The holiday and Yom Kippur War sentiment effects: the Tel Aviv Stock Exchange (TASE)

Abstract

Several empirical studies reveal that holidays generally create positive sentiment in the stock market, whereas negative events, such as wars or disasters, are accompanied by negative sentiment. However, what happens if a negative event occurs on a holiday? In such a case, we expect two conflicting sentiment effects, which may cancel one another out or, alternatively, one effect may dominate the other. The stock market in Israel provides a unique laboratory in which to test these two conflicting effects, as Israel faced a horrible war on the Yom Kippur holiday in 1973—a war whose influence is still strongly felt today. Indeed, we find two robust effects: A strong and significant positive holiday sentiment effect; and a negative war sentiment effect, which dominates the positive holiday effect. These results, which show how sentiment effects are created, are general and can easily be applied to other events and other markets when conflicting sentiment effects occur.

Keywords:

• Event effect
• Holiday effect
• Market sentiment
• Behavioral finance

Acknowledgements

The authors acknowledge the helpful comments and suggestions of the anonymous referees of this journal. We thank Moshe Barak and Idit Aminov for preliminary data collection and the Kruger Center for Finance of the Hebrew University for its financial support.

Notes

§See, e.g., De Long et al. (1990), Shleifer and Summers (1990), Shleifer and Vishny (1997), Barberis et al. (1998) and Baker and Wurgler (2006, 2007).

¶Market sentiment, mood and emotions are sometimes used in the literature interchangeably. However, market sentiment is a broader notion, as it includes any misperception that can cause mispricing. The mood effect is therefore a special case of market sentiment.

⊥To claim that bad mood affects the market, it must be a given that mood affects people's behavior. Indeed, numerous psychological studies show that mood, even if it is only a transient one, can effect people's decisions in many aspects of life. Mitchell and Phillips (2007), who have reviewed numerous previous studies, conclude that even mild fluctuations in mood can have a significant influence on neural activation and cognition. Similarly, numerous psychological studies find that anxiety, fear, bad mood or depression is associated with a reduced willingness to take risks (see, e.g., Etzioni 1988, Hanock 2002, Mehra and Sah 2002, and many others).

†As explained in the next section and in the appendix, the analysis in this study is based on the period 1990–2008, for which there is formal reliable data on the TASE which can be considered to reflect actual market trading. However, in the appendix, we also briefly present the results corresponding to the pre-war period which are based on unreliable data, and are therefore presented in the appendix for illustration purposes only.

†Generally speaking, media coverage of traumatic events affects people's mood and emotions. Collimore et al. (2008) and many others find that exposure to media coverage of traumatic events provokes an anxiety level that is so strong that it may be associated with post-traumatic stress disorder (PTSD) symptomatology. Moreover, this phenomenon can be long lasting. Bernstein et al. (2007), for example, show that among people who did not suffer from any PTSD when the 9/11 disaster occurred, 5.6% of the people who watched the 9/11 anniversary news coverage suffered from PTSD. That is, the anniversary media coverage provoked PTSD one year after the event.

‡Obviously, these explanations cannot explain the Yom Kippur negative effect in the U.S. market. Indeed, adding the more recent years (2001–2007), we find the negative rates of return effect in the U.S. to be insignificant. Although one may explain the negative rates of return in the U.S. market on Yom Kippur itself by negative sentiment, the volume also significantly decreases on Yom Kippur, which is attributed to the fact that many Jewish people do not trade on this day. This, in turn, may affect prices, as volume and prices may have direct relations—and when volume decreases relative to the normal daily volume, prices decline. Thus, as noted by Frieder and Subrahahanyam (2004), the negative rates of return observed in the U.S. on Yom Kippur itself could be a technical result of the reduced volume, rather than a sentiment effect.

†The list of holidays during which the TASE is closed, and the corresponding months when these holidays may fall are: Purim (February–March), Passover (March–April), the Seventh of Passover (April), Independence Day (April–May), Shavu’ot (May–June), Tisha B’Av (July–August), Rosh Hashanah (September), Yom Kippur (September–October), Sukkoth (October), and Simchat Torah (October).

‡For a review of market anomalies and their meanings, see Schwert (2003).

§Another holiday, Simchat Torah, comes seven calendar days after Sukkoth. However, while this holiday is taken into account when we analyse the holiday effect, it is not relevant to the analysis of the Yom Kippur War effect.

†Note that as we test for the well-known holiday effect, in this case a one-tail t-test is applied.

†Note that the previous day serial correlation is also more apparent in the case of small firm stocks, which conforms to the non-synchronous explanation for this phenomenon, known to be stronger in small firm stocks.

†In search of such data, we contacted the TASE, the Israeli Central Bureau of Statistics, as well as known Israeli data companies, all of which claimed that no formal data is available.

†The Israeli Central Bureau of Statistics continued publishing daily data until 1974. Generally, no daily data is available for the period between 1975 and 1980. From 1981, the TASE published the ‘Share and Convertible Index’, which is based on a different methodology. Moreover, the daily data corresponding to the period from the late 70s to the 80s suffers from several drawbacks because it covers the notorious period spanning between 1979 and 1985, when the inflation rate in Israel reached an uncontrollable three-digit rate, exceeding 445% in 1984. Even more troublesome, from 1972, the major banks in Israel were blamed for interfering in their stock trading by illegally manipulating the banks’ stock prices. Therefore, trading during this notorious period cannot be considered ‘free market trading’ and prices did not accurately reflect assets’ values. Thus, while the data corresponding to the period of 1967–1972 is not reliable, and is based on a very small and immature market, the data corresponding to the period of 1979–1985 suffers from the same drawbacks and is, in addition, biased and related to an inefficient market.