As sources and volumes of news have grown, so has the techniques to gather, extract, aggregate and categorise them. Important news can result in large positive or negative returns. However, owing to many news sources, we need to ask a fundamental question: Is news analytics profitable in every situation or are there some pitfalls that need to be avoided?
Information flow in News Analytics
Source: Applications of news analytics in finance: A review, OptiRisk Systems
Profitable Avenues for Quantitative News Analytics
Small cap firms
- Small Caps react more strongly to news sentiment as compared to Mid or Large Caps.
- Large Caps attract more attention from analysts giving singnificant media coverage.
- Small caps remain elusive, so news regarding them creates greater sentiment impact
Stocks with low Beta
- Stocks with low beta do not fluctuate a lot when the broader market index changes.
- With their low volatility and less exposure to the market, they are sensitive to sentiment fluctuation
- VIX acts as a dynamic indicator reflecting volatility or "fear" in the markets.
- When markets are relatively passive with low VIX, sudden events spike sentiment indicators.
- In more volatile markets, people react less strongly to positive news and more to negative news.
Soft vs hard news
- Soft news, which are opinion based, have a much weaker short-term reaction as compared to hard news, which are fact based.
- Hard negative news has the ability to plummet the stock prices.
Pitfalls for Quantified News Analytics
- News analytics sometimes fail if the heading and text has double negative.
- e.g. When Osama Bin Laden was killed in 2012, news bots treated this as negative and sold stocks
- Occurance of double negatives "Osama Bin Laden" and "killed" triggered trading signals
- Stale news regarding an event or a company acts as a trigger for sentiment indicators.
- E,g. An old news of 2002 regarding United Airlines' alleged bankruptcy, picked up by Google 2008
- This led to massive eradication of investor's wealth as the stocks of UA crashed immediately.
- Generally, news based traders rely on trusted news agencies.
- However, in case of hacked accounts, the news can turn out to be fake.
- E.g. Twitter account of Associated Foreign Press(AFP) was hacked in 2013 and a false news of explosion in White House was tweeted. Immediately,
- Dow Jones dropped 0.8%, only to recover later.
If you are interested in learning more about how Quantified News Analytics work and how trading strategies are designed based on that, have a look at this webinar. Also, check out our flagship course "Executive Programme on Algorithmic Trading", and get trained to start Algorithmic Trading on your own.