Terrorism and Stock Market Linkages: An Empirical Study from a Front-line State

Abstract

This article investigates the impact of prolonged terrorist activities on stock prices of different sectors listed in the Karachi Stock Exchange (KSE) by using the newly developed terrorism impact factor index with lingering effect (TIFL) and monthly time series data from 2002 (January) to 2011 (December). Johansen and Juselius (JJ) cointegration revealed a long-run relationship between terrorism and stock price. Normalized cointegration vectors are used to test the effect of terrorism on stock price. Results demonstrate a significantly mixed positive and negative impact of prolonged terrorism on stock prices of different sectors and show that the market has not become insensitive to the prolonged terrorist attacks.

Keywords

Prolonged terrorist attacks financial risk financial markets stock price

Introduction

After the 9/11 attacks, globally terrorist activities have increased by 23 per cent per year (Global Terrorism Database [GTD], 2013).¹ Pakistan has paid a very high price for being a front-line ally in the ‘War on Terror’ and the people of Pakistan have been the main target of terrorist attacks. Since the joining of Pakistan in the ‘War on Terror’, terrorist activities in the country have escalated manifold. According to the Global Terrorism Database (GTD), from 1 January 1991 to 31 December 2011 there had been 5,211 terrorist attacks out of which there had been 1,610 terrorist attacks up till 9/11, while there had been 3,601 terrorist attacks from 9/11 to December 2011 in Pakistan. The government has not paid enough attention to resolve the root causes of terrorism, resulting in a stress on Pakistani economy and people of the country (Shahbaz, 2013) (Table 1).

Table 1.

Terrorist Events in Pakistan from 2002 (January) to 2011 (December)

Year	Rank	Incidents	Fatalities	Injuries	Property Damaged
2002	11	40	96	310	13
2003	10	29	119	181	10
2004	5	58	270	658	17
2005	7	65	78	257	32
2006	5	144	244	579	55
2007	2	196	1,144	1,983	72
2008	2	491	1,008	1,739	157
2009	2	650	1,377	3,417	308
2010	2	682	1,519	2,865	345
2011	2	910	1,468	2,459	458

Source: Global Terrorism Database.

Terrorism is a tactic designed to set in oppressiveness and fear, aimed to attain a religious, political, economic or social objective. It is defined as an action with a purpose focused on explicit targets scattered over numerous geographical locations, and is highly unpredictable. Unlike other disasters, where the predictions for the occurrence of such events are possible, there is a very little possibility of predicting terrorism. It predominantly comprises the use of explosives and firearms. Moreover, the reaction time after identification of a potential terrorist event is substantially limited when compared to other disasters. This innate suddenness of terrorist events may initiate societal reaction by disrupting markets.

The act of terrorism on the target population and the impact on its economy have not been consistent or straightforward in instilling fear and uncertainty in the besieged population. Investors’ confidence and financial market instability have been used widely as indicators of fear owing to terrorism (Arin, Ciferri & Spagnolo, 2008; Eldor & Melnick, 2004; Shahbaz, Shabbir, Malik & Wolters, 2013). The terrorist attack of 9/11 on World Trade Center, the financial and economical hub of the USA, has been successful in short-term terror objectives. As a result, the New York Stock Exchange (NYSE) trading operations remained inactive for an entire week till 17 September, while in the intervening week Dow Jones fell by 14 per cent (Lenain, Bonturi & Koen, 2002).

Similarly, a suicide terrorist attack in Bali Island on 12 October 2002 killed 202 people, mostly tourists, and injured 240 people. The Indonesian stock market noted a consistent negative impact on the first trading day and it dropped by 25.53 per cent. The sector analysis showed that on the first day of trading, the investors’ capital reduced between 3.28 and 8.39 per cent and remained persistently negative for 5 days (Ramiah & Graham, 2013).

In contrast, on 7 July 2005 early morning terrorists carried out multiple bombings in London, killing 52 civilians and injuring over 700, the worst terrorist incident in the history of United Kingdom since 1988 Lockerbie bombing and first ever suicide attack in the country.

However, the London Stock Exchange (LSE) somewhat remained uninterrupted, besides initial decline in the Financial Times Stock Exchange (FTSE) 100 (London stock index) to approximately 4 per cent. It adjusted to the shock and trading continued to closed down to only 1.6 per cent. The varying reactions of the LSE and NYSE encourage to inquire whether terrorism has an impact on financial markets or are they relatively insensitive to terrorist incidents. Moreover, if financial markets are sensitive to the act of terrorism, are the effects momentary?

Studies suggest that the fear of terrorism can inhibit investors from making investment decisions freely and that societal confidence weakens. They also indicate that when terrorism becomes a frequent part of the environment, investor confidence may not be affected and market continues to function normally. Consequently, fear from terrorism wears away when it becomes a routine element over time and financial traders are discounted in financial markets. Effects of ongoing terrorism on the economy of a country were assessed in numerous studies such as those by Shahbaz (2013), Alam (2013), Suleman (2012), Arin et al. (2008), Eckstein and Tsiddon (2004) and Eldor and Melnick (2004), and have revealed immediate daily reaction of financial markets to the act of terrorism. Whilst negative effects have been associated with a single terrorist act, answers to the significance of prolonged terrorism involving mass casualty are less clear. Societal adaptation to a man-made disaster (terrorism) or natural disaster may lead to the ‘normalization’ of routine life and subsequent recovery of economic decision-making. This article seeks answers to the following questions: What are the consequences of prolonged terrorism on stock markets? Do terrorist attacks have an industry differential effect?

The rest of the article is organized as follows. In the next section, literature review, research objectives and rationale of the study are discussed. Then the section methodology describes the sample, data collection technique and the measurements of the terrorism impact factor with lingering effect (TIFL). The test of the proposed model using normalized cointegrating coefficient is performed and the findings are presented in the results section followed by the discussion and implication section where the findings are expounded.

Literature Review

Since the involvement in the ‘War on Terror’, the gravest problem faced by Pakistan is terrorism.

From the incident of 9/11 to December 2011, there had been 3,601 terrorist attacks of all types (GTD, 2013). These attacks have directly damaged the economy of the country by dropping foreign direct investment (FDI), threatening the confidence of domestic investors and upsetting financial markets, successively lowering economic growth, increasing unemployment and increasing poverty (Shahbaz, 2013).

Alam (2013) explored the relationship between terrorist activities in Pakistan and the stock market movement by using ‘Terrorism Impact Factor’, developed to study the overall impact of terrorist activities rather than an event study. No significant relationship was found between stock market returns and terrorism in the short run; however, in the long run terrorism had a negative impact on stock returns.

Aslam and Kang (2013) using Karachi Stock Exchange (KSE)-100 index time series daily data and terrorist attack news spanning the years from 2000 to 2012 investigated the effect of 300 terrorist attacks on the stock market of Karachi-Pakistan. Results indicate that the terrorist attacks have significant negative effects on KSE-100 index on the day of attack and 1 day before the attack; however, this effect is short lived. Empirical findings also indicated that the impact of attack varied across locations, types and severity of attacks.

Essaddam and Karagianis (2014) investigated the effect of terrorism on stock return volatility of American firms targeted by terrorist attacks. The act of terrorism was identified as an important risk factor in explaining the volatility of stock returns. Using an event study approach and bootstrapping technique, it was found that volatility increases on the day of the attack and remains significant for at least 15 days following the day of the attack showing lingering effect. This effect varies according to the characteristics of the country in which the incident has occurred. It was also found that firms operating in wealthier or more democratic countries face greater volatility in stock returns relative to firms operating in developing countries.

Ramiah and Graham (2013) using event study methodology probed how Indonesian equity market reacted to the terrorist attacks on USA (9/11), London (7/7), Spain, India and even Indonesia. Results show that equity portfolios were adversely affected by the September 11 attacks and Bali bombings. Domestic terrorist attacks also had a negative impact on Indonesian capital market, whereas London, Madrid and Mumbai terrorist attacks had no significant impact on Indonesian capital market.

Graham and Ramiah (2012) used an adaptive expectations hypothesis and event study methodology to estimate the impact of five terrorist attacks (9/11, Bali, Madrid, London and Mumbai) on all sectors listed in Japanese stock markets. Parametric and non-parametric tests were used to investigate whether the systematic risk and returns were affected by these events or not. In general, increase in systematic risk of some Japanese industries was documented whilst strong negative impact on returns was recorded for most of the industries on the first day of trading following the 9/11 attacks.

Chesney, Reshtar and Karaman (2011) investigated the impact of 77 terrorist events that occurred in 25 countries on their financial market indices. Twelve years’ financial market’s daily price indices starting from 4 January 1994 to 16 September 2005 were used and the behaviour of stocks, bonds and the commodity market was analyzed using an event study approach. They found that terrorist attacks significantly affect European, American, Swiss and global markets. Along the same lines, Kollias, Papadamou and Stagiannis (2011) used event study methodology to examine the impact of terrorist attacks on 11 March 2004 in Madrid and 7 July 2005 in London; the two major terrorist events on three major stock exchanges is Spain (Madrid, Valencia and Barcelona) and the LSE. Empirical findings point to similar reactions on the event day, but noted that significantly different recovery periods (in days) are needed for stock markets to rebound. Furthermore, the negative effects were short lived.

Drakos (2010) examined whether terrorism exerts a significant negative impact or not on daily stock market returns in a sample of 22 countries. In contrast to previous research, which focused on selected major terrorist acts, it tested for overall terrorist activity; it also explored whether this effect is a function of the level of psychosocial impact caused by terrorist incidents or not. The results supported negative returns on the day of a terrorist attack, which amplifies with the increase in psychosocial effects.

Brounrn and Derwall (2010) examined the effects of major significant terrorist attacks on stock markets of major economies of the world. Using event study method, their analytical results suggested that terrorist attacks affect stock prices mildly negatively. They compared these price reactions to those from natural unanticipated disasters, such as earthquakes, and concluded that price declines following terror attacks are more pronounced. However, in both cases it bounced back within the first week of the aftermath. Comparison of price responses of different industries showed different reactions and were strongest for the local markets. Results suggested that financial industry reacted strongly to terror events but recovered swiftly except for 9/11 attacks which caused long-term effects on the financial industry.

Ramiah, Cam, Calabro, Maher and Ghafouri (2010) investigated the impact of five terrorist attacks (9/11, Bali, Madrid, London and Mumbai) on different sectorial equities listed on the Australian Stock Exchange. Significant short-term negative abnormal returns were reported around the 9/11 attacks and to a limited extent the Madrid and London bombings. Furthermore, weak positive equity response to the Bali bombing and no response from the Mumbai attack in the Australian market were observed.

Peleg, Regens, Gunter and Jaffel (2011) found that prolonged terrorism may affect society’s resilience and sensitivity. They found that Israel’s financial market remained sensitive to each act of terrorism but have shown sustained psychological resilience and indicated no apparent overall market shift. This is an evidence of erosion of sensitivity to prolonged terrorism. In other words, ‘normalization of terror’ was observed following an extended period of continued suicide bombings.

Cam (2008) studied the impact of the 9/11, Bali and Madrid bombings on 135 industry equity indices in USA using event study method. The empirical evidence showed that industries reacted differently to terrorism. Following 9/11, airline, hotel and leisure industries recorded negative abnormal returns, whilst water, defence and telecommunications industries showed positive abnormal returns, which confirm the findings of Bruck and Wickstrom (2004). Bali and Madrid attacks had little impact on the US industry equity returns, suggesting that US industries were only moderately sensitive to these attacks.

Garvey and Mullins (2008) studied the London financial market with a focus on terrorist events carried out during 1998–2004 and examined how they have been perceived among participants of the London financial market. Financial options was used as a measurement tool for measuring terrorist risk perception. If risk amplification/attenuation occurs, one would expect to see changing levels of demand in the options markets. Data from the London options market suggest a high degree of sensitivity to terrorist events and reveal a vulnerability in the financial markets.

Eldor and Melnick (2004) studied the impact of suicide attacks (on Jewish population) on financial markets; the results exhibited long-run effect of suicide attacks on financial markets. They did not find any evidence for market unresponsiveness towards terror attacks over time and financial markets continued to incorporate NEWS-related terror attacks as they occur efficiently.

Through the following literature review, we can assume that investors do not necessarily react negatively to terrorist attacks. Investors tend to respond negatively only when they perceive an increase in the expected risk and market players may react if they perceive that the terrorist attack will have an impact on expected returns. Moreover, it is possible that stock markets do not react negatively on days surrounding a major terrorist attack. However, we believe that markets can respond differently to different attacks and that the variability in risk and returns differs significantly across different sectors within an economy.

Many studies have investigated the relationship between terrorism in Pakistan and other aspects of Pakistan’s economy. Previous researches in this area fall roughly into two groups. The first group of studies tests the main determinants of terrorism in Pakistan. Papers in the second group examine the effects of terrorism on economic growth. A brief review of these studies is presented in the following paragraphs.

Shahbaz et al. (2013) used economic reasons for the increase of terrorism in Pakistan rather than focusing on the political and institutional effects on terrorism. Capital and trade openness were integrated with economic growth to explore the causal relationship between terrorism and economic growth. Long-run relationship and direction of causality between the variables were tested. Empirical results confirmed a long-run relationship between economic growth, terrorism, capital and trade openness. A bidirectional causality is found between capital and terrorism, capital and trade openness and trade openness and terrorism. A unidirectional causality was noted running from economic growth to terrorism.

Shahbaz (2013) using annual data from 1971 to 2010 investigated the relationships between terrorism, economic growth and inflation. The analytical results demonstrated long-run linkages between terrorism, economic growth and inflation in Pakistan; furthermore, they revealed that an increase in inflation will escalate terrorism whilst economic growth also contributes in escalating terrorism. Besides, this bidirectional causality was also noted for inflation and terrorism. Consequently, it is concluded that lowering the inflation will reduce terrorism. Furthermore, it deduces some problems for policy-makers in their quest for economic growth, as growth would result in an increase in terrorism activities.

Raza and Jawaid (2013) investigated the impact of terrorism on tourism in Pakistan by using annual time series data for the period from 1980 to 2010. Empirical evidences confirmed long-run negative relationship between terrorism and tourism and also confirmed the unidirectional causal relationship between terrorism and tourism, where causality runs from terrorism to tourism.

Methodology

Based on literature review, it is identified that the earlier research can be divided into the following two categories: event studies, where the behaviour of the financial markets was studied for the period before and after the occurrence of terrorist events (Aslam & Kang, 2013; Chesney et al., 2011; Drakos, 2010; Eldor & Melnick, 2004; Graham & Ramiah, 2012; Peleg et al., 2011), and terrorism index studies, where the impact of terrorism on financial markets was studied through country-specific terrorism index (Alam, 2013; Arin et al., 2008). Furthermore, these indices were constructed on subjective determination of terrorist events by the researchers which may raise doubts about their being unbiased.

To study the impact of terrorism on the financial market of Pakistan where non-stop series of acts of terrorism are occurring all over the country, event study methodology will not be able to accurately estimate the risk associated with each of the event. However, there is a need to conduct a study, which can envision the overall impact of terrorist activities, not just a single activity, or a selected number of activities index as in the studies mentioned earlier.

To fill this gap and to have a holistic picture of impact of terrorism on the financial market of Pakistan, we constructed an index similar to the index of the Institute for Economics and Peace (2012) and called it TIFL. For this index, the impact factor for every terrorist incident that occurred during the study period and listed in the GTD (2013) was calculated and a monthly series of TIFL was developed for testing the hypothesis.

Data

Financial data: We used monthly stock price data from Datastream for the period 2002 (January)–2011 (December), for a total of 120 observations for 13 industries.

Terrorism data: For the second data set, we fabricated a new terrorism impact factor index using publicly available information on terrorist events mainly provided by the GTD (2013). The terror events that occurred during the weekend are summed up to the last Friday’s figure. Data cover all terrorist events that occurred in the sampled period from 2002 (January) to 2011 (December). In earlier researches, mostly event study methodology was used to evaluate the effect of terrorist events (Abadie & Gardeazabal, 2003; Brown & Warner, 1980; Chen & Siembs, 2004). Moreover, where index is constructed, the researchers subjectively determined terrorist events for the construction of their index (Aslam & Kang, 2013; Chesney et al., 2011). We used monthly terror index, which was constructed using all terrorist activities that have taken place from 2002 (January) to 2011 (December) listed in the GTD. The GTD recorded 3,061 events during the mentioned time period. Using GTD terrorism index scoring method, daily terror index (relative impact of incidents) was constructed and later on converted into monthly terror impact factor index. There are four factors counted in the monthly score:

Number of terrorist events that occurred in a particular month.

Number of fatalities due to terrorist events in a particular month.

Number of injuries caused by terrorist events in a particular month.

Estimated level of property damage from terrorist events in a particular month.

Each of the factors is weighted differently; the daily terror index is defined as the sum of 3 × number of human casualties + 0.5 × number of people injured + number of terrorist attacks + 2 × level of total property damage² occurring each day which was later on converted into monthly. Finally, 5-month weighted average is applied to importantly reflect the lingering psychological effect of terrorist acts over time.

Figure 1 shows the severity and intensity of the terrorism impact factor. The height of the bars indicates the severity, whereas the intensity can be observed by the density of the bars. It can be observed that after 2007 the severity and intensity both have increased drastically.

Figure 1.

Source: Global Terrorism Database.

Model

The aim of this study is to test the impact of prolonged terrorist attacks on the overall stock market through its index and on different industries listed in KSE. The basic model we used for this research is given below:

S P = T I F L + I + ε_{t}

(1)

where SP represents the total market index/sector index, TIFL is the focus variable of the study and I represents the set of control variables in the model, comprising fdi (Pilinkus, 2009), net portfolio investment (pi)(Jawaid & Haq, 2012) and remittances (rem).

Augmented Dickey–Fuller (ADF) unit root test is used to examine the stationary properties for time series variables. The ADF test is based on the equation given below:

Δ Y_{t} = α + β T + γ Y_{t - 1} + \sum_{j = 1}^{k^{'}} δ_{j} Δ Y_{t - j} + ε_{t}

where ∆Y_t is the first difference of the time series Y_t, α is the constant, β is the coefficient of time trend, k is the maximum lags order of the autoregressive process and t is the error term also known as white noise. The ADF test provides cumulative distribution of ADF statistics; it determines whether the estimate is γ = 0 or γ < 0. The series is said to be stationary if the test statistic is less than the critical values from Fuller table.

Johansen and Juselius (JJ) (1990) and autoregressive distributed lag (ARDL) cointegration methods are used to examine the long-run relationship between tifl_t and x_it. The JJ cointegration test is based on λ_Trace and λ_max statistics. First, the ‘trace test’ cointegration rank ‘r’ is as follows:

λ_{T r a c e} = - T \sum_{j = r + 1}^{n} \ln (1 - λ_{j})

Second, λ_Max maximum number of cointegrating vectors against r + 1 is presented in the following way:

λ_{\max} (r, r + 1) = - T \ln (1 - λ_{j})

where λ_Max is the maximum eigenvalue statistic to test the null hypothesis that there are r cointegrating relations against the alternative hypothesis that there are r + 1 cointegrating relations in the system. λ_Trace, the trace statistic, tests the null hypothesis that there are at most r cointegrating relations in the system, against the alternative of m, that is, the number of cointegrating vectors is less than or equal to r. T is the sample size, ln is the logarithm and λ_j is the eigenvalue. The critical values for λ_Max and λ_Trace test statistics have been tabulated by Pesaran et al. (2000).

If there is one cointegrating vector for each model, it shall be identified by normalizing market index for overall market model response and by normalizing sector index for each sector model. The cointegrating vector is normalized based on the argument that (overall market and sector) indices of a stock market can be estimated using equation (1).

Empirical Results

In this section, the results of the casual relations between terrorism, measured through newly developed terrorism index TIFL, and stock prices are reported and discussed.

The results of the ADF unit root test statistics are reported in Table 2. The lag length is used to estimate the ADF unit root. Test statistics are based on the Akaike information criterion. All studied variables are found to be non-stationary at level and stationary at first difference, which hold for the existence of unit root in all variables.

Table 2.

Stationarity Test Results

Variables	Level		Difference
	C	C&T	C	C&T
ER	1.141	–1.918	–5.409	–7.901
FDI	–1.743	–1.714	–10.152	–10.117
PI	–1.774	–1.830	–2.952	–4.049
REM	1.306	–2.731	–4.323	–5.213
TIFL	–1.734	–3.319	–15.181	–15.118
Market index	–0.207	–1.304	–8.857	–8.861
Oil and gas	0.665	–1.145	–8.729	–8.838
Auto and parts	–0.371	–1.955	–9.120	–9.127
Industrial	–1.838	–1.835	–7.776	–7.746
Utilities	–0.683	–1.257	–9.600	–9.606
Consumer	–2.385	–2.597	–7.740	–7.811
Tobacco	–1.674	–1.655	–8.590	–8.575
Chemical	–0.002	–2.366	–7.955	–7.997
Financial	–1.507	–1.583	–8.257	–8.226
Health and care	–0.912	–1.445	–8.052	–8.030
Telecom	–1.842	–1.983	–10.230	–10.246
Travel and leisure	–1.436	–2.277	–8.438	–8.414

Source: Authors’ estimations.

The results of the cointegrating vectors are reported in Table 3. λ_Trace and λ_Max test statistics are computed with unrestricted intercepts and no trends in vector autoregression (VAR). For the overall market index, the λ_Trace and λ_Max test statistics show that there is one cointegrating vector at the 5 per cent significance level and similarly, it also exists for all studied sectors at 5 per cent. In general, existence of one cointegrating vector for all indices is not rejected. This study normalizes the cointegrating vector by X_it indices (Effiong, 2014; Wong, 2010; Ziramba, 2010).

Table 3.

Cointegration Test Results

Model	Null HypothesisNo. of CE(s)	TraceStatistics	5% Critical Values	Max-eigenStatistic	5% Critical Values
Market Index	None	109.509	95.753	44.661	40.077
	At most 1	64.847	69.818	29.229	33.876
	At most 2	35.618	47.856	16.902	27.584
Oil and Gas	None	109.467	95.753	42.354	40.077
	At most 1	67.113	69.818	26.009	33.876
	At most 2	41.104	47.856	19.260	27.584
Utilities	None	137.303	95.753	62.662	40.077
	At most 1	74.641	69.818	40.259	33.876
	At most 2	34.382	47.856	16.914	27.584
Auto and Parts	None	116.548	95.753	43.283	40.077
	At most 1	73.264	69.818	28.939	33.876
	At most 2	44.325	47.856	18.874	27.584
Consumer	None	97.338	95.753	40.707	40.077
	At most 1	56.631	69.818	23.480	33.876
	At most 2	33.150	47.856	13.630	27.584
Industrial	None	107.132	95.753	45.964	40.077
	At most 1	61.167	69.818	27.287	33.876
	At most 2	33.879	47.856	10.892	27.584
Tobacco	None	178.133	95.753	75.344	40.077
	At most 1	102.789	69.818	60.168	33.876
	At most 2	42.620	47.856	22.406	27.584
Chemical	None	153.947	95.753	63.599	40.077
	At most 1	90.348	69.818	38.155	33.876
	At most 2	52.193	47.856	28.534	27.584
Financial	None	191.984	95.753	86.270	40.077
	At most 1	105.713	69.818	63.952	33.876
	At most 2	41.760	47.856	19.280	27.584
Health and Care	None	184.733	95.753	75.847	40.077
	At most 1	108.886	69.818	63.871	33.876
	At most 2	45.015	47.856	20.737	27.584
Telecom	None	149.656	95.753	75.745	40.077
	At most 1	73.911	69.818	43.206	33.876
	At most 2	30.705	47.856	13.376	27.584
Travel and Leisure	None	141.259	95.753	69.353	40.077
	At most 1	71.905	69.818	34.780	33.876
	At most 2	37.125	47.856	15.367	27.584

Source: Authors’ estimations.

Normalized cointegrating vectors are reported in Table 4. The maximum likelihood ratio test statistic³ is used to test whether the coefficient of explanatory variable is zero (Hall, 1989). The results of the likelihood statistic are rejected mostly at 1, 5 or 10 per cent level. Therefore, the explanatory variables are generally important to be included in the estimation.

Table 4.

Results of the Normalized Cointegration Vectors

Model		ER	FDI	PI	REM	TIFL
Market Index	Coefficient	–2.256	1.327***	2.310***	1.267***	–0.771***
	t-statistics	–0.288	4.896	3.071	3.151	–2.988
Oil and Gas	Coefficient	7.377	2.547**	10.330***	5.753***	–4.323***
	t-statistics	0.232	2.338	3.472	3.440	–4.089
Utilities	Coefficient	–8.409	–2.230***	9.806***	0.317	0.367
	t-statistics	0.468	4.684	8.198	0.351	–0.672
Auto and Parts	Coefficient	9.003	1.666**	10.945***	2.668**	–1.443*
	t-statistics	0.380	2.155	4.513	2.130	–1.908
Consumer	Coefficient	49.697**	–0.284	11.732***	–0.203	2.983**
	t-statistics	1.755	–0.297	4.556	–0.140	–3.285
Industrial	Coefficient	8.127	0.164	2.822***	0.156	–0.883**
	t-statistics	1.053	0.650	4.060	0.406	–3.532
Tobacco	Coefficient	–6.890	2.564***	11.619***	–0.613	2.637***
	t-statistics	–0.459	6.540	10.505	–0.814	5.757
Chemical	Coefficient	–20.922	–2.130***	10.591***	1.389	0.725
	t-statistics	–1.115	–4.201	8.197	1.472	1.241
Financial	Coefficient	–97.825***	7.157***	8.891***	2.183**	4.119***
	t-statistics	–5.198	14.256	6.766	2.334	7.151
Health and Care	Coefficient	–4.639	0.713***	1.614***	–0.087	0.518***
	t-statistics	–1.518	8.912	8.110	–0.583	5.630
Telecom	Coefficient	2.545	11.371***	–38.075***	2.184	–0.093
	t-statistics	0.040	6.535	–8.748	0.688	–0.048
Travel and Leisure	Coefficient	152.541	–13.866***	54.363***	–0.829	1.258
	t-statistics	1.654	–5.725	8.944	–0.179	0.447

Source: Authors’ estimations.

Note: *significant at 10%, **significant at the 5% and ***significant at 1%.

Table 4 summarizes the empirical results for prolonged terrorist attacks, measured through TIFL, whereby the evidence of significant negative as well as positive impact is found for different industries.

The results reported show a significant negative effect on overall market and oil and gas variables. The negative effect specifies that besides personal injuries and the loss of lives of targeted population, terror attacks also have economic costs, which reduce firms’ expected profits. Furthermore, it indicates that industries with a negative reaction to domestic terrorism are more susceptible to overseas assaults.

However, four industries—financial, consumer, tobacco and health and care—exhibited statistically significant positive relationship with the stock price. The financial industry index, which includes insurance and banks as sub-industries, had the highest positive coefficient.

This result was expected as terrorist risks prolonged and increased after engaging in war on terror, leading to increased demand for insurance. Also after the events of 9/11, Pakistan assumed a wide range of dangerous responsibilities for the international campaign against terrorism and received substantial assistance. Pakistan’s debt was unscheduled for 28 years, which provided a relief of about $1 billion a year in debt servicing. Strict scrutiny of Muslim account holders in foreign banks in Western countries impelled Pakistanis to repatriate their foreign accounts back to Pakistan, mainly through official channels. As a result, remittances went up. Private investors in the Middle East also felt that it would be less risky to invest their surplus capital in Pakistan than in Europe and America. These external factors increased bank deposits and their profitability.

Finally, for consumer, tobacco, health and care and travel and leisure industries, there seems no obvious reason for this positive movement. However, it could be that these industries were perceived as ‘refuge’ industries when compared to other risky industries, showing the investor sentiments in shifting their investment to less risky assets during the periods of high-risk events (Liu, Longstaff & Pan, 2003; Zaighum, 2014).

Conclusion

The extant body of literature argues that stock markets provide services that increases economic growth by mobilizing savings, whereas terrorism decreases the trust of local and international investors and negatively affects investment. Terrorism in Pakistan has a calamitous effect on its society and it is a phenomenon causing not only fatalities and injuries but also real economic cost.

Government statistics show that the direct and indirect economic costs of terrorist attacks in Pakistan have crossed USD 70 billion during the studied period. In this article, we examined the effect of terror attacks, measured through TIFL, on stock price of listed companies in the KSE. Our investigation exhibits that over the studied period of 10 years (2002–2012), both intensity and severity of the terrorist events caused the Karachi stock market index (KSE-100) to fluctuate differently and have shown a significant negative effect. This is evidence to believe that investors are sceptic about the future performance of the stock market and are reluctant to invest. A long-term relation between them was also eminent which rejects the plausible idea of terrorism becoming part of the routine life, that is, ‘normalization of terror’. Moreover, it indicates that the market is efficient in integrating the terror news but remains unsuccessful to accumulate evidence supporting unresponsiveness to terror attacks over time. Consequently, it is concluded that Karachi stock market is functioning in an efficient way and promptly reflects the effects of terror attacks.

This article also tested the impact of terrorist attacks on different industries listed in the KSE-100 index. It was noted that various industries responded differently to terrorism. Some industries experienced decrease in price, while others recorded an increase. The financial, tobacco and health and care sectors experienced a rise in prices. In contrast, oil and gas, auto and parts, industrial and telecom sectors experienced a fall in prices. The analytical results of this article highlighted the similarities that exist between TIFL and risky events. Likewise, this article also provides some clues for investors desiring to reduce the risk related to terrorism through industry diversification. If concerned about this risk, investors should hold assets that can react positively to terrorist attacks or, alternatively, assets that have little or no negative sensitivity to terror shocks.

Furthermore, global capital markets today are tightly interlinked; news spreads rapidly (especially bad news) with quick spillover, or contagion, effects (Chittedi, 2015). Abbas, Khan and Shah (2013) investigated the existence of volatility transmission and the long-run relationship among Pakistan’s regional equity markets including China, India, and Sri Lanka. They found ample evidence of bidirectional transmission of volatility between India and Pakistan. Moreover, the null hypothesis of no volatility transmission between Pakistan and China cannot be rejected. A closer examination of data indicates that volatility spillover amongst the four countries is mostly from a larger market to a smaller market. Some evidence of spillover from a smaller market to a larger market was also found.

A number of policy insights can be drawn from this analysis. First, South Asian neighbouring countries should work jointly to curb terrorism. More developed South Asian countries must assist less developed neighbours to protect themselves and to recover from terrorist attacks and should take a leading part in controlling countermeasures against a common terrorist threat. Second, for terrorist groups, such as Taliban and Al-Qaida, coordinated government actions are needed to root them out because any Asian foothold that these terrorists achieve will allow them to pose greater risks throughout the region. Finally, Western countries also have a real interest in eliminating these terrorist groups; therefore, Western countries can greatly assist South Asian countries’ efforts to address such common terrorist threats which bolster the case for joint Asian efforts, supported by Western help.

Footnotes

Acknowledgements

The authors are grateful to the editor and anonymous referees of the journal for their valuable comments to a previous version, and to Syed Tehseen Jawaid and Syed Ali Raza for their research support. We are also thankful to the participants of ICMESR 2015 in Karachi for their valuable comments on a previous draft of this article. Usual disclaimers apply.

Notes

References

Abadie

, & Gardeazabal

(2003). The Economic Costs of Conflict: A Case Study of the Basque Country. American Economic Review, 93(1), 113–132. doi: 10.1257/000282803321455188

Abbas

, Khan

, & Shah

S.Z.A.

(2013). Volatility transmission in regional Asian stock markets. Emerging Markets Review, 16, 66–77. doi:10.1016/j.ememar.2013.04.004

Alam

(2013). Terrorism and stock market development: Causality evidence from Pakistan. Journal of Financial Crime, 20(1), 116–128.

Arin

K.P.

, Ciferri

, & Spagnolo

(2008). The price of terror: The effects of terrorism on stock market returns and volatility. Economics Letters, 101(3), 164–167.

Aslam

, & Kang

H.G.

(2013). How different terrorist attacks affect stock markets. Defence and Peace Economics, 26(6), 634–648. doi: 10.1080/10242694.2013.832555

Brounrn

, & Derwall

(2010). The impact of terrorist attacks on international stock markets. European Financial Management, 16(4), 585–598.

Brown

S.J.

, & Warner

J.B.

(1980). Measuring security price performance. Journal of Financial Economics, 8(3), 205–258. doi: 10.1016/0304-405x(80)90002-1

Bruck

, & Wickstrom

B.A.

(2004). Economic consequences of terror: Guest editors introduction. European Journal of Political Economy, 20(2), 293–300.

Cam

(2008). The impact of terrorism on United States industries. Economic Papers, 27(2), 115–134.

10.

Chen

A.H.

, & Siems

T.F.

(2004). The effects of terrorism on global capital markets. European Journal of Political Economy, 20(2), 349–366. doi: 10.1016/j.ejpoleco.2003.12.005

11.

Chesney

, Reshtar

, & Karaman

(2011). The impact of terrorism on financial markets: An empirical study. Journal of Banking and Finance, 35(2), 253–267.

12.

Chittedi

K.R.

(2015). Financial crisis and contagion effects to Indian stock market: DCC-GARCH analysis. Global Business Review, 16(1), 50–60.

13.

Drakos

(2010). Terrorism activity, investor sentiment, and stock returns. Review of Financial Economics, 19(3), 128–135.

14.

Eckstein

, & Tsiddon

(2004). Macroeconomic consequences of terror: Theory and the case of Israel. Journal of Monetary Economics, 51(5), 971–1002. doi: 10.1016/j.jmoneco.2004.05.001

15.

Effiong

E.L.

(2014). Exchange rate dynamics and monetary fundamentals: A cointegrated SVAR approach for Nigeria. Global Business Review, 15(2), 205–221.

16.

Eldor

, & Melnick

(2004). Financial markets and terrorism. European Journal of Political Economy, 20(2), 367–386.

17.

Engle

R.F.

, & Granger

C.W.

(1987). Co-integration and error correction: Representation, estimation, and testing. Econometrica: Journal of the Econometric Society, 55(2), 251–276.

18.

Essaddam

, & Karagianis

J.M.

(2014). Terrorism, country attributes, and the volatility of stock returns. Research in International Business and Finance, 31, 87–100.

19.

Garvey

, & Mullins

(2008). Contemporary terrorism: Risk perception in the London options market. Risk Analysis, 28(1), 151–160.

20.

Global Terrorism Index (GTD) (2013). Retrieved 15 Dec 2013, from http://www.start.umd.edu/gtd/

21.

Graham

M.A.

, & Ramiah

V.B.

(2012). Global terrorism and adaptive expectations in financial markets: Evidence from Japanese equity market. Research in International Business and Finance, 26(1), 97–119.

22.

Hall

S.G.

(1989). Practitioners corner: Maximum likelihood estimation of cointegration vectors: An example of the Johansen procedure. Oxford Bulletin of Economics and Statistics, 51(2), 213–218. doi: 10.1111/j.1468-0084.1989.mp51002006.x

23.

Institute for Economics and Peace (2012). Global Terrorism Index 2012 (Report). Retrieved December 2013, from http://reliefweb.int/sites/reliefweb.int/files/resources/2012-Global-Terrorism-Index-Report.pdf

24.

Jawaid

S.T.

, & Haq

(2012). Effects of interest rate, exchange rate and their volatilities on stock prices: Evidence from banking industry of Pakistan. Theoretical and Applied Economics, 8(8), 153–166.

25.

Johansen

, & Juselius

(1990). Maximum likelihood estimation and inference on cointegration—with applications to the demand for money. Oxford Bulletin of Economics and statistics, 52(2), 169–210.

26.

Kollias

, Papadamou

, & Stagiannis

(2011). Terrorism and capital markets: The effects of the Madrid and London bomb attacks. International Review of Economics and Finance, 20(4), 532–541.

27.

Lenain

, Bonturi

, & Koen

(2002). The economic consequences of terrorism (OECD Economics Department Working Papers, No. 334). Paris: OECD Publishing.

28.

Liu

, Longstaff

F.A.

, & Pan

(2003). Dynamic asset allocations with event risk. The Journal of Finance, 58(1), 231–259.

29.

Pesaran

M.H.

, Shin

, & Smith

R.J.

(2000). Structural analysis of vector error correction models with exogenous I(1) variables. Journal of Econometrics, 97(2), 293–343. doi: 10.1016/s0304-4076(99)00073-1

30.

Peleg

, Regens

J.L.

, Gunter

J.T.

, & Jaffel

D.H.

(2011). The normalisation of terror: The response of Israel’s stock market to long periods of terrorism. Disasters, 35(1), 268–283.

31.

Pesaran

M.H.

, Shin

, & Smith

R.J.

(2000). Structural analysis of vector error correction models with exogenous I(1) variables. Journal of Econometrics, 14, 97(2), 293–343. doi: 10.1016/s0304-4076(99)00073-1

32.

Pilinkus

(2015). Stock market and macroeconomic variables: Evidences from Lithuania. Economics and Management, 14, 884–891.

33.

Ramiah

, Cam

, Calabro

, Maher

, & Ghafouri

(2010). Changes in equity returns and volatility across different Australian industries following the recent terrorist attacks. Pacific-Basin Finance Journal, 18(1), 64–76.

34.

Ramiah

, & Graham

(2013). The impact of domestic and international terrorism on equity markets: Evidence from Indonesia. International Journal of Accounting and Information Management, 21(1), 91–107.

35.

Raza

S.A.

, & Jawaid

S.T.

(2013). Terrorism and tourism: A conjunction and ramification in Pakistan. Economic Modelling, 33, 65–70.

36.

Shahbaz

(2013). Linkages between inflation, economic growth and terrorism in Pakistan. Economic Modelling, 32, 496–506.

37.

Shahbaz

, Shabbir

M.S.

, Malik

M.N.

, & Wolters

M.E.

(2013). An analysis of a causal relationship between economic growth and terrorism in Pakistan. Economic Modelling, 35, 21–29.

38.

Suleman

M.T.

(2012). Stock market reaction to terrorist attacks: Empirical evidence from a front-line state. Australasian Accounting Business and Finance Journal, 6(1), 97–110.

39.

Wong

H.T.

(2010). Terms of trade and economic growth in Japan and Korea: An empirical analysis. Empirical Economics, 38(1), 139–158.

40.

Zaighum

(2014). Impact of macroeconomic factors on non-financial firms stock returns: Evidence from sectorial study of KSE-100 index. Journal of Management Sciences, 1(1), 35–48.

41.

Ziramba

(2010). Price and income elasticities of crude oil import demand in South Africa: A cointegration analysis. Energy Policy, 38(12), 7844–7849. doi: 10.1016/j.enpol.2010.08.044