By Professor Albert Wang, Auburn University, and Professor Michael Young, University of Virginia.
Abstract
Using two sources of household data, we show that in response to psychological distress induced by terrorist attacks, households significantly reduce their trading activity and equity ownership, are less likely to enter the market, and increase the value of their savings accounts. We find that the effects of attacks are stronger in months with more casualties and increased news coverage, as well in households with a male head and in those located in the same state as the attack. Further tests on security selection suggest that our findings are consistent with risk shifting and inattention as implied by an anxiety-based utility model.
Keywords: Terrorism, Stock Market Participation, Household Finance, Attention, Anxiety
1. Introduction
In recent years, behavioral finance has made significant advancesin understanding the way investor sentiment affects the stock market. (1) However, much less is known about the extent to which emotion and mood affect retail investors’ decision-making individually (Loewenstein, 2000). Understanding these effects is important, because as Campbell (2006) notes, the gap between optimal and observed investor behavior can be partially explained using nontraditional behavioral models that include loss aversion or sentimental bias. To formalize this idea, Caplin and Leahy (2001) develop a recursive utility model that incorporates anticipatory emotions into a riskaverse agent’s preference for risky assets. The model predicts that investors experiencing anxiety will hold fewer risky assets and increase their allocation of safer assets that provide more secure returns. Using increases in terrorist attacks in the United States as exogenous shocks to investors’ mood, we aim to empirically examine the effect of anxiety on household trading decisions. (2)
Terrorism provides a better setting to examine sentimental bias, in the form of psychological distress, for three reasons. First, the fallout from terrorism is largely emotional, and is characterized by increased feelings of anxiety, fear, and stress. Second, terrorist attacks avoid the economic effects that hinder previous studies using the financial crisis, natural disasters, or stock market movements as emotional shocks. (3) Third, it provides a setting with clean identification to examine the extent to which psychological distress causes investors to deviate from optimal behavior, as it is difficult to argue that attacks are expected by investors.
To determine the effects of terrorism on household trading, we use two sources of individual investor data and a comprehensive list of terrorist attacks from the past 30 years. While many previous studies on terrorism use only a handful of large events, we use a sample of 457 salient attacks. This approach is similar to Wang and Young (2019) and consistent with the findings in Drakos (2010) that attacks causing even “minor” psychosocial effects can alter behavior. A comprehensive list of terrorist attacks is taken from Enders et al. (2011), and pared down to ensure that all attacks are likely to be salient to household investors. (4)
Using brokerage data from 1991 to 1996, we find that households significantly reduce their trading activity and limit their exposure to equity markets in the month following an increase in terrorist attacks. A one standard deviation increase in the number of attacks, roughly equivalent to going from the mean of two attacks in a month to four attacks, leads active traders to reduce their monthly net trades by 20%.(5) This equates to a drop of $1,472, relative to an average monthly net trades of $7,235. Breaking net trades into buys and sells, we find that households reduce the value of both their buys and sells in response to an increase in attacks.
Using household survey data from the Panel Study of Income Dynamics (PSID), and aggregating the number of salient attacks each year from 1984 to 2013, we find that households significantly reduce their level of market participation and their share of risky assets in years with higher terrorist activity. Following a one standard deviation increase in the level of terrorism, akin to going from 9 attacks in a year to 16 attacks, there is a 2.4% drop in the likelihood that households own any equity. Additionally, households not previously invested in the market are less likely to enter. In a shift away from risky assets, households respond to attacks by decreasing the share of their risky assets by 3.3% and increasing savings by 2%. Lastly, in evidence that individuals experiencing anxiety increase their expectation of the variance of future consumption, we document a drop in durable goods consumption after attacks. (6) These initial results provide compelling evidence for the causal effects of psychological distress on risky asset ownership.
To further examine the effect of attacks on anxiety, we examine the cross section of attacks and the responses by different individuals. We start by measuring the severity of attack months in two ways: the number of people killed and total news coverage of the attacks. Consistent with larger attacks and more news coverage causing higher levels of distress, we find stronger results on trading in months with more people killed and with increased news coverage of the attacks. Using household location data, we find that the effect of terrorism is stronger in households more likely to experience psychological distress: local households and those in large metropolitan cities. (7) In addition to the location of the investor, gender and marital status are significant predictors of how individuals cope with psychological distress. Consistent with prior work, we find that married households with a man designated as the head significantly reduce their trading activity. (8)
Suggested Citation:
Wang, Yan Albert and Young, Michael, Terrorist Attacks and Household Trading (August 28, 2019). Available at SSRN: https://ssrn.com/abstract=3445583
Comments