Using Google Trends and Baidu Index to analyze the impacts of disaster events on company stock prices
| Pages | 350-365 |
| DOI | https://doi.org/10.1108/IMDS-03-2019-0190 |
| Date | 05 November 2019 |
| Published date | 05 November 2019 |
| Author | Ying Liu,Geng Peng,Lanyi Hu,Jichang Dong,Qingqing Zhang |
| Subject Matter | Information & knowledge management |
Using Google Trends and
Baidu Index to analyze the
impacts of disaster events
on company stock prices
Ying Liu
School of Economics and Management,
University of Chinese Academy of Sciences, Beijing, China and
Key Laboratory of Big Data Mining and Knowledge Management,
Chinese Academy of Sciences, Beijing, China
Geng Peng
School of Economics and Management,
University of Chinese Academy of Sciences, Beijing, China
Lanyi Hu
Academy of Mathematics and Systems Science, Beijing, China, and
Jichang Dong and Qingqing Zhang
School of Economics and Management,
University of Chinese Academy of Sciences, Beijing, China
Abstract
Purpose –With the ascendance of information technology, particularly through the internet, external
information sources and their impacts can be readily transferred to influence the performance of financial
markets within a short period of time. The purpose of this paper is to investigate how incidents affect stock
prices and volatility using vector error correction and autoregressive-generalized auto regressive conditional
Heteroskedasticity models, respectively.
Design/methodology/approach –To characterize the investors’responses to incidents, the authors
introduce indices derived using search volumes from Google Trends and the Baidu Index.
Findings –The empirical results indicate that an outbreak of disasters can increase volatility temporarily,
and exert significant negative effects on stock prices in a relatively long time. In addition, indices derived
from different search engines show differentiation, with the Google Trends search index mainly representing
international investors and appearing more significant and persistent.
Originality/value –This study contributes to the existing literature by incorporating open-source data to
analyze how catastrophic events affect financial markets and effect persistence.
Keywords Stock market, AR-GARCH, Crash incidents, Search volume index
Paper type Research paper
1. Introduction
Various studies have described the impacts of disaster events on stock markets. For
example, Barrett and Davidson et al. studied the effects of aviation accidents on stock
returns in 1987; Lamb (1995) and Angbazo and Narayanan (1996) examined the
effects of hurricanes; and Shelor et al. (1992) studied earthquakes and market fluctuations
in 1991 and 1992. In recent years, the related research is also abundant. Robinson and
Bangwayo-Skeete study the economic and financial impact of hurricanes and tropical
Industrial Management & Data
Systems
Vol. 120 No. 2, 2020
pp. 350-365
© Emerald PublishingLimited
0263-5577
DOI 10.1108/IMDS-03-2019-0190
Received 30 March 2019
Revised 16 June 2019
7 September 2019
Accepted 8 September 2019
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0263-5577.htm
This work was sponsored by the National Natural Science Foundation of China under Grant Nos
71573244, 71532013, 71871210 and 71850014. The sincere gratitude goes to the anonymous reviewers
for their careful work and thoughtful suggestions that helped improve this paper substantially.
350
IMDS
120,2
storms on small island developing states. Kowalewksi and Spiewanowski (2017) examined
the relation of stock market to natural and man-made disasters and found that the
response of stock market to natural disasters is stronger. Matthew et al. studied the effect
of landfall hurricanes on stock returns and found that value stocks are more prone to
disaster risk than growth stocks. According to Lee et al. (2018), the effect of natural
disaster has spill-over effects. The result of the research showed the 2008 Sichuan
Earthquake in China caused the most substantial contagion effect in the stock markets of
neighboring Asian countries.
Traditional research on this subject has mainly employed intervention analysis and
event study to analyze the effects of such events indirectly (Kowalewksi and
Spiewanowski, 2017; Lanfear et al., 2017). Intervention analysis involves rigorous
statistical modeling to test whether a special event affects a time series factor, along with
its direction and extent (Box and Tiao, 1975). In this way, a hypothetical event is often
presented as a dummy variable. After prior specification is set, the coefficient of dummy
variables represents an intervention effect of the event. Event study is a standard
analytical technique to measure the effects of unexpected events (Mackinlay, 1997).
An event study establishes normal return models within a specific period excluding the
event to estimate cumulative abnormal returns (i.e. the difference between actual and
predicted returns during the event period) and to test their significance from 0. It is often
used to measure the unusual effects of an incident or information disclosure and the
abnormal reactions of stock prices. However, these methods are all indirect measures that
assess the ensuing effects using dummy variables. They cannot directly measure public
attention to outlying events, nor quantify their effects.
Studies in behavioral finance have extensively documented that emotions sway financial
decisions (Nofsinger, 2003). For most of the people, their online behavior can contain their
emotions. Hence, financial market analysis based on tracking online behavior has been
widely considered, and its application to stock market research has drawn scholarly
attention. For example, Mao et al. (2011) compared relationships between different emotion
indices and an average price index, trading volume and the volatility of the Dow Jones
Industrial Average. Additionally, Rao and Srivastava built a predictive model using large-
scale microblog discussions and online search data. Overall, the results showed that
tracking human behavior on the internet can advance our understanding of price
movements in oil, gold, international exchange and stock markets.
Search data present new avenues for research, as it measures the public’s attention to
unexpected events and gives timely feedback on investment dynamics. Google Trends
analyze the keywords searched by Google search users. Then, searches for keywords can be
compared with the Google Trends search volume index. The Baidu Index is a similar and
free massive data analysis service in China. It reveals the keywords of “users’attention”and
“media attention”in the previous period.
In this study, we take the Malaysia Airlines incident as an example, on March 8, 2014,
disappeared Malaysia Airlines Flight 370 attracted worldwide attention. After the accident,
Malaysia Airlines shares fell sharply at the opening of the Malaysia Stock Exchange, which
fell to 16 percent, creatingthe lowest share price in thecompany’s history. How to measurethe
impact of the event on the compan y’s stock price and how long will this impact last? We use
keywords pertaining to Malaysia Airlines’events and stocks from Google Trends and the
Baidu Indexto quantify the incidents. Weapply the composite leadingsearch index proposed
by Liu et al. (2015) to establish the search indices as a proxy variable of incidents, and use a
vector error correction (VEC) model to illustrate the effects of incidents on the airline’sstock
price, as well as the timing of the effects. Subsequently, the autoregressive-generalized
auto regressive conditional Heteroskedasticity (AR-GARCH) model is used to analyze how
incidents affect stock volatility.
351
Using Google
Trends and
Baidu Index
To continue reading
Request your trial