol. 79, No. 1, 365-392.
Huijgevoort, S. v. (2015, November 16). Nabeurs: Effecten aanslag Parijs op beurs vallen mee. Retrieved from RTL Z: https://www.rtlz.nl/finance/beurs/nabeurs-effecten-aanslag-parijs-op-beurs-vallen-mee
Johnston, R., & Nedelescu, O. (2006). The Impact of Terrorism on Financial Markets. Journal of Financial Crime, 367-386.
Law, R. D. (2015). The Routledge history of terrorism. Routledge.
Levine, R., & Zervos, S. (1998). Stock markets, banks, and economic growth. American economic review, 537-558.
Malkiel, B. G. (1989). Efficient market hypothesis. Finance, 127-134.
Malkiel, B. G. (2003). The Efficient Market Hypothesis and Its Critics. Journal of Economic Perspectives—Volume 17, Number 1, 59–82.
Mnasri, A., & Nechi, S. (2016). Impact of terrorist attacks on stock market volatility in emerging markets. Emerging Markets Review, 28, 184-202.
Nesser, P., Stenersen, A., & Oftedal, E. (2016). Jihadi Terrorism in Europe: The IS-Effect. Perspectives on Terrorism, 10(6), 3-22.
Nikkinen, J., & Vähämaa, S. (2010). Terrorism and Stock Market Sentiment. The Financial Review(45), 263-275.
Pagano, m. (1993). Financial Markets and Growth. European Economic Review, 613-622.
Poe, A. C. (2001). Aftershocks of the 11th. HR Magazine, 46(12), 46-51.
S.Narayan, T.-H.Le, & S.Sriananthakumar. (2018). The influence of terrorism risk on stock market integration: Evidence from eight OECD countries. International Review of Financial Analysis, 1-12.
Shell, A. (2004, march 23). Fear of terrorism jolts stock market. USA Today.
Timmerman, A., & Granger, C. W. (2004). Efficient market hypothesis and forecasting. International Journal of Forecasting, Volume 20, Issue 1, 15-27.
Twin, A. (2006, September 11). Stocks: 5 years after 9/11. Retrieved from CNN Money: http://money.cnn.com/2006/09/08/markets/markets_fiveyearslater/index.htm
Yang, J., Min, I., & Li, Q. (2003). European Stock Market Integration: Does EMU Matter? Journal of Business Finance & Accounting, Volume 30, Issue 9‐10, 1253-1276.
Zanini, M., & Edwards, S. J. (2001). The networking of terror in the information age. In J. Arquilla, & D. Ronfeldt, Networks and netwars: The future of terror, crime, and militancy (pp. 29-30). Santa Monica: RAND.
Zhang, Y. J., Fan, Y., Tsai, H. T., & Wei, Y. M. (2008). Spillover effect of US dollar exchange rate on oil prices. ournal of Policy Modeling, 30(6), 973-991.
Self-Reflection
In this part of the bachelor’s thesis, the self-reflection will be written. It will be divided into subchapters for a better structure.
Process and personal competence, weaknesses
The process of this thesis was a difficult one for me, at the University we are used to writing papers, but none of them have been as big and important as this one. At the start, my academic background was weak and my academic writing skills left much to be desired. Also my ability to find the right academic papers was not good enough. In the very beginning, the chosen research question was not well academically substantiated, however after the first group meeting with my supervisor this problem was solved by choosing another research question. Furthermore the thesis lacked structure from the start and this was problematic in the beginning, however, this has gotten a lot better and in the end, this thesis is well structured.
I did overestimate my ability to write, at first I thought that some of the chapters could be written in less than a day, but unfortunately, this is, of course, not the case writing a thesis. Combined with a concentration problem and a planning problem this has been a pain throughout the entire thesis. I was also not always able to understand academic papers after reading them once, this has cost me loads of time. Unfortunately, I am not really organized which resulted in the structure problem at first, luckily this problem was solved during the thesis. I normally have the habit to write just enough, or just not enough for papers. In this thesis that was the problem at first also, but at this point that is not a problem for the thesis anymore.
Process and personal competence, strengths
From the start, this was the topic that I desired. This was a great boost, because of the interest that I had for this topic. The workshops where very helpful for me, I learned a lot about doing research and using data programs. Also, the program Stata which I did not fully understand at that time was explained to me. The recourses that we are able to use studying at the university are fantastic. Many different papers were a real help writing this thesis. The contact with my supervisor worked well, even though his handwriting is sometimes difficult to decode, he always gave good feedback.
Working independently is something that I really learned from this thesis. Normally there are always people around that follow the same subject and can give a lot of advice on the topic, but with the thesis this is different. You are writing your own paper with your own subject and only very few people can help you with that. Also, the determination that I possess helped a lot writing this thesis, especially at points when it becomes a little too much and other people are dropping out, but because of really wanting to finish my thesis I managed to finish it.
Process and personal competence, Lessons Learned
Some of the strengths and weaknesses were real learning points for me. My academic writing skills improved a lot and so did my academic background. Also, the searching for academic papers has been improved since the beginning of the thesis. Working well organized from the beginning is something that I will do from now on, it gives me more structure and calmness which improves the overall ability of writing. Next to this, I learned a lot about the program Stata and about the data searching programs, this will definitely be helpful in future writing.
I could have started earlier with some tasks only by planning better ahead, this is something I will do differently when a new challenge presents itself. This is one of the personal competencies that I should develop to overcome my weaknesses. Furthermore, I have to become communicatively more skilled.
Process and personal competence, Future Plans
Next Year I want to start my master’s program at the Radboud University in Nijmegen, the master that I want to participate in is the corporate finance and control master. After this, I want to get a job that fits me. To be able to reach these goals this bachelor’s thesis thought me to work more organized, not only by writing more structured but also in other parts of my life, like for instance planning and being more communicative.
For me, the most important current personal competencies to achieve my future goals are my determination and my self-confidence. Furthermore, being able to talk to others very easily (verbally strong) also helps a lot in achieving goals. New skills that I want to acquire are: being able to plan better for myself, also I want to become better at reading and understanding academic papers.