Fed to track market sentiment on Twitter
The US Federal Reserve (Fed) has created a new measure of credit and financial market sentiment using Natural Language Processing on Twitter data. The Fed's report on the subject noted that the Twitter Financial Sentiment Index (TFSI) was highly correlated with corporate bond spreads and other price and survey-based financial condition measures. The report, which stated that overnight Twitter financial sentiment helped predict market returns the next day, said, "Moreover, we show that the index contains information that helps predict changes in US monetary policy. The deterioration in Twitter financial sentiment the day before the Federal Open Market Committee (FOMC) announcement estimates the size of restrictive monetary policy shocks. We also document that financial sentiment worsened in response to unexpected tightening of monetary policy." The report noted that a real-time financial sentiment index was created using Natural Language Processing on Twitter data within this scope, and that tweets about the Fed played a dominant role in terms of financial sentiment on FOMC days.