تمایلات بازار و قیمت سهام: مورد حوادث حمل و نقل هوایی

Behavioral economic studies reveal that negative sentiment driven by bad mood and anxiety affects investment decisions and may hence affect asset pricing. In this study we examine the effect of aviation disasters on stock prices. We find evidence of a significant negative event effect with an average market loss of more than $60 billion per aviation disaster, whereas the estimated actual loss is no more than $1 billion. In two days a price reversal occurs. We find the effect to be greater in small and riskier stocks and in firms belonging to less stable industries. This event effect is also accompanied by an increase in the perceived risk: implied volatility increases after aviation disasters without an increase in actual volatility.

Bad mood and anxiety may affect investor decisions; anxious people may be more pessimistic regarding future returns, tend to take less risk, or both. Anxiety creates a negative sentiment that can affect investment decisions and corresponding asset returns.1 In this study we examine large-scale aviation disasters. Our hypothesis is that aviation disasters affect people's mood and increase their anxiety which negatively affects the investment in risky assets. Therefore, we expect to observe negative rates of return in the stock market following aviation disasters. Indeed, we find significant evidence that aviation disasters negatively affect stock prices for a short period of a few days. The effect found in this study encompasses both an event effect and a mean-reverting reversal effect two days after the event. There is more than one possible interpretation of the investors’ reaction to news of aviation disasters: 1. Investors who are “not fully rational” (see, e.g., Lee, Shleifer, and Thaler, 1991) react irrationally to the immediate news on aviation disasters and after two days revert back to their normal behavior. It is also possible that sophisticated investors exploit the relatively low prices; hence, a price reversal occurs. 2. Investors have a state dependent utility function of the type U(C,X), where C stands for consumption and X=0,1 indicates the presence of negative sentiment following aviation disasters (X=1) or the absence of negative sentiment (X=0). Thus, if for example U(C,1) is characterized by a higher degree of risk aversion than U(C,0), our results can be explained within the expected utility framework. Yet, even in this case the switch between U(C,1) and U(C,0) falls in the category of behavioral economics, as mood affects preference and, in particular, it affects the degree of risk aversion. Both the event effect and the reversal effect are examined in this study in various ways. This study shows that the effect is highly significant and remains intact under rigorous robustness checks. Fig. 1 presents the main findings of this study, the statistical analysis of which appears in the following sections.Fig. 1 depicts the cumulative average residuals (CARs) around the dates when aviation disasters occurred. The figure shows that on the first day after a disaster (t=1), when the media are typically flooded with disturbing pictures about the event and horrible stories about casualties (rather than when the occurrence of the disaster is known to some people), there is a sharp decline in average rates of return. This decline is almost 10 times larger in absolute terms than the average daily rate of return during the observed period. This decline represents an average market loss of more than $60 billion per aviation disaster, whereas the upper bound on the actual economic loss involved with these events is roughly estimated at $1 billion per disaster. Moreover, we find that the event effect is followed by a reversal effect. On the third day after the event occurs (t=3), there is an increase in returns that is about half the magnitude of the first day's decline. This reversal tendency persists for several days afterwards; the market fully reverts back to its mean average about 10 days after the decline. What can one learn from the coexistence of the event effect and the reversal effect? If the market loss were due to the actual economic loss resulting from the disaster rather than due to the mood and anxiety effect, we would not expect to find a reversal effect at all. The fact that there is almost a complete price reversal is one more element in favor of our hypothesis asserting that excess anxiety induces the effect, and presumably when anxiety subsides or when sophisticated investors exploit the effect, a price reversal occurs. To further study the event effect, we conduct several complementary analyses. First, we show that the decline in stock prices after aviation disasters is accompanied by a corresponding increase in perceived volatility, as measured by the VIX and VXO versions of the Fear Index, which has been proposed by Baker and Wurgler (2007) as a potential proxy for market sentiment. As we do not find a similar increase in actual volatility, this suggests that anxiety following aviation disasters affects the perception of volatility. Second, motivated by the prediction of Baker and Wurgler (2006) that a sentiment effect will be larger in stocks with valuations that are highly subjective and difficult to arbitrage, we test whether there is a difference in the magnitude of the effect in portfolios constructed by volatility, size, and industry. Indeed, we not only find the effect to be highly robust and to exist in all studied portfolios, but the results also conform to Baker and Wurgler's (2006) theory; a relatively larger event effect is found in small firms, in more volatile stocks, and in the stocks of firms belonging to less stable industries. If investors’ response to the event induces a flight to safety, it can take place in various ways. It may be confined to holding cash intended for stock purchases for a few more days, or it can also induce a shift of a portion of the investments from risky assets toward safer assets such as short-term U.S. Treasury securities and the safe haven of the U.S. dollar. To test whether the effect spills over to the bond and currency markets, we search for possible effects on U.S. Treasury securities with various maturities and on the U.S. dollar exchange rates. Although we find price changes in the expected direction, the changes are insignificant. One possible explanation for this is that the flight to safety is executed through a variety of assets; hence, the event effect on each asset is diluted. An alternative explanation is that investors postpone investing in risky assets and hold more cash in their normal daily trading activity. Our findings shed new light on the role of information inflow, its psychological effect on investors’ decision-making processes, and on the way this process is corrected, i.e., the way markets become more efficient over time. The event effect is found one day after an aviation disaster has occurred and it lasts for two days. On the third day a market correction process begins and this process continues for several days. We also find a clear association between the relevancy of the disasters to U.S. investors and the magnitude of the event effect. Namely, the strongest effect is found in American-oriented disasters; a relatively weaker effect is found in European-oriented disasters, and the weakest effect is found in all other disasters. This association may also be related to the extent of public attention, media coverage, and the speed of information inflow. With regard to the speed of information inflow, we find that the event effect is seen more rapidly over the last three decades than in the previous three decades. This result is consistent with the fact that over the last three decades detailed news has become available much more quickly than in the period before that. Similarly, we find that the event effect of disasters occurring on U.S. soil has a quicker impact on the U.S. market than when the effect corresponds to faraway disasters. Finally, we show that the effect is substantially weaker in transport, industrial, and miscellaneous disasters (see Appendix A), most likely because these disasters lack some psychological resonance that we see in the case of aviation disasters. The remainder of the paper is organized as follows. Section 2 provides the theoretical background and justification for the selected variables. Section 3 describes the data, presents the events and the corresponding hypotheses tested in this study, and explains the methodology used in the empirical analysis. Section 4 presents the empirical results and the robustness tests. Section 5 concludes the paper.

In this study, we find that aviation disasters are followed by negative rates of return in the stock market accompanied by a reversal effect two days later. As the transitory decline in the stock market is more than 60 times larger than the direct economic loss, we look for an explanation of this discrepancy in the realm of behavioral economics. Indeed, psychological studies show that exposure to media coverage of aviation disasters can provoke bad mood, anxiety and fear which may induce people to be more pessimistic, not to take risks, or both. Therefore, the hypothesis of this study is that with the increased anxiety following aviation disasters there is a short-term reduction in the demand for risky assets, which in turn affects stock prices. When anxiety subsides, or when sophisticated investors exploit the effect, a reversal in the stock market takes place. We find that the effect is largest for disasters corresponding to American airline companies, smaller but still highly significant for disasters corresponding to European airline companies, and completely disappears for disasters corresponding to the rest of the world's airline companies. In addition, consistent with the prediction of Baker and Wurgler (2006) that market sentiment has a larger effect on stocks with valuations that are highly subjective and difficult to arbitrage, we find the effect to be larger for small firms, firms with more volatile stocks, and firms belonging to less stable industries. While it is possible that anxiety induces an increase in the degree of risk aversion, we find that on the event day the implied volatility, as reflected in the VIX and VXO, significantly increases, which may imply that aviation disasters also affect the perceived volatility. The hypothesis that fear and anxiety, rather than real economic factors, affect people's decisions after aviation disasters is supported by the fact that we find no evidence for a change in actual volatility after aviation disasters. Is there also a flight to safety on event days? Here the results are less clear-cut. Although on event days, on average, the yields on short-term Treasury securities decrease and the U.S. dollar strengthens against other currencies, these two changes are insignificant. It is possible that selling risky assets and investing in various relatively safe assets dilutes this phenomenon; the price of each safe asset changes in the predicted direction, but as there are a variety of relatively safe assets, the change corresponding to each asset is insignificant. The event effect is significant under various definitions of the event day. However, the empirical evidence shows that the effect is more related to the arrival of detailed and disturbing information to the public attention, than to the arrival of the first news on the event itself. Moreover, the effect corresponding to disasters that occur on U.S. soil begins earlier than the effect corresponding to disasters that occur far away. Similarly, over the last three decades, the effect begins earlier than in the three decades preceding that, probably due to the development of the communication media. An interesting area for future research is to study whether the market is efficient, namely whether one can obtain abnormal returns by executing an investment strategy based on the findings of this paper. Another possible avenue along this line is to examine the change in the price of options due to the increase in the perceived volatility, and whether a profitable position can be established in the option market when transaction costs are incorporated.