It’s been a busy month for exchange news. First, CME Group and Google signed a unique agreement that saw the tech giant invest in the Chicago-based exchange, which will move its trading technologies to the Google Cloud. Then Adena Friedman of Nasdaq announced that the exchange would migrate its matching engine to Amazon Web Services beginning next year. AWS also added FactSet datasets to its Amazon Data Exchange.

And yesterday, December 6, the New York Stock Exchange announced a sweeping shake-up of top leadership positions. Most notably, president Stacey Cunningham, who stepped into her role in 2018, will move to the NYSE board of directors while Lynn Martin, president of Ice Fixed Income and Ice Data Services at the Intercontinental Exchange, takes over Cunningham’s seat. (See the remainder of moves that were announecd here.)

If you want to know about Martin, you can read this profile by Max Bowie from 2017 or this December 2020 story about her work to create a unified offering with fixed-income data at its core after a spate of acquisitions by Ice.

What I find most interesting about this move is that, in her professional duties, Martin has been focused on data for the better part of the last six years, while Cunningham hailed from the ops and sales side of the organization. Does this move represent a shift in priorities—namely data becoming the main focus—for the exchange?

Now, Martin does have ops in her background, too. Before she took over the data business, she ran Ice’s clearing business. Prior to that, she spent 14 years at NYSE in derivatives business development and other roles before the Ice acquisition, including a year as CEO of the NYSE Liffe US market business line. So maybe it’s simply a matter of her varied experience that led her to the job and not an indication of the growing importance of data. But keep in mind that NYSE is one of the US consolidated tape administrators, which has been a cash cow for it over the years, and there is likely some anxiety over the possibility of that revenue line disappearing under the SEC’s Sip shakeup.

To be clear, exchanges will still get paid for the data that goes to the tapes, but they’ll no longer get an additional cut for operating them. One way of looking at this is to say that they need someone with Martin’s chops to help devise new (possibly data-led, leveraging her recent experience) revenue lines to replace any impact from the Sip shake-up.

Or they think the new Sip environment presents opportunities and Martin is best poised to identify and execute on them.

Or if you want to be quite cynical, combine this long-term revenue line potentially diminishing with the fact that NYSE is an old-school equities listings marketplace that’s facing new competition from the likes of the Members Exchange, aka Memx, it could make this job a thankless, unwinnable task…

Finally, what might be even more interesting is what becomes of Cunningham. While she’s not leaving entirely, the announcement makes vague references to her starting a new chapter. That could simply mean taking a career break before moving to a new company, or maybe it means she’s being lined up for a different TBA job. Recklessly speculating, could it mean taking over for Ice CEO Jeff Sprecher should he look to relinquish the throne (unlikely?), or maybe she’s going to run some sort of spinoff or head up a TBA acquisition (longshot?).

Time will tell on all fronts. But if you have thoughts, I’m interested: anthony.malakian@infopro-digital.com.

And speaking of the importance of data professionals…

Give me a C! Give me a D! Give me an O(H-NO)!

I was speaking with the chief data officer of a large bank recently, and one of the first things he told me was this: “What we’re trying to do is put ourselves out of the CDO business and live in an environment where our value system is embodied in our technology platforms and our operating processes.”

It’s not often that you hear someone say that they want to put their current job function out of business—but it’s not so unheard of when it comes to CDOs. At the FAANGs, the title of chief data officer doesn’t exist. Granted, that’s also because they’re given fancier titles than something as pedantic as CDO, but these companies were born and bred in the cloud, and data is used differently at these organizations than it is at highly regulated financial institutions, which are run by legacy platforms and mainframes, and which have gone through numerous acquisitions over decades.

His contention was that, traditionally, the CDO role was a regulatory and compliance function in the wake of the Financial Crisis, and a data governance role. The push has been in creating a single source of truth and consolidated implementations of organizational single sources of truth. Additionally, is there referenceable integrity, i.e., can this information be verified against public sources?

These things combined set the foundation that allows for completing regulatory and compliance functions, while building proper data governance throughout the organization. But that has to become table stakes; as the need for more data and analytics tools grows, the CDO needs to become more of a business enabler that spearheads things like AI/ML projects and deploying event-modeling techniques, among a litany of other duties.

“If we do our jobs right, we can give you phenomenal insights on what’s going on around the world,” he said. “When we expose the technologies and techniques that we’re using to get from this legacy company to something where we don’t have a CDO, I get to go and do something else because we’re good at it.”

A second investment bank CDO echoed that sentiment, saying that firms need to make the role less defensive and more offensive.

“How do we move beyond the millions and billions we spend on reg compliance and lineage and justifying model explainability, and moving it to a way that becomes meaningful?”

For that CDO, the answer was to become more proactive.

He said that about four years ago, “the job was all regulatory and all defensive. What I’ve tried to do is look at the literally $100 million that we were spending on various regulatory initiatives, and I thought that there had to be a better way to do it.”

He identified four work streams across different regulatory programs that were all conducting similar data functions. He combined those work streams so that they were done once across the organization. While it didn’t solve all the bank’s problems, he said that it did help to cut costs and time working on governance.

“After that, I was given a bit more to do. Now, today, I have reg and I have commercialization—what can I do with my client data, my counterparty data? Can I create my own market data from the various bids and quotes that we do all day long? The job now also includes the tech direction.”

What that means is that the software development lifecycle, or SDLC, process at the bank now has a tollgate: the CDO, himself. In some ways, it sounds like he’s consolidating power in the way Robert Moses created New York’s roads and parks systems via the New York City Department of Parks and the Triborough Bridge and Tunnel Authority.

“I have to sign off before any program for it to go forward. If they’re spending any money that involves touching an interface or a database or a migration of an application stack, they have to come to my data design authority, where we get to stipulate how they do it and where they do it,” the second CDO said. “So think about market data in the cloud—that’s how I prioritize what data goes into the cloud. If they come and they’re going to rebuild their app stack, I say, ‘Fine, go build your app stack, but decouple all the data, and let’s put it into the cloud so that everybody else can use it.’ It’s the enterprise’s data, not the exotic equities desk’s data.”

As I’ve written about previously, modern trading firms need to own the data, and they need to contextualize it, too. That’s how you create alpha throughout an organization. Sounds simple enough, but if you don’t have the foundation in place that the first CDO talked about, you can’t do this efficiently, which drives up costs, as the second CDO pointed out.

Next phase

Whenever people talk about the CDO, it feels like the conversation is always something along the lines of creating a data-driven culture, of getting senior management buy-in, of creating proper data governance and efficient taxonomies, and other stuff that’s…well…quite boring.

But how do you get the traders and portfolio managers and data scientists and analysts to see the value of their CDO? It’s not via the regulatory and compliance siloes, and it’s not by explaining the importance of data governance and whatnot. It’s by helping them to see data across the organization—democratizing it—and helping them to contextualize it and create something from it that has never been done before.

To do that, the first bank CDO said, you should never do things for one purpose, and you should always leave budgeting room and time for ambitious, forward-looking projects.

“We’re all inundated with regulation, but I try to ensure that we’re never doing something for one reason. If I’m doing something for regulators because they’re trying to achieve market transparency, timeliness, accuracy—if I can’t flip that into something that’s good for the enterprise, then they fail and I waste a lot of time and money just checking a box.”

The CDO for a large asset manager notes that you also need to automate the governance process to allow business users to move faster.

“Process has become so heavy,” he said. “Our initial goal was to make data available to everybody—whoever wants to use it. But just for running a three-week PoC, I have a three-month cycle of governance in order to get it approved.”

And finally, a third bank CDO said that the CDO role of the future cannot just be about the data—as more workloads move to the cloud, the role needs to evolve and become a key component of infrastructure projects and migrations.

“You have to think about data not just in terms of content, but also in terms of architectural components. As we start to see all this blend together, one of the key things is to really plan our infrastructure. So it’s not just about data, you have to have access. … What we’ve learned is that sometimes we can get all the data from a place, but getting it there is only half the battle. If we don’t have all the right capabilities architecturally to enable people to use it, then it’s a long process.”

I do hope that the conversation around the CDO role does evolve. So often, it’s a boring, back- and middle-office conversation. But if data is king, the CDO will need to be treated at least like a prince by the front office—but first they must show they’re worthy of being treated like royalty by bringing in royalties.

The image accompanying the story is “Fight between a Tiger and a Buffalo” by Henri Rousseau, courtesy of the Cleveland Museum of Art’s open-access program.