RavenPack News Sentiment Data Outperforms During Coronavirus Crisis

As the coronavirus outbreak spreads around the world, we have gone far beyond just a health crisis, with people self-isolating at home or complying with state-forced lockdowns. Businesses have suffered, and have had to either close or scale down as a consequence, thus impacting real jobs and the economy. News continues to play a key role in staying informed about the current state of the crisis, influencing both market sentiment and asset prices.

News Sentiment Data outperform random portfolio strategies during the three crisis periods, showing strong statistical significance

• A long-short U.S. large-cap strategy of buying stocks with extreme positive sentiment, and selling stocks with extreme negative sentiment delivered Information Ratios between 2.6 - 2.8 during the Global Financial Crisis
• A similar strategy delivered Information Ratios between 3.6 - 4.2 during the Chinese stock market selloff in 2015-16
• During the recent Coronavirus crisis, the strategy delivered Information Ratios between 5.8 - 6.8

Access Full White Paper

Negative Sentiment has a Greater Impact

While both negative and positive sentiment has proven predictive for future asset prices, the former tends to have a greater impact. Due to the presence of more negative news during crisis periods, it seems reasonable to assume that news sentiment should work well during such times.

In our latest paper, we look at three significant drawdown periods: the Global Financial Crisis of 2008-09, the China-driven selloff of 2015-16 including the Greek debt crisis and the devaluation of the Renminbi, and finally, the 2020 Coronavirus health crisis.

We show that sentiment-driven strategies outperform during all of these volatile periods. We validate this by running two flavors of backtest strategies on long-short, market-neutral portfolios, constructed using the RavenPack sentiment signal.

About RavenPack's Coronavirus Data-Driven Insights

Visit our Coronavirus Data-Driven Insights Hub where we are pulling together all our analytics, information and insights about the novel coronavirus.