News is the prime factor which affects prices of financial assets, everything else is secondary. However, owing to the huge volume of news information continuously released by modern electronic communication, it becomes increasingly difficult to process all the information in a timely manner.
To track news events, a computer system would do a better job than a pair of human eyes, given its ability to (i) respond immediately in microseconds, (ii) process vast amount of information and (iii) do it with no downtime.
The challenge is therefore in building smart linguistic analytical tools which can convert a piece of news text to quantified numbers – numbers which a trading strategy can use to make objective quantitative trading decisions.
This presentation looks at the methodologies by which news texts are converted into a set of numbers. The various quantified factors are then explored in broad detail – the ways in which they are calculated, probable pitfalls in their usage and how to avoid them.
Key Factors in News Based Trading
- Market Impact capability (a.k.a. News Type)
- Secondary factors like volume of news, search engine trends, social media
In the next section of the presentation, the profitability of quantified news analytics based trading strategies is discussed. Profitability is studied w.r.t. various holding periods, different categories of news events (soft/hard, scheduled/unscheduled), the sensitivity with respect to equity sectors, stock beta scores, market VIX, the market capitalization of the stock, etc.
In the last segment of the presentation, we look at pitfalls and case studies of failures. The market impact of some of these failures also highlights the extent to which news analytics has become mainstream in some of the developed trading geographies.