By Roy Kirby, Head of Core Products, Financial Information, SIX
Since ChatGPT burst onto the scene in late 2022, the media’s obsession with AI has been unrelenting. Almost overnight, a wave of speculation spread through financial markets, with everyone from journalists to fund managers posing lofty questions around its potential impact on the front office.
The consensus by now is that AI will indeed revolutionize stock picking and trading. Analysis suggests AI is already a half-decent stock picker. And a recent study we conducted with Coalition Greenwich, which canvassed the views of 67 global buy and sell-side firms, revealed respondents feel AI applications will offer the most value by generating better investment decisions. But despite the largest asset managers rallying to integrate AI into their strategies, it would be hyperbolic to argue the tech has radically transformed the role of traders and fund managers. The robot revolution has not arrived – at least, not yet.
The obvious question now is what may be hindering the march of the machines? The potentially hefty cost of developing or acquiring the technology is one obvious factor. The industry’s reluctance to rely too heavily on complex new tech is another. But there is a somewhat overlooked aspect of the debate – the need for more universal use of a key bridging technology between software platforms.
Attention to APIs
Application programming interface (API) technology – which acts as a bridge between different software applications, allowing them to share data seamlessly – ultimately represents the portal through which AI can access and analyze a huge variety of market data.
It pays to view API tech as a network of high-speed highways between cities. Just as these highways allow for faster, more efficient movement of goods and people, APIs enable seamless, real-time data flow between systems. This fluid exchange of information is therefore key to allowing AI to access and analyze market data from a wider array of sources.
Given the increasingly complex nature of most modern investment strategies, the ability to glean insights from a wealth of different data sources across a broadening asset class spectrum mustn’t be downplayed. This was also reflected in our research with Coalition Greenwich, which found that most firms now rely on two or three market data providers, with larger players often tapping five or more depending on the complexity of their investment and trading strategies. Simply put, firms won’t be able to fully leverage their shiny new AI tools across diverse market data sets without the proper API infrastructure in place.
Equally important in today’s fast-evolving investment landscape is the ability to access and process real-time data. This is another advantage API technology brings to new AI programmes. APIs support real-time data exchange, which is critical for AI models that require up-to-date information to generate the most accurate insights and predictions. Moreover, as high-speed trading expands and market volatility persists, the need for speed when it comes to data will only intensify.
A new age beckons
The good news for the AI bulls is that a more widespread adoption of API technology among financial institutions could be right around the corner. Revisiting our recent study with Coalition Greenwich, 70% of respondents feel the preferred market data interface will shift towards APIs over the next three to five years.
While less efficient market delivery methods like desktop solutions – which are still extremely common in the industry – will remain in use, FTP, SFTP and other types of file transfer are expected to decline. Inefficiency, added risk and the need for manual interventions associated with this legacy technology are some of the leading drivers behind the decline, but new and convincing reasons to shift to API tech emerge on an almost daily basis.
This could have dramatic implications for the use of large language models in trading and investing. It would support the development of more sophisticated AI-driven investment strategies, allowing for rapid adaptation to shifting market conditions across asset classes and their respective data sources. Indeed, it may beckon in a new era of investing and trading – one characterized by AI-assisted decision-making and seamless API connectivity.
