Understanding the Market Impact of News Sentiment Signals: From High-Frequency Event-Driven Signals to Low-Frequency Macro-Sentiment Indicators
Date: 15 April 2015, Wednesday
Time: 3.00PM – 4.00PM BST
Presenter: Elijah DePalma, Thomson Reuters
About the webinar:
Financial markets are becoming increasingly efficient at incorporating news information into security market prices.
For scheduled economic news releases the latencies of market reactions are on the order of milliseconds and microseconds. Using a record of microsecond, time-stamped tick data from Thomson Reuters Tick History, we take a granular look at the market impact of US economic news surprises on trading and market by price level activity surrounding liquid index futures contracts.
For unscheduled news events, market reactions can be delayed by seconds or minutes. Thomson Reuters News Analytics (TRNA) is a natural language processing system that provides real-time linguistic and sentiment analytics on financial news, and TRNA can be used as an HFT event detection system to identify market moving news events. In this study we use TRNA over US and UK markets to identify news events which were followed by abnormal price returns, volatilities, and trading volumes at the one-second level. In addition, TRNA is expanding to include native Japanese natural language processing capabilities and using these capabilities, we identify Japanese language news events which significantly impacted JP equity markets at the one-second level.
We can also use TRNA to construct market-wide, macro-sentiment indices by aggregating news sentiment scores over broad universes of companies or asset classes. We present recent research which demonstrates the significant, macro-behavioral influence of market sentiment on the future performance of market anomalies and fundamental style factors over monthly time horizons.
Elijah DePalma is currently working in the most exciting business at Thomson Reuters – Machine Readable News and News Analytics – generating alpha over mid- to long-term trading horizons utilizing innovative quant signals from financial newswires and social media sources. He started his career with Thomson Reuters in Feb 2012, initially providing research support for Thomson Reuters MarketPsych Indices – a compelling product which provides macro-level, financial insights based on principles of modern psychology and behavioral finance. Prior to coming to Thomson Reuters, he completed a PhD in Applied Statistics from University of California, Riverside.
Registration link: https://attendee.gotowebinar.com/register/8526128097500516097