How can AI help in investment decisions? And if there are challenges, how does your platform help to resolve those challenges?
As to why investors in general need AI, there are enormous amounts of data out there, and there is an ongoing battle over that available data. The industry as a whole now produces all kinds of data-based financial reporting and statements, and investors and industry players alike can buy really well-structured data as a result. AI has the ability to study massive amounts of this data and identify patterns.
Let us assume we identified a stock pattern today, and we want to figure out what to do next: buy or sell. AI can find somewhat similar patterns that existed in history and then analyze what happened right after. Knowing what happened after the pattern in the past may suggest what may happen in the future from today.
We can identify patterns for stocks, Forex, ETFs, mutual funds, and even currencies. With that said, some patterns will not work for certain stocks; that is why people need a complete picture, including discovery, testing, and a presentation of results.
What if there is a challenge and if they are having a problem identifying the patterns? How does then AI support this kind of investor?
Challenges can also be patterns. Let us assume there is a significant drop in the market today; AI can go back through historical data and find similar significant drops in the market to come to pattern-based conclusions, such as which particular stocks continue to go down and which stocks tend to quickly bounce back. And in that regard, AI helps to solve the challenges in conjunction with human involvement, where humans can take these signals and use them for making better trading decisions.
That perspective raises the question: can AI effectively trade or manage a portfolio without any human involvement? So far, there is only one recorded example, a hedge fund claiming no human involvement. In all other cases, at this moment, humans have some kind of involvement. Today, the best minds in the finance industry are working on solutions that can help interpret challenges or anomalies in the market, including significant drops or significant jumps. Beyond AI, many companies use robots to work on these solutions, too. They look at the expense ratio and come up with the best-case scenario – we are talking about the fully automated robots which can solve the challenges that arise.
Are there any security challenges in data processing of this type?
Data security challenges are the same whether AI is involved or not – you have to be secure either way. With that said, you do need to protect against the black swans – when something unexpected happens, and the AI can react and perform a problematic money maneuver.
Think about the verification challenges when people put driverless cars on autopilot, and the driverless car sees something unexpected. There is a chance it will crash, like Tesla demonstrated recently when the human fully relied on autopilot. When it comes to AI and investing a lot of money could be on the line.
So you see AI as a future of investment platforms? How is your platform leveraging AI differently?
Ans: Yes, absolutely. It is an enormous amount of power, and no human being can compete with the speed and volume of this power when applied to trade.
Here is the main difference with AI in our approach: to make it convenient for our users, we test a lot of strategies in advance, and that means that a typical investor gets access to a secure cloud. In our secure, local cloud, we run a lot of pre-calculations over different strategies. We run tens of thousands of different strategies simultaneously. We don’t know what is going to happen with these tens of thousands of strategies, but we know that if the user on our site wants to use one of them, then it is going to be pre-calculated. That way, the person has more immediate access to our data and analysis. And that is our main feature – that a person can use our AI on request.
Sergey Savastiouk, Ph.D. has served as CEO for several hi-tech start-up companies and nonprofit organizations and is currently the Founder and CEO of Tickeron, an artificial and human intelligence platform delivering unparalleled trading insights and analysis. As a retail investor, he spent 15 years developing his proprietary trading and quantitative algorithms (now Tickeron’s A.I.).