AlphaBig DataNewsResearchTradingtrading strategy

One news sentiment provider is all you need – not really (Part 1)

As RelateTheNews engages with prospective clients we occasionally hear the refrain “I already have a news sentiment provider” which the quick or quantitative follow with “What makes you different?” Both of these are great starting points for a conversation in which the value of proprietary data created from standard source material can truly be explored. For the purposes of this discussion I will address first the statement about already having a news provider; then speak briefly about the challenges and opportunities contained within a how are you different question.

For firms that do not employ a robust quantitative investment and risk management process one news provider may be sufficient. Yet any firm that determines a single provider of alternate market data to be enough out of hand is most definitely missing out on opportunity. Why is it so easy to think one provider of alternate market data is adequate? First is the understandable baseline that market (pricing, trading data, etc) and fundamental data can be delivered by a single provider and the firm will have all the factsso to speak. This is true because these sources of data are provided without interpretation directly from the markets, companies or regulators – often through a consolidator such as Thomson Reuters, Factset or Bloomberg. Second, is an over-simplification that just because two providers of alternate market data use the same source material (ie social sentiment or news sentiment) that the output (investing, trading or risk management signals) is the same or will provide the same form of alpha. Clearly two good reasons that may cause a firm to stop evaluation and adoption of more than one alternate market data provider. And yet, the reality is that two alternative market data firms using the exact same sources but very different proprietary analysis processes will inherently create different signals as output. These different signals may have overlap that is essentially random. They may have different leading periods or may even be contrarian indicators. The nature of the proprietary algorithm and the resulting proprietary output is exactly why firms should evaluate multiple vendors in the alternate market data arena even when they may appear on the surface to be similar. This difference in output from similar inputs can actually increase capabilities when applied effectively within the investment process.

Check back next week for part 2 of this discussion.


For all institutions seeking to employ alternate market data it is critical to understand that the value is not in the inputs but rather the unique signals generated by each provider’s proprietary analysis algorithms. We look forward to answering the crucial value and differentiation questions as well as demonstrating the alpha generating capability of RelateTheNews market news sentiment data as the evolution of alternative market data for quants and asset managers continues.