The IMD Wrap: Dining on data, from pay-as-you-go to all-you-can-eat

Looking back, I didn’t appreciate it at the time. How could I? A kid raised on a goat farm in rural Sussex in the south of England, thrust into the hurly burly of London, working just off the crowded tourist spot filled with flashing lights and huge billboards that is Piccadilly Circus. I was just looking for a quick takeout lunch, and didn’t fancy braving the long lines for a soggy sandwich, a half-empty bag of chips and a soda at the Boots pharmacy chain.

So, I took a dark side street and found myself in front of a tiny, hole-in-the-wall Japanese restaurant. Why not, I thought. I didn’t think I’d ever had Japanese food of any variety before. I ducked under the flag draped in the doorway and was welcomed with a cup of hot tea while I waited for a tiny bento box—probably the most authentic bento box I’ve ever had in the years since.

Like I said, I didn’t appreciate it at the time. I didn’t know any better; didn’t understand how a few simple ingredients paired together with deft hands and years of experience of combining flavors and textures could be so sublime. It was like watching Anthony Bourdain—already many years into his show No Reservations—waxing lyrical about Croatia and calling himself stupid for not having visited the country earlier.

I tell you this purely for context. That was my first ever experience with sushi. My second—where I’ll start getting to the point of its relevance to this column—was courtesy of Telerate veteran Dane Thacker, who had just joined interdealer broker GFI and wanted to outline the broker’s data offerings over lunch. A somewhat posher setting than my first sushi experience, his choice of restaurant featured a conveyor-belt style of small plates—an upscale version of the model popularized by Yo! Sushi and others, where you pick what you want as it passes in front of you, and your bill is tallied based on the number and color of empty plates or bowls stacked in front of you when you’re done.

This may sound oddly familiar if you’re a data manager reconciling a data invoice. Here is some food for thought for you (and for sushi lovers when counting the plates): What if you just want one item per plate? Or want to combine different slices of your own choice of fish? What if you don’t like—and don’t want to pay for—the oshinko pickles that garnish your sea urchin or otoro?

The advantage of the pay-as-you-go model is that you can pick—and pay for—only what you need, rather than having to subscribe to bundles of services that include things you don’t need along with the things that you do, padding the bill

The concept of pay-as-you-go is more widely associated with cafeteria-style meals rather than fine dining, or low-cost mobile phones or internet cafes rather than your own 24/7 wired internet service—a cheaper option for small bites of food or data, but which would be more expensive if extrapolated to meet full-service needs.

The advantage of the pay-as-you-go model is that you can pick—and pay for—only what you need, rather than having to subscribe to bundles of services that include things you don’t need along with the things that you do, padding the bill.

So, to put this in the context of market data, imagine being able to dissect market data feeds into their constituent parts—such as specific securities, and individual exchanges on which they trade. For example, say you only want to trade a handful of securities, and you’re only interested in their activity on one particular exchange. Say you have a theory about the behavior of one stock in relation to a handful of others on one exchange. Traditionally you’ve needed to buy that exchange’s entire universe of hundreds of equities data just to get data on the handful of equities you’re interested in.

Jason Mendez, director of supplier strategy and operations at “data wrangler” Crux Informatics, says he has spoken to prospective customers who want to dissect data into “more digestible slices” such as by region or asset class—and even some who may want just a few dozen data fields instead of the myriad available. But Mendez says Crux is also seeing interest from data providers looking to understand this new trend, and to know which of their data fields clients find most valuable and whether their products need to be re-engineered to respond to how clients want to consume them.

This is one of the aims of Databento, a startup with a scary amount of industry experience at banks, hedge funds and high-frequency trading firms, which aims to make data more accessible, faster. With a few clicks, you can literally search for, select, and buy the live or historical market data you want from their website.

“Buyers want to know what data are they buying, and how much they’re spending,” says co-founder and CEO Christina Qi. “So we’re making data accessible and reducing the time it takes to get access to data.”

Its platform, launched earlier this year, offers data from about 40 exchanges and trading venues, and already has more than 1,000 users among asset managers, hedge funds, research and trading functions, fintech firms and brokerage startups, as well as technologists seeking to train artificial intelligence applications on finance.

A key element of making the data purchasing process quick and seamless is that where a data license is required, Databento automatically fills out the contract for users.

Chris Petrescu, founder and CEO of CP Capital, which provides data strategy advisory services, says this is a game-changer—especially for firms that want to move quickly and perhaps don’t have armies of lawyers poring over contracts for hours.

“Databento’s interface is so seamless. To be able to point and click and get that data out of a legacy data provider is nearly impossible,” Petrescu says. “I’ve filled out those data license agreements in the past. For small firms, to have that automated is a big deal.”

There are other players in the space: Polygon.io; Quodd—which is hosting a webinar on the closely tied topic of on-demand data this week—via its API marketplace; Theta Data, which currently offers historical data to retail and active traders, and says it is readying an institutional offering; and Able Markets. There appear to be plenty of companies willing to serve the retail markets with pay-as-you-go pricing, but not so many offerings are aimed at institutional consumers.

Perhaps that’s the economics of it: As mentioned earlier, pay-as-you-go often proves less attractive economically when your needs scale upwards.

So, for larger firms, pay-as-you-go’s addressable market may be—at least, initially—limited to new use cases that don’t warrant a major new data contract, or subtle replacements of existing services.

On the new use case side, the beauty is the elasticity of pay-as-you-go. Like the increased cloud adoption that is enabling many of these new use cases to spin up a business, try it out, test the workloads, and move it into production or scrap it, switch off the resources and start from scratch with something else, the ability to turn data on and off at will is going to be a major factor in the cost-effectiveness of testing new theses and business models, and in firms’ ability to be agile and fast in adapting to and exploiting new opportunities.

“I’m seeing people gravitate toward that for, for example, a research component. Say a hedge fund hires 50 interns; they can supply them with Databento data immediately rather than contracting for a non-standard use case with an exchange,” Petrescu says. “With the move to the cloud, it’s trending toward once you get outside the top echelon of firms, these hedge funds, emerging managers, portfolio managers and smaller investment advisors aren’t concerned about storing massive amounts of data—they just want to be able to click and go.”

Another industry exec agrees that the economics are crucial, adding that firms are unlikely to adopt new offerings if they grow their budget overall: Anything new must either reduce overall costs or replace something else.

Of course, it’s not just about cost. Suitability depends very much on the needs of each firm and the individual departments that might need data. One research analyst at a fund manager that is trying out pay-as-you-go data confirmed that the firm uses Polygon but declined to provide specific details of where and how the firm is using it.

And a data manager at a major hedge fund says that while their organization is “certainly open to the concept, especially for non-production uses (i.e., perception, oversight, or ‘tinkering’ vs. actual ‘touch the money’),” they note that they have yet to find a provider that meets their firm’s needs.

There are, of course, grey areas: Data on-demand and pay-as-you-go are not quite the same thing, though arguably you need to be able to provide one to then offer the other. And I freely admit that this discussion merely touches the tip of the iceberg—though I think in this case, it’s a good kind of iceberg to encounter.

The bottom line? I’d be shocked if any top-tier institution revealed that it’s using data on a pay-as-you-go basis at any scale. But, I think that for smaller firms with more limited portfolios that have the flexibility to think strategically about their data spend, there’s an opportunity here to take advantage of new ways of accessing old types of data. This is perhaps especially true for startup funds with a laser-focused objective, which can use this model as a stepping stone toward bigger, enterprise-wide data services as their companies and needs grow and evolve. Those firms can deploy their limited startup capital toward investments, rather than data and infrastructure, lowering their costs and allowing them to deliver better returns, faster.

And that, as Mr. Bourdain would have said, doesn’t suck.

Still, I’m open to being shocked and surprised by new and unexpected use cases. If this is an area where you have thoughts—or perhaps, you’ve already begun to put your thoughts into action—I’d love to hear from you. Write me at max.bowie@infopro-digital.com, tweet me (or “X” me or whatever you call it now), or message me on LinkedIn. And maybe you'll be in my next seaweed wrap. ... Er, I mean IMD Wrap!