Several studies have examined the effect of terrorism on stock returns. In this paper, we
attempt to add to and expand upon that literature by using a sample of both terrorist and
military events, largely compiled from documents issued by the US Department of State,
over the period 1963-2012. We measure the effects of the events in our sample along
several dimensions: in the aggregate; comparatively across industries; by each event’s
predicted level of impact; by the type of event (terrorist versus military); by the location
of the attack (USA or outside the USA); and by whether the USA was, directly or by
proxy, the primary target of the attack. In addition to conrming some of the ndings of
the pre-existing literature, we also provide some new results that hopefully will add to
our body of knowledge in this area.
Studies on terrorism vary considerably, but most are specic in some way. For
example, many focus on just one or a few events, a short time period or on specic types
of events. Karolyi (2006) summarizes much of the nancial and economic research on the
effects of terrorism. This research nds that, generally, acts of terrorism do in fact hurt
the stock market. Arin et al. (2008) nd that terrorism affects both the returns and the
volatility of stocks, especially in emerging markets. Baumert et al. (2013) argue that
while the market indeed reacts to terrorism, the effect has diminished in recent years,
while Kollias et al. (2011) nd no clear pattern or change over time for data specic to the
London and Athens exchanges. Similar to Arin et al. (2008), however, the authors do nd
that the smaller market (Athens) is more sensitive to terrorist attacks than is the larger
market (London). Kollias et al. (2011) examined the 2004 Madrid and 2005 London
attacks and found widespread negative returns in Spain along with a slower market
rebound following the attack. Other studies focusing on specic events include Chen
and Seims (2004),Berrebi and Klor (2005),Abadie and Gardeazabal (2003) and Guidolin
and La Ferrara (2005).
Some studies consider the effects of terrorism from other angles. For example,
Johnston and Nedelescu (2006) study the September 11, 2001, New York City and 2004
Madrid attacks from both a market reaction and crisis management standpoint.
Chesney et al. (2011) compare terrorism to nancial crashes and natural disasters and
give investment advice on which industries offer better diversication and better
protection against the acts of terrorism on the whole.
Finally, Karolyi and Martell (2006) examine 75 attacks targeting publicly traded
rms between the years 1995 and 2002. The authors nd a negative stock market return
around such events, especially for those attacks that involved “human capital losses”
such as kidnappings. Interestingly, there is not much industry-level research on the
impact of terrorism. Much of the research that does exist in this area focuses on 9/11.
Cummins and Lewis (2003) examine the effect of the attacks on September 11, 2001, on
property/casualty insurers, Brown et al. (2004) on insurance companies, Kallberg et al.
(2005) on real estate investment trusts (REITs) and Doherty et al. (2003) also on
insurance companies. Alternatively, Drakos and Khutan (2003) examine the effect of
terrorist attacks on the tourism industry in Greece, Israel and Turkey.
This study corroborates the general ndings in the existing literature and also
provides some new results. First, we examine a broad sample of 28 large-scale terrorist
and military events occurring over a nearly 50-year period beginning with the
assassination of John F. Kennedy in 1963 and ending with the attack on the US Embassy
in Libya on September 11, 2012. The breadth of this sample differs from most of the