Can Google Trends Affect Sentiment Of Individual Investors? The Case Of The United States

Aleksandra Rutkowska , Agata Kliber

Abstract

As the empirical studies show, investor sentiment is a significant factor in financial markets. The large-scale development of the technology has led to widespread access to information in real time also to individual investors, which in turn has also led to the inflow of Big Data to market analysis. One of the sources of such data is the ability to track the phrases searched for in the web search engines. In our research we verify whether investor sentiment is affected by, among others, a daily Google keyword search called “Google Trends”. We consider measures of US investors’ sentiment calculated from survey studies – the AAII index. We investigate changes of sentiment and its volatility, which can be interpreted as nervousness of the market participants. We estimate a set of GARCH models with explanatory variables in conditional mean and variance. We confirm that negative keyword searches are connected with the decline of the investor confidence. The overall effect of a negative search is stronger than positive. Older searches have a weaker influence on investor sentiment than new ones – no lagged search proved to be significant.
Author Aleksandra Rutkowska (WIiGE / KMS)
Aleksandra Rutkowska,,
- Department of Applied Mathematics
, Agata Kliber (WIiGE / KMS)
Agata Kliber,,
- Department of Applied Mathematics
Journal seriesMathematical Economics, ISSN 1733-9707, (B 10 pkt)
Issue year2018
Vol21
No14
Pages51-70
Publication size in sheets0.95
Keywords in Englishinvestor sentiment, GARCH models, behavioural finance, Google Trends
DOIDOI:10.15611/me.2018.14.05
URL https://www.dbc.wroc.pl/dlibra/publication/134857/edition/70311
Languageen angielski
Score (nominal)10
Score sourcejournalList
ScoreMinisterial score = 10.0, 14-04-2020, ArticleFromJournal
Citation count*2 (2020-09-19)
Cite
Share Share

Get link to the record


* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.
Back
Confirmation
Are you sure?