- The global sentiment index on the coronavirus monitor reaches a two-and-a-half month high
- Black gold hits rock bottom - how can news analytics help oil traders turn loss into opportunity?
- With major Central Bank meetings in focus, what can text analytics data bring to the table?
Light at the end of the Tunnel?
As countries have eased measures and life starts to get back to normal, the global Covid-19 news sentiment index (GSI) which is a feature of our news monitor, has surged higher, reflecting a substantial turnaround in overall media optimism regarding Covid-19, potentially indicative of the pandemic passing its peak.
The ‘V’ shaped reversal in late March and early April at -70 (the index only goes as low as -100 from a +100 maximum) was the first sign sentiment might have hit rock bottom and be about to start a recovery.
Sentiment briefly dipped again in mid- April but since then has rebounded sharply, suggesting this could be a base from which a major longer-term recovery is mounted.
It could have major repercussions for financial markets suggesting asset prices could also further recover. This is especially likely to be the case as economic activity starts to return back to normal, and demand picks up again.
If you want to stay informed about the latest changes in coronavirus news analytics and conduct your own research feel free to access our monitor.
Can News Analytics Anticipate Oil Price Moves?
Oil fell to historic lows recently after international lockdown restrictions and recession fears annihilated demand.
Many traders were caught out by the fall, including Singaporean-based Commodity trading company Hin Leong, which went bust as a consequence.
The question is, could RavenPack news analytics have helped traders anticipate the coronavirus-crash, or not?
Research suggests it could, for example, a 2018 study using RavenPack news sentiment data showed that oil trading strategies performed profitably, especially over a short-term, 24-hour, trading horizon.
The other key finding of the study was that the strategy was even more profitable during periods of high volatility (vol) when it achieved a 21.3% annualized return compared to 9.8% where vol was not a factor, and high vol perfectly describes the current period.
The infographic below shows how well news sentiment predicted next-day returns for oil over the Covid-19 period since the start of the year.
Other studies have shown news analytics also has forecasting value over longer time horizons.
A study by Brandt and Gao , for example, found the main factor driving longer-term changes in oil was actually the outlook for economic growth in oil demand countries.
They went on to use RavenPack text analytics to analyze macroeconomic news sources relating to growth and found these could accurately forecast changes over the next 20-days.
“Oil prices rise gradually around positive GDP growth and economic growth news, and they decline around negative news,” said researchers.
Could the model be used to signal whether now is a good time to buy?
Unfortunately, although good at picking tops the model worked less well at picking bottoms since pessimism remains pervasive at market lows.
However, not all variables lost their leading qualities at market lows, news for Housing and Interest Rates, for example, still retained a leading edge even at bottoms, suggesting these might form part of a discrete model for navigating drawdowns in the future.
Can News Analytics Help Forecast Interest Rates?
Given the economic body-blow delivered by Covid-19, it is unlikely policymakers are going to feel ready to raise interest rates anytime soon, in fact, from what they have been saying at recent meetings, they are likely to continue cutting them or employing other non-standard measures to ease the fallout.
Regardless of what happens, there is a growing corpus of research that shows that news analytics can help with forecasting central bank policy moves by analyzing relevant news in the run-up to scheduled meetings.
The same model that successfully anticipated oil prices mentioned above also shows news sentiment is effective at anticipating changes in interest rates.
“Similarly, the news index seems to lead the actual increase of interest rates (FOMC announcement). The news index goes down around 2001 and rises around 2004, earlier than the actual interest rate release. This is quite interesting, as we know that information on change of interest rates is kept confidential before the FOMC meetings (see, for example, Bernile et al. (2016)).” Say Brandt and Gao, authors of “Macro Fundamental Events or Geopolitical Events”.
A similar study undertaken by researchers at the Bank of International Settlements (BIS) showed news analytics performed similarly well in predicting interest rates changes at meetings of the Bank of Indonesia (BOI).
Subscribe to COVID-19 Updates