Automated trading has already upended markets. Now AI could shake up stock picking and investment advice.

Samantha Lee/Business Insider

• Investing has long been upended by automated trading strategies driven by artificial intelligence, but the availability of data and processing techniques is increasing exponentially.
• Business Insider spoke to experts who described the challenges and opportunities that have arisen for market liquidity, stock picking, and the core process of investing itself.
• This article includes the top opportunties for AI and investing, as well as a deep dive into Yewno, which has co-created AI-driven indexes.
• Read how AI is transforming retail, transportation, media, and more in other articles from our special report, How AI is Changing Everything.

Programmatic strategies that are designed to buy or sell stocks like humans, based on specific criteria, are here to stay.

Their pervasiveness is often noticed when markets are under pressure, or when a headline triggers a flash crash that's so violent it's blamed on 'the algos.'

They were referring to algorithmic trading strategies programmed to trade stocks like humans. These methods, which incorporate artificial-intelligence technologies, are so pervasive that they've become part of the market's overall structure.

Take the February 2018 market sell-off, for example, when the Dow Jones Industrial Average lost the most points ever in a single day. The magnitude of the move was swiftly blamed on automated trading strategies like volatility targeting, which is programmed to decrease a fund's allocation to stocks as the market becomes more unstable.

That's just one example of how these strategies can worsen the magnitude of market disruptions. Marko Kolanovic, JPMorgan's global head of quant and derivatives research, has found an inverse relationship between volatility and the ease of trading - or liquidity. When machines are programmed to sell because there's plenty of selling already going on, it creates a negative feedback loop like the one that occurred last February.

But the increasing role of these strategies in investing is not all negative. After all, they strip out some of the emotional impulses that leads to wild price swings in single stocks. And they actually add some stability to the market on a day-to-day basis, according to Kolanovic.

What has and hasn't changed about investing

When dissecting the impact of AI on their industry, it's important that investors do not lose track of how things worked long before complex technologies were involved.

That's because the fundamental process of investing has not changed over time, according to Barry Hurewitz, the global head of UBS Evidence Lab, a major provider of big data sets.

Investing remains an information-processing business that requires studying competing points of view - from analysts, investors, and companies - and drawing educated conclusions.

"The core job of what it takes to make investing decisions hasn't changed," Hurewitz told Business Insider. He added, "What is changing is the amount of data that needs to be processed and the availability of data."

As data and the AI technology designed to process it have boomed, so has Wall Street's interest. One downside of this, Hurewitz said, is that some investors are borrowing strategies from successful quant firms without applying the technologies in an appropriate manner. Another danger, he added, is that trendy buzzwords are distracting investors from tried and tested strategies.

Ultimately, it's not enough to only have data at your fingertips. How you leverage AI to generate alpha, or outperformance compared to a benchmark, is more important than just having access to information, said Ruggero Gramatica, the founder and CEO of Yewno, a provider of data sets aggregated with AI technologies like machine learning.

Gramatica's firm is one of several using AI to disrupt the traditional stock-picking process.

He also flagged the boom of so-called robo advisors and investing products that allow people to automate the investing process. This disruptive trend will continue to drive the cost of investing down, he said. While it's a challenge for traditional stock pickers, it creates an opportunity for firms spearheading the new technologies.

Henry Nicholls/Reuters

Three big opportunities for AI in investing

Stock picking: Growing

The traditional stock-picking process - of poring over financial statements, meeting with management teams, and consciously making buy/sell decisions - is being disrupted by automated stock pickers. There are many examples of 'Ai-ception': creating an exchange-traded fund of companies in the AI industry and picking its constituents solely with AI.

Investment advice: Growing

The boom of so-called robo advisors and investing products that allow people to buy specific baskets of stocks is going to continue to drive the cost of investing down, according to Gramatica. That's a challenge for traditional fund managers, but an opportunity for those at the forefront of new technologies.

High-speed trading: Growing

The breed of investor that conducts high volumes of trades in mere milliseconds has been expanding rapidly in recent years - and AI-driven machine learning is a big part of that. Those tools allow computers to sift through more data than humanly possible. The main catch is that the output is only as good as the data the machines are being fed.

AI in the real world: Picking stocks without stock pickers

Gramatica's firm Yewno has co-created a handful of AI-linked indexes - including two with the Nasdaq - that track global companies in the industry, as well as the Stoxx Global AI Index. It also leveraged AI to create three cannabis indexes for Nasdaq.

What better way to populate an ETF of AI companies than through the technology itself? Yewno used a knowledge graph to aggregate reams of data sets on the AI industry, identify the most important trends, and link these to the companies spearheading them.

The returns of the indexes show that investors' enthusiasm on AI is strong. Both the Stoxx Global AI Index and the Nasdaq Yewno Global AI and Big Data Index gained 19% this year through July 11, versus a 20% gain for the S&P 500 and 25% for the Nasdaq Composite.

According to Gramatica, the application of knowledge graphs in investing - built to gather scattered data and refine it for specific queries (in this case, the best AI companies) - is the industry's biggest opportunity.

"The adoption of full-fledged knowledge graphs that are fed openly with the right amount of data and diversity of data is something that is really at the beginning," he said. "I can see that in the next five years, it will be a must-have for everyone."

It's the same technology Google uses to enhance its search results. Basically, Google's knowledge graph gathers and processes reams of information on the internet to present the most relevant answers to search queries.

"There really is too much information that is interconnected somehow," Grammatica said. "You might have an impact on your portfolio from something that is happening now in Iceland, and you won't even know. Either you have an army of analysts that keeps track of anything at any point in time, hoping to find all the connections and develop intuition, or you're always behind. You'll always miss something."

The Takeaway

Photo supplied by Yewno

"AI platforms will massively improve the ability to generate hypotheses or investment theses - and those investment theses will become more precise and will be more rich with information."

-Ruggero Gramatica, founder and CEO of Yewno