In this example, RavenPack news analytics is applied to forecasting a macroeconomic data release
Despite showing a rise of 167k new jobs, July’s ADP employment report came in sharply below the above 1 million expected by economists. Could investors have foreseen the miss? We find that a bigger than usual deterioration in sentiment for Donald Trump often precedes macroeconomic misforecasts, and today’s data was no exception as Trump’s sentiment had been falling sharply over the last week.
“News sentiment towards Trump is directly positively related to the state of the economy. Some of the chatter about Trump that happens during the week before a key economic release will be about that release and therefore positive/negative views expressed on that release (including opinions about whether or not the release is likely to surprise consensus expectations and in what direction) will be contributing to the Trump sentiment captured by RavenPack.” Says Inna Grinis, Senior Data Scientist at RavenPack.
“This is a good occasion to illustrate the advantage of news analytics vs traditional surveys. Consensus estimates are gathered by asking a number of economists with different forecasting accuracy for their predictions of key data releases at different points in time. By contrast, the sentiment expressed in the news the week before the release would incorporate the latest information available on the state of the economy and could therefore be a more timely and accurate predictor of the actual data release. This is something our Quant Research Team at RavenPack is exploring further.”
To find out more about how your organization can benefit from RavenPack data or to generate more meaningful macroeconomic insights - as in the example above - request a free trial.
The RavenPack Analytics Platform scans over 20,000 news sources to derive sentiment data on over 250,000 individual entities, tracking 6,800 different market-moving events, it is available on visual dashboards or via web APIs.Subscribe to US Election Updates