A weekly review of news articles regarding Market Sentiment through Natural Language Processing
6 May 2019
This week we are going to see how AI is used to predict Market Sentiment and how News Sentiment still adds alpha
Artificial Intelligence has a tremendous potential to impact the way data is processed. AI has the power to enormously enhance the potential and capabilities of the people and processes working with it. AI provides an efficient method to quantify Market Sentiment by eliminating the emotional and reactionary filters from the process of reading Market Sentiment.
Quant Trading Conferences have become hubs of flyers from data vendors and panel discussions on News Sentiment. Members of the QTS team did not have the time and resources to build a natural language processing engine to turn raw news stories into sentiment and relevance scores. Data vendors do the job for them. Not much Alpha could be extracted due to Alpha decay that happens due to the high competition between the vendors.
By Oliver Bertold, Yukka Lab,Chief Business Development & Co-Founder at YUKKA Lab Effectively using AI requires adventurous leadership AI is the buzzword on everyone’s lips right now and with good reason—according to a recent survey by Finextra/OpenText of financial industry professionals, more than half of the responders felt that AI was mainstream now or would…
Today AI is being used in a range of applications, starting from Banking services to fraud detection apps. Usage of AI apps has triggered fears among us now that this technology is going to be replacing us a s humans.
We still have no idea of the unlimited potential of AI. AI is not just processing of a hug amount of data. Artificial intelligence is when a computer is programmed to learn from a process and adapt.
Initially when cars were invented, people were critical of its usefulness, but once its’ pitfalls were corrected and its’ features were enhanced, there was no looking. Similarly, once we solve the pitfalls of AI, it can be used to greatly enhance the numerous processes and human capabilities.
Understanding the Market Sentiment is a key investment that gives you an edge over other competitors. In the financial market that is driven by information flow, without the data processing capabilities of AI, there is no way one can handle the information overload.
Natural-language processing (NLP) and machine learning have been integrated to create Augmented Language Intelligence (ALI) to augment our abilities to analyse key factors in sentiment including global and local events, shifts in government policy, leadership changes, accidents.
Test-driving AI in the financial sector is a good start to build an AI foundation with strong engagement, innovation and decision-making, that can sustain the next 100 years or more.
Nowadays it is nearly impossible to step into a quant trading conference without being bombarded with flyers from data vendors and panel discussions on news sentiment.
It is impossible to step into a quant trading conference without viewing flyers from data vendors and panel discussions without news on sentiments. QTS has worked hard to extract such data but were unsuccessful.
News sentiment can be extracted from central, alternative, pre-processed data, which is called alpha. But nowadays, this is proving to be a difficult task as rival competition too generate similar articles that makes the data decay and hence alpha cannot be generated.
Data vendors are relied upon for this task as we cannot create a customised natural language processing engine that can generate our alpha. Two-Sigma hedge fund ($42 Billion) sponsored a news sentiment competition that received free data of 200 US stocks, its prices and volume from Thomson-Reuters. This was a leading source of news sentiment.
The evaluation criterion of the competition is effectively the Sharpe ratio of a user-constructed market-neutral portfolio of stock positions held over 10 days. This is conveniently the Sharpe ratio of the “alpha”, or excess returns, of a trading strategy using news sentiment.
That’s all for the week. We’d love to hear your thoughts on these articles and anything else data related! Email us at firstname.lastname@example.org