Sometimes I feel there’s a strong tie between technologists and hipsters—and I’ve spent the last 16 years living in Williamsburg, Brooklyn, which makes me something of an expert in the field of hipsterdom.

“Oh, you’re migrating to the cloud? How nice. We were born in the cloud—we’re cloud-native,” says the tatted-up, twentysomething, mustachioed dude drinking some pretentious cappuccino or an India Pale Ale made with passion fruit, peach, vanilla, and caramel. It’s the equivalent of, “Yeah, well I saw MGMT when the Music Hall of Williamsburg used to just be Northsix before it sold out.”

Of late, I’ve been writing a fair amount about how banks, asset managers, exchanges, and vendors are migrating to the cloud. It’s indisputable that this is the direction that the industry is heading. What is in dispute is just how real these migrations are. This can be seen in a recent story we published looking at Nasdaq’s migration of its MRX options exchange to AWS’s cloud—the first piece of a 10- to 15-year migration plan for the full exchange.

From that story:

Some argue that Nasdaq’s and CME’s shifts to AWS and GCP are not true cloud migrations, in the sense of moving everything to virtualized instances on someone else’s computer in someone else’s datacenter. Rather, say these critics, these are simply lift-and-shift exercises that don’t truly take advantage of cloud’s revolutionary promise of elastic scalability.

Nasdaq executives concede that initially the exchange’s move to AWS might not be a “true” cloud migration yet, but say the exchange is content to move slowly and incrementally. As Nasdaq’s head of alternative data, Bill Dague, told the Waters Wavelength podcast, “This [migrating to AWS] is definitely an experiment in the sense that we want to see what it takes to run an exchange cloud where the gaps are things like multicast and other really common technologies in the exchange world that are hard to do in the cloud.”

By their nature, startups have always had the advantage of being nimble. But in a world where most everything is delivered via an “as-a-service” model, being born in the cloud can serve as a major differentiator. Conversely, being that it’s easier to get a product to market faster via the cloud, the competition is fearsome. And here’s the thing: It’s not as though the incumbents are standing still—simply look at Nasdaq (or Bloomberg, Broadridge, CME, FactSet, or Refinitiv) as proof of cloud’s growing importance in the capital markets.

Just as APIs are becoming an increasingly popular data delivery mechanism in the capital markets, not all APIs are created equal. Similarly, not all cloud projects are equal. That’s a common refrain we’re hearing from industry folks when we bring up the Nasdaq or CME projects, much less bank migration projects, where there’s a lot more skepticism because of the legacy systems that are still so heavily relied upon.

As we recently wrote, mainframes are still heavily relied upon by banks. A recent whitepaper stated that 60% of those surveyed still use a mainframe, and 44% of those don’t plan on retiring their mainframe(s). Of the world’s top 50 banks, 88% still use a mainframe. And companies like IBM are still heavily investing in this staple of bank technology, with cybersecurity top of mind, as well as considerations for the eventual capabilities of quantum computing.

So when we talk about cloud migration projects, what we’re really talking about is a hybrid of lift-and-shift versus re-platforming. While it’s the latter that everyone wants to get to, going solely down that path is not usually viable for most. But without re-platforming, firms can’t take full advantage of all the wonders the cloud has to offer.

Still, talking with one cloud expert, they think critics are being a bit too harsh if they expect an exchange or bank to completely re-platform every core piece of their respective tech stacks, right off the back.

“It is true that most of the lift-and-shift migrations are not then capable of the ‘best of the cloud’, there can still be major benefits in terms of network simplification, decommissioning data centers, etcetera,” they say. “The true ‘cloudification’ comes from adopting ‘platform-as-a-service’ capabilities instead of running ‘machines’ at all. The critics might be overly harsh here as iterating the architecting post-lift-and-shift with everything already in the cloud is easier than iterating with a mixed architecture. So the critics are correct, but perhaps a little too harsh.”

With PaaS, you don’t need to “run” the machines—you just connect to a database and use it. Traditionally, you needed to worry about the operating system, the memory, the networking, etcetera. “With PaaS, you outsource all that,” they say. As such, the question becomes what gets shipped to a specialist, and what stays in-house (and, thus, will ultimately be re-platformed).

It’s a journey getting to the cloud, and banks, asset managers, exchanges, and legacy tech companies have to start somewhere. The hardware versus software debate isn’t going anywhere. The first “Waters Wrap” column I wrote, which was published March 2020, looked at how some vendors that had successful legacy systems were in the process of going through the painful task of sunsetting those systems to better embrace the cloud. While Nasdaq is starting to migrate one of its options exchanges to AWS, the process of getting all of its venues to the cloud will take anywhere from 10-15 years.

This is all to say that this is a topic we’ll be writing about for a long time. Hopefully, we’ll highlight the lessons learned as firms make these transitions. If you’re a cloud hipster or a new convert, we want to hear from you: anthony.malakian@infopro-digital.com