Mind-reading the market: Hong Kong’s Amareos claims to offer investors an edge by analysing news articles and social media

Start-up uses artificial intelligence to process millions of documents from last 24 hours and paint a picture of how the market you are interested in is thinking

PUBLISHED : Wednesday, 16 March, 2016, 7:11pm
UPDATED : Wednesday, 16 March, 2016, 7:16pm

A financial technology start-up in Hong Kong has developed a tool to analyse millions of news articles and social media posts in a matter of seconds so that investors can see the psychological state of the market.

Amareos visualises market sentiment by analysing up to two million published articles to give the market sentiment of major stock indices, commodities, currencies and views on individual countries.

“What we’ve done is combine all of this using artificial intelligence to be able to digest in a day everything written in the past 24 hours and to provide that with visualisation tools into the simple market sentiment updates,” said Philippe El-Asmar, co-founder of Amareos.

The company is among a growing number of fintech start-ups offering investors a way to easily monitor the thousands of articles related to various markets and social media postings using artificial intelligence. They can then use the information to make prudent trades.

Amareos has partnered with Thomson Reuters to provide its analysis on the company’s Eikon platform. In 2014, Thomson Reuters added sentiment analysis of Twitter to Eikon.

Bloomberg integrated tweets into the data offered on its terminals in 2013 and last year expanded this with alerts on spikes of tweets related to companies and sentiment analysis.

“In the financial industry, there is a lot of room for AI to play. Besides understanding people’s sentiment, one can use machine learning to understand how to act in the best interest of the users,” said Professor Yang Qiang, head of the computer science and engineering department at Hong Kong University of Science and Technology.

“For example, algorithms used in AlphaGo can be used to judge the trends of the financial markets, and recommend the best action to take to optimise one’s financial goals,” said Yang, referring to the Google-designed AI supercomputer that recently beat the South Korean grandmaster of Go in a five-match tournament.

Launched in March 2015, Amareos currently tracks 100 individual stocks but plans to expand this to 8,000 single stocks.

The company’s data can also map how investors view countries and whether their feelings towards those states are changing.

“On China, even though the sentiment - current and outlook - is quite negative, it seems interesting that it’s not driven so much by emotions,” El-Asmar said.

“There’s a decent amount of certainty which tells me that it sounds like people are bracing for a tough road ahead.”

When a user selects the currencies options, Amareos gives the top three currencies for both positive and negative sentiment, as well as the top three currencies that are creating a buzz.

Individuals can access Amareos visualisations for US$20 a month, while institutions can buy access fof between US$10,000 and US$15,000 a month for a research-based subscription.

Hedge funds or other investors can buy access to the raw data collected by Amareos for their own simulations for US$50,000 to US$100,000 a month.

The start-up said it plans to expand its service to cover market chatter in Chinese and Japanese.

Amareos is one of eight start-ups taking part in the SuperCharger accelerator supported by Standard Chartered, Baidu and Tuspark.