I spent last week listening to experts in artificial intelligence talk about what AI can and will bring to the markets and the broader world.
What is patently clear is that AI is here now and is only going to expand. As Jay Biondo, product manager, surveillance at Trading Technologies told me, it’s everywhere already – from Amazon’s Echo (usually called Alexa) to Netflix’s suggested picks. It’s part of Charles Schwab’s brokerage service which will allow its brokers to service thousands instead of hundreds of customers, according to Jack Boyle, vice president derivatives & trading support at Charles Schwab, who spoke at last week’s WealthTech conference.
It is firmly in the hedge fund space today, as funds look for new ways to generate alpha.
A May report from Future Perfect Machine named Bridgewater Associates, Renaissance Technologies, DE Shaw, Two Sigma, Winton Capital Management, Schonfeld Strategic Advisors, PDT Partners, Man Group and Citadel as big AI users and cited a Barclays PLX survey that said “62 percent of hedge funds now use some type of AI process in collecting information, finding best execution, identifying market momentum and scanning information sources for signals and obscure patterns.”
That falls in line with another study published last week from Adobe and Econsultancy of 700 senior leaders of financial services and insurance firms that showed 61 percent are using or plan to use AI in the next 12 months.
It’s well into the compliance and regulatory space, too. Trading Technologies jumped into the AI space with the purchase of Neurensic last October and announced last week that the surveillance tech was fully integrated into the TT trading platform. AI is even being applied to voice communications and can search for certain keywords – a particularly interesting development for OTC trading as well as surveillance.
The bottom line is, if you are not using artificial intelligence or planning on it, you’re outnumbered and probably outgunned. AI technologies can now incorporate articles into the algos used in trading systems. Digesting news from hundreds or thousands of information sources in microseconds can suddenly make that trading system faster and smarter than any human could, Biondo said.
And that is just one small slice of the information pie. Speaking at the WealthTech conference Morgan Slade, CEO at CloudQuant, said AI is being applied to only 50 or so of the 1500 alternative data sets available today. In the next few years, those alternative data sets may top 6,000, he said. Researchers who can mine that data fastest using AI may have the next edge.
It is also being applied to performance, helping firms parcel out a segment of a trading day and see how and where a strategy works or is failing. It can help accelerate the testing phase of a trading strategy and it can be applied to trading simulations. In other words, a technology to test the algos.
And even though AI appears to be technology that is magically popping up everywhere, it still has multiple challenges, primarily in interpreting unstructured data, Biondo said. Simple formulaic data is relatively easy for AI programs, but when it comes in various formats from different exchanges, it’s not a simple obstacle to overcome – today.
If AI can and is already doing all of this, it does bring up the question of what happens to the human dimension of this industry. The humanity of the trading floor, for all its pluses and minuses, was what drove so many people into stocks, options and futures markets. Of course those days are long gone, replaced by click trading, then algo trading and high frequency trading and now AI trading. Biondo is not a doomsayer who says all the jobs will soon vanish, but he acknowledges that some positions will be automated. That will then push the industry to find ways to use our domain knowledge more productively using AI. Legal experts can access hundreds or thousands of cases in an instant to help research a case for a client. Judges may do the same for a more just and unbiased judgement.
The fact is that AI is changing the dynamics of every industry in different ways – some positive and some worrisome. But it is here today and will be here a little more tomorrow. The question is how are we going to adapt, and what will it do to this industry?