Market sentiment
Market sentiment is usually considered as a contrarian indicator: what most people expect is a good thing to bet against.
[6] The stock market's demonstration of the situation is often described as all boats float or sink with the tide, in the popular Wall Street phrase "the trend is your friend".
Whaley (2001)[13] and Baker & Wurgler (2007)[9] suggest Chicago Board Options Exchange (CBOE) Volatility Index (VIX) as an alternative market sentiment measure.
The study shows that retail investor transactions "...are systematically correlated — that is, individuals buy (or sell) stocks in concert".
Initial public offering (IPO) of a company generates a big amount of information that can potentially be used to proxy investor sentiment.
It is not surprising that high investments in advertisement of a particular company results in a higher investor attention to corresponding stock (Grullon et al. (2004)[24]).
In particular, according to Da et al. (2014):[14] "Although market-based measures have the advantage of being readily available at a relatively high frequency, they have the disadvantage of being the equilibrium outcome of many economic forces other than investor sentiment."
[27] A sample of 1000 households with total investments equal or higher than $10,000 are interviewed to construct UBS/Gallup Index of Investor Optimism.
[28] Mentioned above survey-based sentiment indexes were reported to be good predictors for financial market indicators (Brown & Cliff (2005)[29]).
Under the third direction, researchers propose to use text mining and sentiment analysis algorithms to extract information about investors’ mood from social networks, media platforms, blogs, newspaper articles, and other relevant sources of textual data (sometimes referred as news analytics).
A thread of publications (Barber & Odean (2008),[12] Dougal et al. (2012),[31] and Ahern & Sosyura (2015)[32]) report a significant influence of financial articles and sensational news on behavior of stock prices.
Tetlock (2007)[33] suggests a successful measure of investors’ mood by counting the number of "negative" words in a popular Wall Street Journal column "Abreast of the market".
The usual way to analyze the influence of the data from micro-blogging platforms on behavior of stock prices is to construct special mood tracking indexes.
The easiest way would be to count the number of "positive" and "negative" words in each relevant tweet and construct a combined indicator based on this data.
Nasseri et al. (2014)[36] reports the predictive power of StockTwits (Twitter-like platform specialized on exchanging trading-related opinions) data with respect to behavior of stock prices.
An alternative, but more demanding, way is to engage human experts to annotate a large number of tweets with the expected stock moves, and then construct a machine learning model for prediction.
Overall, most popular social networks, finance-related media platforms, magazines, and journals can be a valuable source of sentiment data, summarized in Peterson (2016).
This approach is supported by results from Simon (1955),[43] who concludes that people start their decision making process by gathering relevant information.
Most studies are using Google Trends (GT) service in order to extract search volume data and investigate investor attention.
An increasingly important role of Internet search data is admitted in cryptocurrency (e.g. Bitcoin) prices forecasting (Kristoufek (2013a)[51]).
The authors report a strong evidence of abnormally negative stock returns after losses in major soccer competitions.
However, Abudy, Mugerman and Shust (2022)[55] document a positive stock market reaction following a victory in the Eurovision song contest.
The evidence that the number of sunlight minutes in a particular day influence the behavior of a trader is presented in Akhtari (2011)[57] and Hirshleifer & Shumway (2003).
[65][66][67] Since most retail currency traders are unsuccessful,[68] measures of Forex market sentiment are typically used as contrarian indicators.
[69] Some researchers report Internet search data (e.g. Google Trends) to be useful in predicting volatility on foreign currency markets.
