We’ve all seen it before: Someone comes up with a brilliant idea that’s ahead of its time. They gleefully demonstrate it to the market, thinking everyone will share their vision. But typically conservative decision-makers don’t see its full potential, and so their idea becomes a hammer looking for a nail, and instead of trying to shift a stagnant base, they shift their own model, and let potential paying clients steer the product in the direction of whatever their most pressing need is at that particular moment. Compliance? Yes, it can do that. Margin calculations? Absolutely! HR systems? Um, OK … Lead generation databases? Suuuuure …

And over time, sometimes the vision gets lost in pragmatism and becomes a plain old hammer. Cloud isn’t so much a hammer looking for a nail; it’s more a Swiss Army knife with a bunch of weird attachments that you can’t figure out what they do—because you haven’t yet encountered the problems that these are designed to solve. And that’s because those problems are still a few years away. But before we look forward to those challenges, let’s first take a look back.

People have been using the cloud for at least 15 years, before Amazon Web Services, Google Cloud, and Microsoft Azure were household names and had refined their offering for financial markets. The start of the 2000s saw the birth of grid computing—massive architecture of interconnected compute resources—and this evolved into “grid-as-a-service,” and ultimately into the catchier-named “cloud.” The first time I encountered the term “cloud” was a sales rep for a reference data management vendor describing how they processed the vast amounts of data involved in creating their security master. By a decade ago, the cloud was well understood—if not widely adopted beyond web and data hosting. Nasdaq was an early adopter that sought to capitalize on cloud’s potential and to entice and reassure financial clients, creating a financial markets-dedicated service dubbed FinQloud in partnership with AWS. Though Nasdaq exited the venture, it later regrouped with AWS in 2020 to build a new market data cloud.

There were also innovators who saw greater potential early on: For example, one FinQloud client, Tradier, set out to build a brokerage-in-a-box entirely in the cloud, a model that others such as Clear Street—a startup prime brokerage founded by former Knight Capital (now Virtu) execs, which recently hired industry veteran analyst and strategist Brad Bailey as head of market intelligence—are taking to the next level today. And as early as 2012, derivatives pricing and analytics provider Numerix was leveraging the cloud to serve clients’ complex pricing requirements, harnessing the ability to “burst to cloud” during predictable and resource-intensive periods each week or month, such as for calculating credit value adjustment, where compute requirements would take much longer to perform using existing resources.

But back then, while firms were all indirectly using cloud already because their suppliers were busy using it, they still resisted it publicly—that is, until they realized cloud had the potential to address the costliest challenge of all: costs themselves.

Let’s face it: What persuaded financial firms to shift storage and compute processes en masse to the cloud—and even (gosh!) the dreaded “public” cloud—wasn’t a collective visionary realization of the future needs that data processing would require, but rather the promise of significant cost reductions over firms’ existing spend on in-house infrastructure: more power and storage without increasing—and possibly even while lowering—their budgets.

But don’t pat yourself on the back just yet. First, firms need to be just as critical of their cloud setup as they are of their internal tech spend. A simple “lift and shift” won’t automatically reduce costs—in fact, some firms have been surprised when their bills didn’t immediately plummet, and some even increased. To take advantage of what cloud offers from a cost perspective, firms need to review their usage carefully.

The beauty of cloud is its elastic and on-demand access to resources—the ability to spin up additional compute power on the fly when you need it and then shrink it again to normal levels when you don’t. If you think that having the full resources of the cloud at your disposal means you can leave the lights on 24/7, that defeats the object of the cloud altogether. And no, your bills won’t go down. Just because the cloud offers seemingly infinite resources doesn’t mean you should have them switched on all the time, because then you’ll be paying for them all the time.

Instead, weigh the costs not against today’s requirements, but against what you plan to be doing and expect to gain in several years’ time. Plan for beyond just “infrastructure for the now,” says Christine Johnson, a former Hentsu executive who is now CEO and founder of data platform startup Ingenii.

“There will be a massive technological convergence over the next three to five years between quantum computing and artificial intelligence that will allow you to compute simultaneously on a single machine and deliver more power than we can comprehend. If you’re a hedge fund manager wanting to research multiple datasets that would ordinarily take weeks, that’s huge,” Johnson says. “That ability to consume massive amounts of data in parallel is going to lead to the evolution of AI. And you can’t do that on clunky old architectures—you have to be in the cloud and using quantum.”

Finally, we’re seeing the full potential of those early visions of cloud coming to fruition, and potentially being surpassed. The next stage in cloud’s development will open new doors for firms willing to walk through them, who are prepared to look at cloud as more than a cost play.

In a couple of upcoming articles, I’ll be taking a look at the next questions raised by the evolution of some of the issues discussed in this column, such as how market data platforms of the future will take advantage of cloud, and where you “burst” to if you’re already hitting the ceiling of your cloud environment. So watch this space. And if you have thoughts on these topics, please get in touch!