Can you tell me about your journey at Finatext?
I’ve been in finance for more than 15 years now. I started on the trading floor, engineering and maintaining trading systems before running both European and Global production tech teams across Deutsche Bank and HSBC.
For most of my career I worked for major banks and investment brokers, but as time went on I found that the drive to innovate and evolve was hampered by overly rigorous compliance and regulations put in place because the technology isn’t good enough to automate the processes needed. So I moved on to a new challenge and a change in culture. I reconnected with a former colleague in Japan who told me about his ambition to grow Finatext Ltd globally, and in 2017 we started Finatext UK Ltd, where I took the role of CEO.
Now Finatext UK is bringing their first (FCA-regulated) crypto trading app to the market – Pipster. Using simplified versions of the same type of technology used on trading floors, Pipster speaks on a level that benefits a broader community of people. There’s nothing so complex about investment that should keep the average person shut out and nothing too exclusive about trading anymore.
How are digital solutions aiding the development of your business?
Finatext UK is based here in London but we work in tandem with Finatext Ltd teams based in Japan and Vietnam. We are an image of a global online business relying heavily on the digital solutions and synergy that’s only been viable since the world became connected in this way. We consist of data science teams, product and software developers and plan to lead the industry in terms of digital solutions for trading and investments. So there’s no lack of awareness for the importance of this to our business.
What is Nowcast and how does it provide real-time data?
Nowcast is a Big Data Economics Analysis platform founded by Tsutomu Watanabe, now a Technical Advisor at Finatext Ltd, and was brought into the group in August 2016. The firm delivers global economic indicators to established financial institutions in real-time, focusing on the utility of premium data such as POS data, banking data, satellite imaging data and news-text data.
Nowcast strives to create a world where people can get an up-to-the-minute view of the economy. Its aim is to disrupt the conventional methods of economic statistic collection, the basis of policy and investment decisions, as there is an apparent ‘time lag’ between the time of survey and the time of release.
How is big data affecting investment management trends?
Rapid technological advancements have led big data to not only influence how we work, but how we consume information. It’s changing the landscape of financial trading. Machine learning has dramatically improved and today’s computer tools can interpret these huge datasets at a much faster pace, revealing previously untapped trends to make predictions and decisions that even the savviest trader wouldn’t have the capacity for.
Understanding big data has been a challenge for the investment management industry; the increasing complexity of data generation is transforming the way industries operate and this calls for the financial sector to keep up.
What we’re seeing is the use of big data and machine learning to enhance financial models and generate better returns. Big data analytics are now being used in predictive models to estimate the rates of return and make decisions on stock options in real time – something that is quite literally impossible for a trader. Financial analytics is no longer just about the examination of prices and price behaviour alone.
How do social media data and NLP (Natural Language Processing) help assess investor sentiment for stocks and currencies?
Our social sentiment indicator is programmed to understand whether, for example, a tweet’s ‘mention’ of bitcoin was made in a positive or negative light (sentiment). These are then rather simply added up en-masse and then rather complexly processed, to provide an overall score.
Instead of assessing only the ‘investor sentiment’ we take a wider-scoped ‘social pulse’ from a broad and ultimately indiscriminate group. So rather than simply taking the views of people who might be considered to be influencers or market-makers themselves, we incorporate the full demographic and refine the macro-data, to be utilised by investors – our users. We’re constantly working to further clean-up the data and tweak the algorithms to make the service more and more accurate with each evolution. It’s both a science and an art-form.
Social media data has become a new benchmark to gauge investor worth. Is this the new normal in the industry?
The establishment appreciates time-tested solutions only. So this sort of social media data analysis for investor trends will have to go on proving itself before for some time before it is considered the norm. For a generation now, we’ve had the raw data from social media piling up and the processing power to crunch almost limitless numbers. And though previously the organic nature of language made it difficult to obtain anything meaningful from the sheer volume of social media posts, now with modern techniques such as NLP we’re obtaining something of real value. Social media has been considered an immature, toxic and therefore unreliable environment for consumer and investor trends. Like all new generations though, they’re growing up and shouldn’t be ignored.
You have launched this service in Asian markets so far – do you plan on taking it to other regions? If so, where?
Our first edition of a social sentiment tool was developed in Japan for use on stock markets. With Pipster’s social sentiment tools, we are taking it to a more mature (developed) stage and deploying it to forex and cryptocurrency markets. We also plan to develop a portable sentiment product which can be plugged-in globally, to practically any financial asset. The new portable product will provide the same or similar benefits of identifying new trends and forecasting future investor behaviour.
Specifically, why the heightened interest in Asia? What does it say of the market in comparison to global markets?
Our business was founded in Japan where the retail forex market is almost completely centred on its domestic market and yet accounts for over 35% of all retail forex volume in the world. As by-and-large a technologically-progressive society, the ‘average Joe’ is simply more likely to use a hand-held device for activities such as investments and trading. We’re seeing more and more uptake for this style of activity and investment here in the UK as time progresses and the younger generations drive the consumer demands and business trends.
What are the biggest advantages of using a social data-driven model?
The world of data and modelling has changed drastically over the last few years. Every day, we generate 2.5 quintillion bytes of data, a number that wouldn’t be anywhere near as massive without the contribution of social media. To give you just one example, Facebook has more than 2 billion monthly active users and generates 4 new petabytes of data per day.
Social data provides qualitative detail and a quantitative scale. It reveals social trends and invaluable real-time insights that can make a significant impact on investor sentiment. The raw voice of consumers can be collected, providing context that simply isn’t available through traditional research methods.
The speed of social data is another factor that assists in making trading decisions. Due to the sheer speed at which updates can be posted and shared by the masses, social media has become a key way for traders and investors to stay up to date with breaking news.
With a large debate around user privacy, what are the challenges of a social data-driven model?
The user agreements for social media accounts are clear in that certain information and posts shared online, such as on Twitter, are public domain. Whether or not the mass-acquisition of this data by businesses becomes a leading topic of debate for change or regulation, is one thing. However, it is hard to see how relevant companies like Facebook and Twitter would be without the user-analytics aspects of their business available to the wider market. Monopolized access to this sort of data, by these tech giants, would also provide its own very serious ethical (let alone commercial) challenges.