EP1014: Opportunities for innovation

This interview talks about the financial sector is undergoing a massive shift from physical trading floors to cloud-based ecosystems that prioritise speed and global accessibility. This evolution is democratising market infrastructure, allowing smaller banks to utilise the same high-powered tools as major global institutions. However, many firms struggle with costly system fragmentation and siloed data, which act as a hidden tax on their ability to scale and innovate. To overcome these hurdles, institutions are increasingly seeking regulated technology partners like Nasdaq to provide secure, managed services and reliable governance frameworks. The future of the industry lies in the integration of AI and real-time data, which will enable more intelligent risk management and automated decision-making. Ultimately, the most successful organisations will be those that simplify their digital architecture through proactive collaboration and transparent regulatory engagement.

Transcript

SPEAKER_00 0:00

Welcome to the deep dive. I mean, if you're joining us today, you're probably looking to get thoroughly informed on capital markets without, you know, the usual information overload.

SPEAKER_01 0:08

Yeah, exactly.

SPEAKER_00 0:09

So today we are jumping right into a really fascinating interview from the February 2026 edition of the IBSI FinTech Journal.

SPEAKER_01 0:19

Right. That's the one by reporter Pujasharma.

SPEAKER_00 0:21

Trevor Burrus, Jr. Yes. She sat down with Valerie Bannert Thurner. And she is the executive vice president and chief revenue officer of financial technology at NASDAQ. So if you track capital markets at all, you know we are right in the middle of this massive architectural paradigm shift.

SPEAKER_01 0:37

Aaron Ross Powell Oh, absolutely. I mean we're tearing down those proprietary on-premise servers that have basically run global finance for the last 40 years.

SPEAKER_00 0:44

Trevor Burrus Right, moving into cloud-native multi-tenant environments.

SPEAKER_01 0:47

Aaron Powell Yeah. And it's not just like a software patch, it's a complete rewiring of the central nervous system of global finance.

SPEAKER_00 0:54

Aaron Powell Which is our mission today for you, the listener. We're going to decode this invisible plumbing. What does this mean for the velocity of money, institutional risk, and you know, the future of AI in capital markets?

SPEAKER_01 1:05

Aaron Powell Because it really changes everything.

SPEAKER_00 1:07

Aaron Ross Powell Exactly. But why should a casual learner even care about back-end financial systems?

SPEAKER_01 1:12

Aaron Powell Well, you really should care because understanding these systems reveals the future of market fairness, AI, and global connectivity. I mean, the stakes are totally structural here.

SPEAKER_00 1:22

Trevor Burrus, Jr.: Right. It's the foundation.

SPEAKER_01 1:23

Exactly. When an entity like NASDAQ begins migrating core matching engines and risk management systems to the cloud, it just signals a complete reevaluation of how market infrastructure even operates. Bannert Thurner makes a really compelling case that this modernization is driven by an absolute necessity for elasticity and resilience.

SPEAKER_00 1:44

Because for decades, competitive advantage in finance was really defined by proximity, right?

SPEAKER_01 1:50

Yes, like who had the fastest proprietary servers physically co-located right next to the exchange. But now the architecture itself is changing those rules of engagement.

SPEAKER_00 1:59

Okay, let's unpack this. Because before we talk about the specific problems banks are facing today, we really need to understand this huge technological shift. The transition to cloud environments, specifically NASDAQ's partnership with AWS, fundamentally changes the barrier to entry. Aaron Powell Right.

SPEAKER_01 2:16

If we connect this to the bigger picture, this is basically the democratization of Wall Street. Smaller and mid-sized banks used to just have to compromise on their tech stack.

SPEAKER_00 2:24

Right, because they obviously couldn't afford the massive fixed costs of building a tier one risk management system from scratch.

SPEAKER_01 2:31

Aaron Ross Powell Exactly. But by shifting to a software as a service model hosted on AWS, those massive fixed capital expenditures just become variable operational costs. Yeah. So a mid-sized bank in an emerging market can spin up the exact same AI-driven risk analytics that a Wall Street primary dealer uses.

SPEAKER_00 2:50

So they're using fully managed solutions like NASDAQ Calypso. It's almost like a local indie filmmaker suddenly getting to rent the exact same CGI supercomputers used by the biggest Hollywood studios.

SPEAKER_01 3:01

That is a perfect analogy. You are effectively commoditizing the infrastructure and flattening the whole global playing field.

SPEAKER_00 3:09

But wait, if everyone has the same tools now, why isn't every bank instantly succeeding? Like we've watched major institutions pour billions into digital transformation over the last decade, and it often just results in massive write downs.

SPEAKER_01 3:23

Yeah, because just lifting and shifting legacy code into an AWS instance doesn't actually fix a fundamentally broken data architecture.

SPEAKER_00 3:33

Aaron Powell Right. So if the world-class tools are available, the lag points to a deeper friction. The playing field is theoretically leveled, but institutional agility is still wildly uneven.

SPEAKER_01 3:43

Aaron Powell And that friction is rooted in what the interview describes as the hidden tax of complexity. Most major financial institutions are basically operating on a Frankenstein architecture.

SPEAKER_00 3:52

Aaron Powell A Frankenstein architecture. I love that term.

SPEAKER_01 3:54

Aaron Powell It's so true though. Over the last 30 years, as new asset classes emerged and electronic trading evolved, banks rarely ripped out their foundational systems to rebuild them cleanly.

SPEAKER_00 4:03

Aaron Ross Powell I guess the risk of disrupting live trading was just way too high.

SPEAKER_01 4:06

Exactly. So they just built horizontal layers instead, bolting on a new derivatives module here, adding a new foreign exchange matching engine there.

SPEAKER_00 4:14

Trevor Burrus And relying on middleware to just tape it all together. So you end up with a scenario where the equities desk is running on a completely different database than the fixed income desk, exactly.

SPEAKER_01 4:25

And neither system natively communicates with the central ledger. The operational reality of that fragmentation is what Bannert Thurmer calls a reconciliation nightmare.

SPEAKER_00 4:36

Aaron Powell Wait, really? These are the wealthiest institutions on Earth. Why do they have reconciliation nightmares? Can't they just buy a new software package and start fresh?

SPEAKER_01 4:44

Well, it's because the data is so hopelessly siloed across disparate mainframe architectures. You literally cannot achieve a real-time, holistic view of institutional risk because the systems rely on asynchronous batch processing.

SPEAKER_00 4:58

Ah, so they aren't talking to each other in real time.

SPEAKER_01 5:00

Aaron Powell Right. Say market volatility spikes. System A calculates one margin requirement for your equities exposure, while system B calculates a completely different requirement for your currency hedges.

SPEAKER_00 5:11

And because they don't share a data schema.

SPEAKER_01 5:13

Exactly. Human analysts or fragile API patches have to manually reconcile those positions at the end of the trading day just to ensure the firm isn't over-leveraged.

SPEAKER_00 5:23

Oh wow. And in a market environment, moving to T plus one settlement where trades settle essentially, the next business day relying on end-of-day batch processing is a massive systemic vulnerability.

SPEAKER_01 5:34

Aaron Powell You are literally flying blind during market hours. And whenever a regulatory body introduces a new reporting requirement, a bank with a legacy tech stack can't just push a universal update.

SPEAKER_00 5:46

They have to individually recode and validate that change across dozens of fragile interdependent systems.

SPEAKER_01 5:53

Exactly. And every single manual update introduces a new vector for operational failure. It traps institutional capital and maintenance rather than innovation.

SPEAKER_00 6:02

Which means they can't deploy real-time pricing algorithms or advanced AI.

SPEAKER_01 6:06

No, because their foundational data layer is completely fractured.

SPEAKER_00 6:09

Okay, here's where it gets really interesting. Because the traditional response from a bank facing this level of technical debt would be to just issue a massive request for proposal to Silicon Valley, right?

SPEAKER_01 6:20

Oh, absolutely.

SPEAKER_00 6:21

They'd hire a massive software vendor for a multi-year rip and replace project.

SPEAKER_01 6:26

Aaron Powell Right. But Bannert Thurner explicitly outlines why that standard sauce vendor model is failing in capital markets. Treating mission-critical financial architecture like a generic software procurement totally ignores the regulatory reality.

SPEAKER_00 6:40

So what's the alternative then?

SPEAKER_01 6:43

What's fascinating here is her emphasis on the concept of skin in the game. Regulated banks are realizing they cannot outsource core infrastructure to a tech vendor who has never faced a federal audit.

SPEAKER_00 6:55

Or who doesn't even understand the microstructure of market liquidity. Trust is the ultimate differentiator.

SPEAKER_01 7:00

Exactly. Providing enterprise software for a retail company is fundamentally different from providing the execution and risk layer for a systemically important financial institution.

SPEAKER_00 7:10

It's the difference between taking driving lessons from someone who only read the manual versus someone who actually drives an 18-wheeler through rush hour traffic every day. You want the partner who drives the truck.

SPEAKER_01 7:20

I love that. And NASDAQ drives the truck. They process billions of transactions daily and operate matching engines under continuous global regulatory scrutiny.

SPEAKER_00 7:30

They have the institutional scar tissue to prove they can scale cloud infrastructure safely.

SPEAKER_01 7:35

Right. And that operational lineage changes the dynamic from a vendor-client transaction to a strategic partnership. When Nasdaq offers managed services, they provide an operational framework that has already been stress tested against real global financial regulation.

SPEAKER_00 7:51

Aaron Powell Let's look at the mechanics of those managed services actually, because they directly attack that Frankenstein problem we were talking about.

SPEAKER_01 7:57

Yeah, they do.

SPEAKER_00 7:58

The interview specifically highlights NASDAQ Calypso for managing risk, margin, and collateral, and then Axiom SL for regulatory reporting.

SPEAKER_01 8:06

Right. So by migrating to these unified platforms, a bank effectively outsources the maintenance of all that underlying logic. Calypso centralizes your cross-asset transaction data into a single normalized environment.

SPEAKER_00 8:20

So it calculates margin requirements across all those historically siloed asset classes in real time.

SPEAKER_01 8:25

Exactly. And AxiomSLs function similarly on the compliance side. Regulatory reporting is so resource intensive, it requires pulling vast amounts of data and formatting it to meet the highly specific taxonomies of different international regulators.

SPEAKER_00 8:39

So Axiom SL acts as an automated translation layer between the bank's normalized data and the regulator's API.

SPEAKER_01 8:46

Right. The strategic pivot here is about stripping away the operational dead weight. A bank's core competency is pricing risk and serving clients.

SPEAKER_00 8:55

Yeah. Building the API logic to report data to the European Securities and Markets Authority is not a competitive advantage. It's just a utility.

SPEAKER_01 9:02

Exactly. By handing those utility functions over to a trusted partner, the bank frees up massive amounts of engineering bandwidth to focus on actual alpha generation.

SPEAKER_00 9:11

Aaron Powell And that consolidation force is a crucial outcome, right? It creates a unified, pristine data layer, which Kanner Thurner emphasizes is the absolute prerequisite for deploying artificial intelligence.

SPEAKER_01 9:21

Oh, without a doubt, this is the inflection point. You simply cannot train an effective machine learning model on fragmented asynchronous batch data. Aaron Powell Right.

SPEAKER_00 9:29

The model will just hallucinate or reinforce the errors of the legacy systems.

SPEAKER_01 9:33

Spot on. But once you have that unified architecture in the cloud, the convergence of elastic compute power and pristine data unlocks entirely new operational paradigms. AI models can detect highly sophisticated fraud anomalies in milliseconds.

SPEAKER_00 9:49

Just by analyzing cross-asset trading patterns, a human would never spot.

SPEAKER_01 9:53

Exactly. But the source points out that the real magic is the community aspect of this multi-tenant architecture.

SPEAKER_00 10:00

Oh, right. Because in a legacy environment, a bank's infrastructure was a walled garden. Their data, their threat intelligence, it was all totally isolated.

SPEAKER_01 10:08

Yeah. But on a shared cloud-based platform like NASDAQs, you introduce the possibility of collective intelligence. The network effects are profound.

SPEAKER_00 10:16

So if a novel form of synthetic identity fraud or a new algorithmic spoofing tactic targets just one single institution?

SPEAKER_01 10:24

The underlying AI models can analyze that threat vector, adapt, and instantaneously deploy the countermeasure across the entire ecosystem.

SPEAKER_00 10:32

Wow. The whole community benefits from the intelligence generated by a single node. It's essentially a financial hive mind.

SPEAKER_01 10:39

It really is.

SPEAKER_00 10:40

But wait, what does this mean for institutional strategy? If everyone is sharing threat telemetry and pooling best practices, doesn't that neutralize a bank's competitive advantage? Like who actually wins here?

SPEAKER_01 10:53

Well, it forces a redefinition of what constitutes a competitive advantage in the first place. Operational efficiency and basic risk mitigation are no longer areas where banks want to compete.

SPEAKER_00 11:04

Ah, they're just baseline utilities now.

SPEAKER_01 11:06

Right. Having a better fraud detection filter than your competitor doesn't win you institutional mandates. It just keeps you out of the news.

SPEAKER_00 11:12

Aaron Powell So you commoditize the defense to focus all your capital on the offense.

SPEAKER_01 11:16

Exactly. The shared intelligence raises the foundational floor for everyone, allowing institutions to compete on higher order value, like proprietary trading algorithms and complex credit modeling.

SPEAKER_00 11:28

But all this talk of self-learning algorithms in global capital markets triggers immediate systemic alarms for me. We've seen firsthand what happens when automated trading algorithms interact unpredictably.

SPEAKER_01 11:38

Oh, yeah, like the flash crashes where billions of dollars of market capitalization just vanish in minutes.

SPEAKER_00 11:44

Aaron Powell Right. So regulators do not look at cloud-based AI ecosystems and just give them a rubber stamp.

SPEAKER_01 11:50

Aaron Powell No, the regulatory scrutiny is appropriately intense. Modernization is as much about governance as it is about technology. You cannot deploy predictive AI models without mathematical guardrails.

SPEAKER_00 12:03

Aaron Powell And Bennett Thurner specifically notes that NASDAQ aligns its AI governance frameworks with NIST standards, right? The National Institute of Standards and Technology.

SPEAKER_01 12:14

Yes. NIST provides a highly rigorous framework for managing the lifecycle risks of artificial intelligence. Implementing a framework like that in finance requires solving the black box problem.

SPEAKER_00 12:24

Aaron Powell Explainability.

SPEAKER_01 12:25

Exactly. Explainability is the foundational pillar of responsible AI. If a neural network decides to flag a transaction or execute a liquidation, you can't just tell a regulator, well, the algorithm optimized for it.

SPEAKER_00 12:37

Right. The mathematical weights and the data provenance have to be totally transparent and understandable to human risk officers.

SPEAKER_01 12:43

Yes. And the governance framework mandates continuous human oversight for systemic decisions. The AI serves to augment human intelligence processing the data lakes, surfacing the anomalies, but the final authorization remains with the human operator.

SPEAKER_00 12:58

So for you listening, basically, you need a human holding the leash, ready to explain why the AI made a choice.

SPEAKER_01 13:05

Exactly. And the framework also requires continuous monitoring for model drift.

SPEAKER_00 13:10

Right, because an AI trained in a low interest rate environment might make catastrophic miscalculations in a volatile high inflation environment.

SPEAKER_01 13:18

Absolutely. You have to constantly validate the models against real-time conditions, which raises an important point regarding how innovators communicate with regulatory bodies. The source advocates for proactive, radically transparent dialogue.

SPEAKER_00 13:31

You don't just build a multi-tenant AI architecture in a sandbox and try to force it past the regulators.

SPEAKER_01 13:36

No, you bring them into the architectural process. You demonstrate empirically how these technologies actually enhance market integrity.

SPEAKER_00 13:43

Because the ultimate currency here isn't just data, it's trust. Trust that the unified data architecture is resilient. Trust that the AI models are transparent. And trust that the managed service partner really understands the regulatory burden.

SPEAKER_01 13:58

Exactly. And the institutions that successfully navigate this transition won't just be adopting new technology. They will be operating on a fundamentally different, significantly faster foundation than peers who cling to the legacy models.

SPEAKER_00 14:12

We have covered a massive amount of architectural ground today. We started by looking at how the migration to the cloud is democratizing access to tier one market infrastructure.

SPEAKER_01 14:22

And we broke down the Frankenstein problem, how decades of siloed legacy systems created massive reconciliation nightmares.

SPEAKER_00 14:29

Right. We explored why solving this requires heavily regulated partners who have real skin in the game, utilizing managed services like Calypso and Axiom SL to normalize data.

SPEAKER_01 14:39

Yeah, and we saw how this unified data unlocks the power of AI, creating a multi-tenant network where collective intelligence raises the operational baseline for everyone.

SPEAKER_00 14:48

And finally, we examined the rigorous governance required to satisfy regulators, utilizing NIST frameworks to ensure algorithmic explainability and constant human oversight. It's incredible.

SPEAKER_01 15:00

Synthesizing the sheer scale of the infrastructure evolution, Banner Thurner outlines, there is a fascinating strategic paradox to consider, though. Oh. The interview makes it clear that the future of finance relies on community intelligence, shared platforms, and optimized cloud ecosystems. Right. But if all banks eventually shared the exact same highly optimized AI-driven cloud-based infrastructure, what becomes the true competitive advantage for a bank in the future? If the back end is identical, where does the real innovation happen next?

SPEAKER_00 15:32

Oh wow. That is the perfect question to leave hanging. When the foundational infrastructure of global finance becomes perfectly efficient and instantly accessible to everyone, the battleground shifts entirely.

SPEAKER_01 15:42

It's a whole new paradigm.

SPEAKER_00 15:44

Take a moment to think about what happens to the market when the playing field isn't just leveled but perfectly synchronized. Thank you for joining us today and bringing your curiosity to the deep dive.