Ishan Shah is AVP and leads the content & research team at Quantra by QuantInsti, a global pioneer in the quantitative & algorithmic trading domain through its flagship educational offering - Executive Programme in Algorithmic Trading (EPAT®). QuantInsti also offers Quantra®, a self-paced intelligent learning platform, as well as Blueshift®, a platform for quantitative trading strategy development.
Sentiments refer to the opinion or feeling of the people towards a particular security or asset class. These are expressed in the blogs, discussion forums, news article and tweets. Quantifying this text-based information by assigning a positive or negative score to it is known as sentiment analysis. And making a buying or selling decision using the sentiment score is known as sentiment trading.
“News was found to have an impact on the market buying behaviour of the S&P NIFTY index futures”, according to a recent study conducted and published in IIMB Management Review.
If you are a short-term trader and using a time series model to predict future trends, then you can add sentiments to your trading framework to refine your entry and exit signals. For example, the recent corporate tax cut by the Finance Minister caused a positive sentiment, and the benchmark index Nifty rallied 5 percent +. You could have benefited from it by closing all your shorts and initiating fresh long positions. Those who were short and not using this sentiment in their trading decisions would have ended their day’s P&L in deep red. The key is to analyse the sentiment correctly and act on it faster to get maximum benefit.
A common saying is that sentiments drive the prices in short-run, but fundamentals of the business drive the price in the long-run. How can sentiments be crucial to make investment decisions?
You can assess the fundamentals of the business to make a list of stocks to buy and then filter the list based on the sentiment. For example, if you are a quality investor and objectively evaluated that housing and consumer goods demand will rise in India for the next 10 years, and housing finance and consumer goods companies will benefit from it; you can avoid buying housing finance companies as the sentiment around the sector is negative due to recent defaults by real estate developers. Similarly, you can consider buying the consumer companies such as Nestle India, HUL and Asian Paints as the sentiment is positive.
One of the key goals of the investor is to limit the drawdowns and protect the capital. Avoiding the sector or companies with negative sentiment helps to achieve that.
On the other hand, a contrarian or value investor might do exactly opposite, by buying companies with negative sentiments. However, in this case, it is difficult to differentiate whether he/she is doing value investment or catching a falling knife.
Sentiment is something that everyone wants to trade on and add an edge to their portfolios, but there are many false-positive to watch out for.
In 2013, a tweet about an explosion that injured Barack Obama from Associated Press (AP) account wiped out $130 billion in stock value. The markets soon realised that the news was fake as the Twitter account of AP was hacked and the stock prices recovered.
Also, not all positive/negative sentiments news items have an equal impact on stock prices. For example, a corporate tax cut has a higher impact on Nifty50 index than the interest rate cut announced by the US Fed.
It sounds complicated and time-consuming to remove false-positive news items, and to quantify each of these manually. There are sophisticated natural language processing (NLP) algorithms designed for this task. It is a branch of artificial intelligence and can be used to remove unreliable sources and look for confirmation from multiple reliable sources. It is also capable of quantifying the sentiment from the data, thus improving the accuracy and helping in position sizing.
These machine-learning algorithms are also used for tasks to evaluate corporate governance, competitive dynamics, management quality and economic moats of a business. These factors help in making longer-term investment decisions in the equity markets.
However, the NLP is still at a nascent stage, and a lot of development is going on in producing the cutting edge models, which help computers better understand human emotions.
The author is AVP and leads the content & research team at Quantra by QuantInsti.