Effect of weather on security market returns
Table of Contents
TOC o “1-3” h z u HYPERLINK l “_Toc374125926″CHAPTER 2: LITERATURE REVIEW PAGEREF _Toc374125926 h 1
HYPERLINK l “_Toc374125927″2.1.Empirical review PAGEREF _Toc374125927 h 1
HYPERLINK l “_Toc374125928″2.2.Impact of cloudy cover on security market returns PAGEREF _Toc374125928 h 1
HYPERLINK l “_Toc374125929″2.2.1.Impact of sunshine on security market returns PAGEREF _Toc374125929 h 2
HYPERLINK l “_Toc374125930″2.3.Impact of temperature on security market returns PAGEREF _Toc374125930 h 4
HYPERLINK l “_Toc374125931″2.4.Differences in market sensitivity to weather PAGEREF _Toc374125931 h 6
HYPERLINK l “_Toc374125932″2.6. Conclusion PAGEREF _Toc374125932 h 6
HYPERLINK l “_Toc374125933″References PAGEREF _Toc374125933 h 7
HYPERLINK l “_Toc374125934″Appendices PAGEREF _Toc374125934 h 9
HYPERLINK l “_Toc374125935″Appendix 1: Comparative analysis on the impact of weather on market returns PAGEREF _Toc374125935 h 9
HYPERLINK l “_Toc374125936″Appendix 2: Temperature effects o market returns PAGEREF _Toc374125936 h 9
CHAPTER 2: LITERATURE REVIEWEmpirical ReviewStock markets returns have been the subject of many studies for centuries. The first seminal work on this subject can be found in Saunders his work on impact of cloud cover on market returns in 1993 (Keef & Roush, 2007b). Cao & Wei (2005) argue that in investment finance, market returns are sensitive to diverse factors within the market and outside the market. Among the diverse factors that have been investigated in these studies is the influence on weather on stock market returns. There are diverse ways that this aspect can be perceived from a researcher’s point of view. According to Cao & Wei (2005), of the aspects is evidence that there is a correlation between stock market returns and the weather conditions prevailing in a given period. However, Chang et al (2006) observe that existing empirical literature do not have a unanimous consensus on the exact impact of weather conditions on market returns. While some advocate the reasoning that in deed there is a relationship between these two variables, some object to it.
Impact of cloudy cover on security market returnsAnother seminal work supporting the arguments of Yuan, Zheng & Zhu (2006) can be traced to earlier 1990s when Saunders (Dowling & Lucey, 2005) conducted a study to determine the impact of cloud cover on market returns. Dowling & Lucey (2005) also alludes to the seminal works of Trompley which studied the same problem. More recently, cloud cover effect on market returns was performed by Keef & Roush (2007b). Their findings indicated that returns were very low during days characterized by 100% cloudy cover than days with a 0% to 20% cloudy cover. However, the differences in returns were not significant. This had also been observed by Tompley in 1997 and Saunders in 1993 as Dowling & Lucey (2005) indicate. Although not imparting a significant effect on returns, the general deductions of Dowling & Lucey (2005) and Keef & Roush (2007b) indicate that generally there is a negative correlation between cloudy cover and market returns.
Loughran & Schultz (2004) present a quantitative analysis of the impact of clouds on the stock returns in New York and find that the stocks with the highest returns (0.073%) are those studied under overcast relative to a 0.063% average return for clear skies all day. For days with scattered clouds, Loughran & Schultz (2004) found that the New York stock return was averagely 0.046% and 0.054% in cloudy days. Consequently, the percentage of clouds in the sky exhibited some effect on the changes of returns in New York. Moderate clouds (scattered was found to have significant effect on the stock returns. In regard to the economic effect of cloud cover on stock market returns, Loughram & Schultz (2004) argue that investors can enhance their margin of their returns by monitoring the trend in cloud cover. For instance, a day with a overcast will be a very busy day in trading and there will be a significant increase in stock market turnover and returns. However, the economic effect is not significant as the statistical significance because in economics, there are diverse other factors that come to play.
Impact of sunshine on security market returnsThe studies of Dowling & Lucey (2005) and Keef & Roush (2007b) are supported by Akhtari (2010) who take a different approach in evaluating cloudy cover on security market returns. He analyzes the impact of sunshine on stock markets returns using Wall Street as a case study. Akhtari (2010) uses the regression analysis model to find the relationship between cloudy cover (availability of sunshine) and the Dow Jones Industrial Average market index from 1948 to 2010 on an annual frequency.
The findings confirmed the results that were observed in the seminal works of Dowling & Lucey (2005), Keef & Roush (2007b) and Keef & Roush (2007a). There was a negative relationship between cloud cover and the gross Dow Jones Index representing market return. Logically, it can be observed that sunshine will have a positive effect on market returns. In other words, when there is full sunshine, the day is not cloudy and hence returns are high (extrapolating the findings of Dowling & Lucey (2005), Keef & Roush (2007b), Keef & Roush (2007a) and Yuan, Zheng & Zhu (2006)). In regard to modeling the impact of sunshine or cloudiness on security markets, Hirshleifer & Shumway (2003) augment the regression model used in Akhtari (2010).
Using a pooled regression approach, Hirshleifer & Shumway (2003) constrain the intercept and slope to be similar in all areas to be studied for cloudy cover or strength of sunshine. This difference in statistical modeling approach results in an outcome which indicates that there is a very significant correlation between clouds and markets returns. Clearly, there is a divergence in findings between Dowling & Lucey (2005) discussed earlier and those of Hirshleifer & Shumway (2003). However, the methodological approach by Hirshleifer & Shumway (2003) has been criticized on one area: it assumes that errors are independent which Chang et al (2006) & Keef & Roush (2007b) argue is not plausible. The regression model is given as:
EMBED Equation.3 ……Equation I
Where EMBED Equation.3 is the gross return which is measured by the Dow Jones market index in the Akhtari (2010) and Goetzmann & Zhu (2005) on day t. EMBED Equation.3 is the dummy variable representing the month, EMBED Equation.3 is the dummy variable for the specific day in a week, EMBED Equation.3 is the variable for cloudy cover and EMBED Equation.3 is the error term. However, in order to control price movement aspects, Keef & Roush (2003) and Keef & Roush (2005) recommend on having the lagged variable of returns represented by EMBED Equation.3 . Although the nature of the statistical model may change depending on number of variables and subjectivity of the researcher, this study shall utilize the general understanding that regression is a good model in determining the relation between weather and market returns.
In regard to New York, one of the earliest studies to determine the impact of weather conditions on NYSE stock indices was Saunders in late 1990s. To confirm the results of Saunders, Hirshleifer & Shumway (2001) conducted a similar study. The chi-square test statistic of logit regression was very significant at P value of 0.0033. They split the time series data into sub- periods (1982-1989) and 1990-1997. Hirshleifer & Shumway (2001) found that in the first set, the logit coefficient was -0.0136, a chi-square value of 0.68 and P- value of 0.4081. In the second set, the logit coefficient was -0.0578 a chi-square value of 11.38 and P-value of 0.0007. These findings indicate that in New York, the effect of sunshine on stock returns was more significant in 1990s than in 1980s. Not much has been mentioned in regard to the economic significance of the impact of sunshine on market returns. Nevertheless, the economic significance of sun shine on market returns is great. This is because trading will be higher on a sunny day and hence returns will be higher as exemplified by the positive statistical significance explained above. Therefore, there is no difference between statistical and economic significance.
Impact of temperature on security market returnsTemperature is another indicator of weather that has been investigated by diverse researchers to decipher its impact on security market returns. One of the best literatures in this area is that of Cao& Wei (2005). Similar studies have been undertaken by Keef & Roush (2003), Loughran & Schultz (2004), Kamstra, Kramer & Levi (2003) and Pardo & Valor (2003). Temperature has been used as one of the variables in many of the studies and the findings indicate that there is a negative relationship between security market returns and the degree of temperature in the course of the day.
To some extent, there is some relationship between cloudy cover, sunshine and temperature. In other words, days with full cloudy cover are assumed to be colder than those with full sunshine which are presumed to have high temperatures. However, there is some confusion in results when the above empirical studies are assessed critically. In other words, if high temperatures result in low returns (negative relationship) and high temperature is associated with sunny days which, according to Akhtari (2010) have positive relationship, then clearly there is confusion or lack of consensus. Nonetheless, these two results will be expected during the study and contrary finding will not be unusual.
According to Zadorozhna (2009), temperature has diverse effects on returns depending on the market and even economic situation. He argues that in his study to find the impact of weather elements on stock returns, there are many significant positive and negative relationships. However, temperature and evaporation was observed to have an insignificant relationship between temperature and stock returns at 5% confidence level. While there is no literature that could be found which explicitly and succinctly illustrates the impact of temperature on stock returns especially in New York, Worthington (2006) presents one of the best illustrations. Focusing on the decomposition of returns in the last row, first column, the average return in winter which is the coldest season is positive (-0.0262). However, in summer when temperatures are expected to be highest, the average returns are positive (0.0621). The return in the hottest season is higher than the cold season as postulated by the findings of Worthington (2006). However, a general correlation between these two variables (temperature and stock returns) is negative. Worthington (2006) indicates that periods/seasons of prolonged temperature increase will have a significant and adverse effect on returns of certain market portfolios. For example, agriculture based companies’ returns may be adversely affected. If an index like the NYSE Composite consists of significantly high number of firms in sectors easily affected by temperature, the general effect will be observed in the whole market returns.
Differences in market sensitivity to weatherAlthough it has been observed in the antecedent discussion there is clearly a relationship, whether positive or negative, between weather indicators and market returns, there is some confusion. Especially in the subsection on temperature, it was observed that arguments of Akhtari (2010) if extrapolated did not fit those of Cao& Wei (2005). On one hand, there was a report on positive correlation between temperature and returns while on the other there was evidence indicating otherwise. Zadorozhna (2009) clears this confusion and indicates that this is because of difference is sensitivity of stock markets to the elements of weather. He summarizes his studies in appendix 1 below after conducting a comparative assessment of many different markets. The markets were: Romania (BET), Russia (RTS), Ukraine (PFTS), Latvia (RIGSE), Romania (BET), Bulgaria (SOFIX) Croatia (CROBEX) and Ukraine (PFTS).
The results by Zadorozhna (2009) above indicate that there are diverse responses of stock market returns to different elements. For example, temperature up, which represents high temperature recorded, exhibit a positive relationship in some markets and negative in others.
In summary, diverse elements of weather that have been assessed in the antecedent discussion such as cloudiness, sunshine and temperature culminate in one conclusion: evidence of a relationship between weather and market returns. Nonetheless, the lack of consensus on one deduction in each weather sub- element is due to diversity in sensitivity of different markets. For instance, one empirical study indicates that there is negative relationship between temperature and returns while another argues that sunshine is positively correlated with returns. However, the diversity in market sensitivity has saved these literatures from a possible critique.
ReferencesAkhtari, (2010). Reassessment of the Weather Effect: Stock Prices and Wall Street Weather. The Michigan Journal of Business. 1: 51-70
Cao, M., & Wei, J. (2005). Stock market returns: A note on temperature anomaly, Journal of Banking and Finance 29: 1559–1573.
Chang, T., et al. (2006). Are stock market returns related to weather effects? Empirical evidence from Taiwan, Physica A 364: 343–354.
Dowling, M., & Lucey, B. (2005). Weather, biorhythms, beliefs and stock returns – Some preliminary Irish evidence, International Review of Financial Analysis 14: 337–355.
Goetzmann, W. & Zhu, N. (2005), Rain or shine: Where is the weather effect? European Financial Management 11: 559–578.
Keef, S. &. Roush, M. (2003). The weather and stock returns in New Zealand, Quarterly Journal of Business and Economics 41: 61–79.
Keef, S. &. Roush, M. (2005). Influence of weather on New Zealand financial securities, Accounting and Finance 45: 415–437.
Keef, S. &. Roush, M. (2007a). Daily weather effects on the returns of Australian stock indices, Applied Financial Economics 17: 173–184.
Keef, S. &. Roush, M. (2007b). A meta-analysis of the international evidence of cloud cover on stock returns, Review of Accounting and Finance 6: 324–338.
Hirshleifer, D., & Shumway, T. (2003). Good day sunshine: Stock returns and the weather, Journal of Finance 58: 1009–1032.
Kamstra, M., Kramer, M. & Levi, M. (2000). Losing sleep at the market: The daylight-savings anomaly. American Economic Review 90: 1005–1011.
Kamstra, M., Kramer, L. & Levi, M (2002). Losing sleep at the market: The daylight saving anomaly: Reply, American Economic Review 92: 1257–1263.
Kamstra, M., Kramer, M. & Levi, M. (2003). Winter blues: A sad stock market cycle, American Economic Review 93: 324–343.
Loughran, T. & Schultz, P. (2004). Weather, Stock Returns, and the Impact of Localized Trading Behavior. Journal of Financial and Quantitative Analysis, 39(2), 343-364
Lucey, B., & Dowling, M. (2005). The role of feelings in investor decision-making, Journal of Economic Surveys 19: 211–237.
Pardo, A., & Valor, E. (2003). Spanish stock returns: Rational or weather-influenced? European Financial Management 9: 117–126.
Yuan, K., Zheng, L. & Zhu, Q. (2006). Are investors moonstruck? Lunar phases and stock returns. Journal of Empirical Finance, 13(1): 1-23
-781050403225AppendicesAppendix 1: Temperature effects o market returnsTable 1: Temperature and stock return performance (source: Worthington, 2006)