We find that the evolution of stock returns with the market is lower on spillover days compared to launch days, but higher than return satisfaction on non-launch days. We consistently find that the evolution of stock returns with the market is significantly higher on launch days than on non-launch days for both the pre- and post-Soviet breakup periods. More importantly, the difference in return between launch days and non-launch days is significantly larger in the pre- compared to the post-Soviet breakup period.
The stock return movement with the market is largest on RCI's launch days, followed by the US. Second, stock return movement with the market increases significantly more for failed launch events than for successful launches. While Huang, Huang, and Lin (2019) used large jackpot lotteries as an exogenous shock to test the theory in the Taiwanese market, the two later studies use the same shock and replicate Huang, Huang, and Lin (2019) in the Chinese markets.
We distinguish our study from previous studies by linking investor attention to the movement of stock returns with the market.
Satellite Launch Events as Exogenous Shocks to Investors’ Attention
The mean and median differences in abnormal SVI between launch and non-launch days are 2.330 and 0.991 while the corresponding differences for abnormal firm SVI are -0.041 and -0.012, respectively. We denote stock i's adjusted R2s for launch days and non-launch days as R2SL,i and R2NSL,i, respectively. Consistent with our expectation, we find that individual stocks experience an increase in return with the market on launch days.
Specifically, the return correlation coefficient is higher on launch days than on non-launch days, and the increase is statistically significant for both mean and median coefficients. The mean adjusted R2 is 0.103 and 0.085 for satellite launch and non-launch days respectively, marking a statistically significant difference of 0.019 or a 22% increase in economic terms compared to the no-launch adjusted R2. Our results provide the first set of evidence on the increase in stock market returns due to investor distraction from satellite launch events.
In this subsection, we investigate what is the primary factor driving the changes in the correlation coefficient by following Huang, Huang, and Lin (2019) to decompose the difference in the recorded correlation coefficient between launch days and non-launch days into three components, as shown below: . where SL,i and NSL,i denote the correlation coefficients between the excess returns of stock i and the market excess returns on the days of launch and non-launch of the satellite, respectively. In general, we find that stock return movements with the market are lower on the preceding and subsequent trading days than those on the official launch day, but higher than those on non-launch days. The results show that satellite launch events attract some investor attention in the days leading up to the event, and that this attraction peaks on official launch days before gradually dissipating after the event day.
In this section, we analyze the financial implications of our key findings by designing a hypothetical trading strategy that takes advantage of investor inattention on the satellite launch days. We suspect that if a stock moves strongly with the market on satellite launch day and the market price rises (falls) on that day, it is likely that the stock is on average overpriced (underpriced) due to investor inattention. According to our previously discussed conjecture, βi,o,t in Eq. 5) represents the additional sensitivity of the stock's return to market returns, which leads to the stock being overpriced on satellite launch days.
Satellite Launch and Stock Return Comovement: Cross-Border Effects
First, we separate the evolution of stock returns with the market on satellite launch days in the pre- and post-Soviet dissolution and examine the difference. Before the dissolution of the Soviet Union, the Space Race was aggressive between the two Cold War rivals, the United States and the Soviet Union. Although the space race formally ended in mid-1975, the end of the Cold War did not come until the dissolution of the Soviet Union in late 1991.
As two major rivals and two giants in space exploration, geopolitical tensions and space competition between the United States and the Soviet Union could increase investors' focus on satellite launches. Therefore, we suspect that the stock return movement with the market on satellite launch days may be larger in the period before the dissolution of the Soviet Union compared to the period after the dissolution. We then calculate stock returns using the markets on launch days in periods before and after the dissolution of the Soviet Union and on non-launch days.
The results shown in Table 6 show that stock return movements with the market on launch days for both the pre- and post-Soviet collapse periods are higher than those on non-launch days and that. For example, compared to an average correlation coefficient of 0.264 on non-launch days, the average coefficients for the Soviet and post-Soviet periods are 0.332 and 0.283, indicating increases of 26% and 7%, respectively. Furthermore, the average difference in correlation coefficient between launch days in the pre- and post-Soviet eras of 0.049 is statistically significant and economically significant, representing a 17% higher satellite return movement during the Soviet era.
The mean difference of 0.027 in adjusted R2 is economically even more astonishing as it indicates that stock returns with the market on launch days are 30% higher during the Soviet period compared to the post-Soviet period. In general, first, we find that stock returns move more with the market on all launch days than those on non-launch days. Compared to other foreign launch days, the average correlation coefficient for RCI launch days is 17% higher.
Additional and Robustness Tests
We separately re-estimate the approximate return values for these two groups and compare them to the corresponding estimates on non-launch days. These results support our hypothesis about the effect of pioneer launches on satellite-induced alignment of stock returns with the market. As before, we obtain the means, medians, and their backsweep differences for these successful and failed launches.
As expected, the returns for both types of launches are higher than those on non-launch days. Our sample has 161 manned launch events and 139 unmanned launch events for which we separately re-estimate the return motion proxies and test their differences. In addition to the expected pattern that both piloted and unmanned launches generate higher return movements with the market relative to non-launch days, we find that the return correlation coefficient for piloted launches is higher than that for unmanned launches. pilot with 0.027 in average correlation and 0.015 in average adjusted R2.
These findings support our prediction of increased public attention to astronaut space missions, resulting in greater investor inattention to the stock market and an associated increase in the movement of stock returns. Similar to Section 4.1, we use the return correlation coefficient and the adjusted R2 as proxies for the return movement; however, we substitute market excess returns for industry excess returns based on 2-digit SIC codes and estimate these proxies for each sample firm for launch and non-launch days. Consistent with our expectations, we find that individual stocks experience an increase in returns with their related industry on launch day.
However, compared to the changes in return movement with the market in Table 3, the findings in Table 9 are relatively smaller in magnitude and weaker in statistical significance, which is also consistent with our expectations and the results in Huang, Huang, and Lin. (2019). The results in Table 10 show that the increase in return movement for non-aerospace firms remains highly significant on launch event days. As expected, the increases in return movement on departure days are significantly larger in all tests, especially for the averages.
Conclusion
We then link satellite launches to the movement of stock returns and find that stock returns are more similar to market returns on launch days than on non-launch days. Furthermore, our results show that the satellite launch events lead to a gradual increase in returns with the market on days before the event, which peaks on official launch days before declining again on days after the event. We further explore the implication of our key finding by designing a trading strategy that aims to exploit the potential misconceptions resulting from investors' inattention to share specific information during the satellite launch days.
The top panel shows the Pearson correlation pattern of excess stock returns and excess market returns around satellite launch days compared to the baseline correlation for non-launch days. This table shows investor trading activity and the Google Search Volume Index (SVI) for launch and non-launch days. Panel A presents the correlation coefficients of excess stock returns and excess market returns for satellite launch and non-launch days separately.
This table presents the correlation coefficients of excess stock returns and excess market returns for satellite launch days, their preceding and following k days, and other non-launch days separately. This table presents each portfolio's abnormal returns, αp, by regressing its excess returns on the five Fama–French factors using different holding periods. Satellitew is a dummy satellite event that receives a value of one on launch days and zero on other days within the evaluation window.
This table presents the correlation coefficients of stock excess returns and market excess returns for satellite launch days during the Soviet and post-Soviet periods, and for non-launch days separately. This table presents the correlation coefficients of excess stock returns and excess market returns for satellite launches from the US, RCI countries (ie, Russia, China, and India), and other countries. This table presents the correlation coefficients of excess stock returns and excess market returns for days without a launch and for satellite launches classified as pioneer or normal, successful or failed, and manned or unmanned.
This table presents the correlation coefficients of stock excess returns and operating excess returns for satellite launch days and non-launch days separately. The adjusted R2s by regressing inventory excess returns on operating excess returns are also obtained for the respective two groups.
