Seemingly unexpected market turmoil exemplified by the recent downturn during the Coronavirus crisis has led many investors to focus on enhancing risk management.
A growing corpus of research shows that one source that can help them is news analytics data, a subset of alternative data, that has proven useful in forecasting market volatility.
Prepared by RavenPack's data science team, this whitepaper takes the form of a discussion of a selection of the research on the subject and its applications to various asset classes, including stocks, bonds, and credit.
Below are some of the key takeaways:
-
The news is independent and timely when compared to other variables - two key reasons why its integration into risk-modeling can produce more accurate results.
-
News sentiment and volume metrics are useful in forecasting equity volatility and demonstrate important explanatory power, especially when linked to investor type (retail vs institutional) and attention.
-
News analytics can uncover companies with ties through article co-mentions and these can be bundled into ‘network graphs’ to provide differentiated insights around company relationships and for appraising joint risk.
-
The observed spill-over effect from firm-level news and news about broader market drivers can be incorporated as important elements in beta forecasting and in managing the cross-sectional and time-varying properties of systematic risk.
-
Adverse market movements in corporate bonds are negatively correlated to the flow of news suggesting that analyzing news volume can help investors anticipate bond liquidity risk more accurately.
-
News analytics can provide an augmentative approach for assessing the quality of credit ratings, overall company creditworthiness and future relative performance of highly distressed firms.
-
The analysis of news covering negative-ESG events correlates with companies’ ability to manage ESG criteria, with implications for their credit-worthiness and ESG portfolio risk monitoring.