The year is 2049. As usual, Mr. Ericsson wakes up, takes a peek out his window, and the soundproof glass turns the busy street outside into a newsfeed. He looks through the news and hears the signal of his automatic coffee machine indicating his espresso is ready. He taps on the screen of his smartwatch, and the newsfeed turns into a summary of his trading performance for the last 24 hours. The trading bot he developed a couple of years before indicates it has just adjusted its strategy to the current market situation and that yesterday’s loss has now been covered, with profits increasing. “Well done,” Mr. Ericsson says and heads for the kitchen.

A hundred years before, in 1949, Richard Donchain introduced the very first automated trading system that followed a set of rules to buy and sell funds. Financial managers and brokers first got access to such systems in the mid-1990s, and the first such service – Betterment, developed by Joy Stein – emerged on the free market in 2008. Since then, automated trading systems have improved with the development of the IT industry. In 2014, more than five percent of all trades on US stock exchanges originated from automated trading system orders. Sounds amazing, doesn’t it?

Modern “Skynets” trading people’s funds

Nowadays, automated trading systems, or bots, are frequently divided into two main groups:  

  • Those that are unable to independently open and close positions: They analyse the current market situation and provide traders with signals on whether to open or close the position.
  • Independent bots: They are programmed so that they do not require the supervision of the user and are fully automated, therefore opening or closing trades themselves according to the developer’s strategy.

As the progress of all new technologies is exponential, new concepts for improved automated trading programs emerge all the time, most of which focus on artificial intelligence (AI) and machine learning applications.

Machine learning (ML) is a branch of artificial intelligence and is defined by computer scientist and machine learning pioneer Tom M. Mitchell as “the study of computer algorithms that allow computer programs to automatically improve through experience.”

Artificial intelligence (AI), on the other hand, is exceptionally wide in scope. According to Andrew Moore, former dean of the School of Computer Science at Carnegie Mellon University, artificial intelligence is “the science and engineering of making computers behave in ways that, until recently, we thought required human intelligence.”

The prospects of automated trading programs based on AI and ML are huge. Big corporations invest billions of dollars in this field, thus minimising human presence in many spheres, including trading.

We face thousands of stocks to pick every day, and it’s a very daunting task. Today, by using AI, we can actually do all the number crunching and look at all the news media, the social media, blogs, and also the real-time codes, and we can basically scan thousands of stocks in real time that give you the best idea, so that’s where the technology is very good today. In our company, we built something called the Kai Score, so we look at all the fundamentals, the technicals, and also the momentum for the traders, and we come back with a score to rank every single stock…

Alex Lu,
CEO & Co-founder Kavout

There are several AI-based trading solutions on the market already. One is EquBot, developed by IBM, which aims to find the cause-and-effect relationship between markets, companies, and media. It has the ability to gain profound insight into unstructured or structured data.

Another example is Holly, which applies multiple algorithms in more than a million scenarios, filtering trades with a success rate above 60% and profits of 2:1 in order to guide traders and limit the failure rate in deals.

In 2017, there was a situation that highlighted just how much financial experts believe in AI. Goldman Sach’s NY headquarters presented figures showing that over the last decade, 600 traders had been replaced by 200 engineers, all due to the effectiveness of AI-based trading bots. But what will happen next? Will bots conquer the market – or the world?

Drafting the future of algorithmic trading

Machine learning adds a layer of intelligence on top of algorithms by providing powerful tools to extract patterns from data processed all across the globe, giving technology the opportunity to study it in real time and learn independently. The combination of machine learning with quantum computing, blockchain technology, cloud computing, big data, and other technology could possibly create a revolution in trading, such has...

  • The use of nanotechnology in trading
  • HFT trading (est. 74 nanoseconds)
  • The further reduction of latency to less than 20 nanoseconds

Future systems could study all the historical data that we have archived over the course of the entire history of trading, easily analyse it to discover the trends, and teach itself to predict future market movements while trading with multiple accounts and strategies to spread risk. If the trading strategy fails, the system’s self-learning algorithms could adjust its trading to different patterns and alter the rules to match the market conditions.

Now all traders have so many real-time news streams, and mining information from these unstructured data sets becomes very important, so we need new technology – new even to Wall Street – to handle this, but with ML and deep learning, we can now look at all these unstructured data sets and mine lots of trading insights, which we could not do before.

Alex Lu,
CEO & Co-founder Kavout