Heckyl's News Sentiment Index for Nifty turns bearish, signals peaking of bull run

Equity benchmarks Nifty and Sensex have retreated around 2 percent from their record highs seen earlier this month, prompting debates on whether the market has peaked out for the time being.

While investors and experts look for trend indicators, fintech firm Heckyl's News Sentiment Index for Nifty has turned bearish for the 50-share index ending the two-month long bullish phase, pointing towards a market correction.

"Media sentiment for the broad-based index was in the bullish territory since the start of earnings season in April this year. The recent rate hike by the US Federal Reserve, sell-off in technology shares and concerns over impact of GST roll-out on companies weighed on the overall sentiment," Heckyl says.

In the stock market, price movement is hugely dictated by new information and one of the key challenges for investors and traders is to track the sentiment around the market based on news development.

"Sentiment score can give early indication of upward or downward price movement in the stock. It can also offer a profitable trading opportunity by quickly identifying a gap between perception and actual reality," Heckyl says.

The green and red shade in the above graph represent bullish and bearish sentiment respectively and are determined in the following manner.

When short-term sentiment (30-day average) rises above long-term (90-day average), green shading indicates likely uptrend for the Nifty.

When short-term Sentiment (30-day average) falls below long-term (90-day average), pink shading indicates likely downtrend for the 50-share index.

Every news item captured is tagged to a company, currency pair or a commodity. For each news item a sentiment is assigned using Natural Language Processing (NLP) techniques by Heckyl's proprietary Sentiment Analysis Engine. The sentiment can be positive, negative or neutral.

Heckyl’s Sentiment Analysis Engine uses predefined keywords, news categories (eg. M&A, Bankruptcy, Product Launch), business rules and relative positions of the keywords in the news title to determine the sentiment for a particular news item.

Follow @shukla_05sid