NLP Case Studies
RavenPack COVID-19 Update | April 23, 2020
In this issue, media coverage plateaus as daily cases of coronavirus flatten, new research shows RavenPack data performs just as well in a crisis, and an enhancement to our news monitor enables analysis on a country-to-country basis.
Go To Coronavirus News Monitor
The general level of Covid-19 media hype continues to fall, reaching the lowest level for a month, at 53.52% of total news.
For us the 50% level is key level, with a break below potentially signaling a sea-change deceleration in both the reporting and intensity of the virus pandemic. Let’s hope that is where we are going.
The level of panic in the news continues to decline and has now reached a 6-week low of 3.92 after peaking on March 30 at 9.24. This may signal a continued decline in new cases given the indexes’s leading properties for actual cases.
Global news sentiment for coronavirus continues to base in a region between -70 and -60, out of a range of +100 (media-based optimism) to -100 (pessimism about the virus and its effects).
On April 17 sentiment fell sharply to -65 after the number of daily cases peaked again, before rebounding to -49 four days later as the daily case count once again began to slide. So far, though, it has not broken below the -70 lows established on March 31.
The list of politicians ranked on the Covid-19 news monitor’s media exposure table, shows someone called Brian Kemp in 3rd place, globally.
Brian is the governor of the state of Georgia, who has featured prominently in the news because of his decision to relax quarantine measures in an attempt to reopen the local economy.
The most talked-about pharma stock for most of the week has been Gilead Sciences Inc after trials of its remdesevir drug showed a significantly reduced mortality rate amongst severe sufferers of Covid-19.
The average mortality rate for the U.S. stands at 4.69%, whilst the mortality rate for the cohort trialing the new drug was 1.6%.
If given FDA approval the drug could become a major weapon in the fight against the coronavirus, with major profit implications for Gilean if it is manufactured in the millions and exported worldwide.
Far from underperforming during crises periods, stock portfolio investment strategies based on RavenPack news sentiment data actually perform better than a random pick portfolio, according to research produced by RavenPack data scientists .
Media hype, by which we mean the percentage of news talking about something, is highly correlated and even sometimes precedes real-world events, and this is borne out by comparisons of the hype around Covid-19 and different countries.
If you look at media hype in Italy and Spain, for example, it first rose in Italy before Spain, reflecting the former’s higher number of cases earlier on in the crisis.
The two then stabilized at a similar level and finally, most recently, hype in Spain rose to slightly above that in Italy as Spanish cases surpassed those in Italy.
This clearly reflects the complex history of the pandemic in the two countries.
This is just one example of a comparison that can be done using the enhancements made to our Covid-19 news monitor.
Other upgrades include one day change metrics on the Geo-Exposure map tooltips and Media Exposure table, time series charts for the list of Related Topics and topics widened to include Inflation, Mental Health, and Crude Oil.
Please use your business email. If you don't have one, please email us at info@ravenpack.com.
We will process your personal data with the purpose of managing your personal account on RavenPack and offering our services. You can exercise your rights of access, rectification, erasure, restriction of processing, data portability and objection by emailing us at privacy@ravenpack.com. For more information, you can check out our Privacy Policy.
Your request has been recorded and a team member will be in touch soon.
High inflation has returned in developed markets after decades of lying low. In our latest paper, we show how to build an inflation-based asset allocation strategy using sentiment data and we illustrate that sentiment-based strategies outperform models that depend merely on past observed inflation values.
This year's RavenPack Research Symposium brought two intense days of knowledge sharing in London and New York, from 25 top experts in natural language processing, quantitative investing and machine learning. Together, we explored how firms can leverage new language models to generate alpha, better manage risk and respond to calls for more sustainable investment practices.
Human capital is at the heart of value creation. Our latest research demonstrates how unprecedented workforce insights, sourced from over 200 million job postings, can generate more alpha.