Stock Market Sentiment Analysis

Using Sentiment To Inform Investment Decisions

Research in the financial domain shows that news articles and social media can influence the stock market. Both the informational and affective aspects of news can impact stock price, trading volume, market volatility and even future company earnings.

Increasing evidence shows that, online sentiment can help predict subsequent market activity. TheySay is working with firms in the financial sector to help glean clean signals from the internet to inform financial trading, both in real-time and non-real-time settings.

The effect of news on trade prices is asymmetric, news with positive sentiment has been demonstrated to relate to large price increases albeit for a relatively short period of time. Negative news sentiment however is linked to price decreases but with more prolonged effects.

Correlation Between Sentiment and Stock Price

In 2013 a fake tweet led to the Dow Jones plummeting in just 2 minutes and once the hoax was announced the subsequent rebound took just 3 minutes. Dow Jones Industrial Average prediction also becomes more accurate when taking into account the mood across Twitter.

The use of optimistic or pessimistic language in news text influences the reader subtly. Exhaustive studies discovered that markets tend to overreact with respect to bad news.

All authors of news consciously or unconsciously introduce their own bias into their text. Readers of this news consume both the factual and affective elements of news text and react markedly to content containing greater negative sentiment.

Additionally many authors are not impartial with regard to specific stock and the bias they introduce, intentionally or not can influence the reader to trade.

It is important therefore to:-

1. obtain objective reports devoid of manual or emotional bias to inform impartial, factual analysis and trading
2. since the bias or emotion of news does affect trading we also want to track influential authors and articles to predict subsequent market activity

The efficient market hypothesis (EMH) suggests that those investors who are able to gain access to social media sentiment quickly are able to make better trades and maximise their return. Computised analysis of the news text introduces greater objectivity and reduces the authors influence and the readers reaction.

Case Studies

Algo Technology analyses stock market sentiment to inform investment decisions

See how one of our partners is using our technology at