EP1001: The outlook for treasury and capital markets

This podcast revisits an interview with Herve Carrere discussing how financial institutions must modernize outdated legacy systems to manage economic uncertainty and increasingly strict regulations. Teciem executive Herve Carrere, formerly of Finastra, explains that cloud-native platforms and modular architectures help banks improve flexibility, reduce operational risks, and respond faster to market changes. The interview also highlights the growing impact of Generative and Agentic AI, which can automate complex financial processes, deliver real-time insights, and improve decision-making in treasury and capital markets. Carrere emphasizes the importance of continuous software “evergreening” through automated testing and strategic partnerships to keep systems updated without interrupting operations. Additionally, the discussion points to increasing demand for Islamic finance solutions that combine innovation with cultural and regulatory requirements. Overall, the interview presents a future where customer-focused technology and autonomous systems drive efficiency, adaptability, and long-term competitiveness in banking.

Transcript

SPEAKER_01 0:00

Imagine a global newswire breaks, right? Like a central bank on the other side of the world just unexpectedly slashes its benchmark interest rate.

SPEAKER_00 0:08

Aaron Powell Which instantly throws the market into total chaos.

SPEAKER_01 0:11

Aaron Powell Exactly. But within milliseconds, I mean, before a human banker can even blink, let alone pour their morning coffee, an autonomous AI system has already read the news. Aaron Powell Yeah.

SPEAKER_00 0:21

And it's calculated the risk exposure across, you know, a trillion dollars of global assets.

SPEAKER_01 0:26

Right. And executed a massive defensive trade to protect its institution. And uh this isn't some pitch for a sci-fi movie. This is the actual new plumbing being built for your money right now.

SPEAKER_00 0:37

It's happening as we speak.

SPEAKER_01 0:38

So welcome to today's deep dive. We are jumping into a really fascinating, highly technical look at the invisible high-stakes scramble happening right beneath the surface of the global banking world.

SPEAKER_00 0:51

Yeah. And our roadmap for this one is an exclusive June 2025 interview from the IBSI FinTech Journal. It features Herv Carrera, who is the chief product officer of Treasury and Capital Markets at a massive financial technology company called Finastra.

SPEAKER_01 1:05

And he basically pulls back the curtain on the digital nervous system of global finance.

SPEAKER_00 1:10

Aaron Powell He really does. And you know, even if you don't spend your days treating complex derivatives, you really need to understand this shift.

SPEAKER_01 1:18

Because these are the colossal technological systems that safeguard the entire global economy, right? Like this infrastructure dictates how credit flows, how mortgages are funded, how your retirement accounts are protected.

SPEAKER_00 1:30

Exactly. When these systems lag, or you know, if they break, the economic shockwaves hit everyone. We are talking about institutions transitioning from this era of manual, rigid operations to an era of autonomous split-second decision making.

SPEAKER_01 1:47

The stakes could literally not be higher. Okay, so let's unpack this because Carrere paints a picture of an industry that is, frankly, practically buckling under its own weight.

SPEAKER_00 1:56

Oh, absolutely buckling.

SPEAKER_01 1:57

We have these massive financial institutions, some of which hold more wealth than the GDP of entire nations, and they are struggling just to keep the lights on. Are banks basically trying to navigate a modern high-speed digital highway while driving a horse-drawn carriage?

SPEAKER_00 2:11

That is uh honestly a perfect analogy. The interview describes this convergence of severe pressures. First, you have unprecedented macroeconomic volatility.

SPEAKER_01 2:21

Right. Like wild swings in inflation and rapid shifts in interest rates.

SPEAKER_00 2:24

Exactly. Then you layer on constant geopolitical turbulence, which creates these massive, unpredictable ripples across global supply chains and capital markets. Right. But the real friction point, like the thing actually breaking the internal systems, is the regulatory environment.

SPEAKER_01 2:40

Aaron Powell Yeah. The interview spends a lot of time on something called Basil the Feith, which uh sounds like a terrifying movie sequel, but what is it actually forcing banks to do?

SPEAKER_00 2:49

Aaron Powell Well, Basil the Feith is this massive international regulatory framework. And its core goal is basically to prevent another 2008 financial crisis.

SPEAKER_01 2:58

Aaron Powell Okay. That makes sense.

SPEAKER_00 2:59

Historically, banks used their own internal, sometimes highly optimistic mathematical models to figure out how much risk they were taking on and therefore how much capital they needed to hold in reserve.

SPEAKER_01 3:09

Aaron Powell So they were kind of grading their own homework.

SPEAKER_00 3:11

Aaron Powell Exactly. But Basil Faith takes away a lot of that freedom. It forces banks to use incredibly strict standardized models to calculate credit risk and market risk.

SPEAKER_01 3:21

Aaron Powell So they can't just estimate anymore. They have to prove their math using these standardized formulas. I mean, that implies they need a staggering amount of data on every single transaction.

SPEAKER_00 3:32

Aaron Powell Staggering is definitely the right word for it. They must calculate their liquidity risk, meaning, you know, do they have enough cash on hand to survive a sudden panic across massive pools of assets in near real time? And that's just one regulation. You also have AML or anti-money laundering laws, which require banks to monitor these vast networks of transactions to catch illicit funding.

SPEAKER_01 3:54

And the text also mentions new sustainability disclosures, right?

SPEAKER_00 3:57

Aaron Powell Yeah. Environmental, social, and governance or ESG reporting. Banks now have to track the actual environmental impact of the companies they finance.

SPEAKER_01 4:05

Aaron Powell Oh, wow. Think about what that requires. You aren't just looking at spreadsheets of dollar amounts anymore.

SPEAKER_00 4:10

Trevor Burrus, Jr. No, you are pulling in unstructured data about the carbon footprint of a supply chain halfway across the world.

SPEAKER_01 4:15

Aaron Powell If we look at how traditional banks are built, this sounds like a total nightmare because they've been running on legacy systems built decades ago.

SPEAKER_00 4:23

Right. Highly rigid systems.

SPEAKER_01 4:24

Think of those old systems like massive welded steel pipelines. They were built to push water from point A to point B.

SPEAKER_00 4:31

Just basic flow.

SPEAKER_01 4:33

Yeah. And now regulators are asking those same welded pipes to sort the water by temperature, filter out specific minerals, and measure the exact flow rate at a thousand different junctions simultaneously. The pipes can't bend, they just burst.

SPEAKER_00 4:47

That welded pipeline analogy perfectly captures the concept of siloed systems. You have different departments within a capital markets, firm-like trading, risk management, compliance operating on completely different software.

SPEAKER_01 4:59

Software that simply cannot talk to each other.

SPEAKER_00 5:01

Exactly. So when a trader executes a complex structured product, the validation of that trade gets delayed because the data has to be manually reconciled between systems.

SPEAKER_01 5:11

So they are flying blind at the exact moment they need absolute clarity.

SPEAKER_00 5:15

Precisely, which means they must modernize. But historically, overhauling a major bank's tech infrastructure is a famously terrifying prospect.

SPEAKER_01 5:25

Aaron Powell Wait, isn't migrating a bank's entire system famously a decade-long multimillion dollar nightmare?

SPEAKER_00 5:32

Oh, easily. It carries a massive risk of a catastrophic outage.

SPEAKER_01 5:36

Like trying to perform open heart surgery on a patient who is currently running a marathon. You can't just pause a global bank on a Friday, rip out its core foundation, and hope everything boots back up by Monday morning.

SPEAKER_00 5:46

Right, you just can't. But what's fascinating here is how Finastra is solving this exact problem. The underlying architecture of software has fundamentally changed to solve that deployment risk.

SPEAKER_01 5:57

Okay, how so?

SPEAKER_00 5:58

Well, the old way was called monolithic architecture. That massive welded steel pipe you mentioned, changing one line of code could break the entire bank.

SPEAKER_01 6:06

Right.

SPEAKER_00 6:06

Carrera explains that Finastra has moved entirely to a microservices-based architecture.

SPEAKER_01 6:11

Okay, walk us through the mechanics of that. How do microservices change the game?

SPEAKER_00 6:15

Instead of a giant web of entangled code, the software is broken down into hundreds of small independent containers or microservices.

SPEAKER_01 6:24

So it's less like a solid block of concrete and more like a set of building blocks.

SPEAKER_00 6:28

Exactly like building blocks. If you need a new tool to calculate Bathsolva compliance, you don't rebuild the foundation. You just snap that specific functional block into the existing framework.

SPEAKER_01 6:39

So you're basically swapping out the engine parts while the car is still driving down the highway.

SPEAKER_00 6:44

Through what they call an API-enabled ecosystem, yes. Application programming interfaces allow these separate blocks of code to talk to each other seamlessly.

SPEAKER_01 6:54

So banks can select and integrate only the specific functionalities they need, minimizing the blast radius if something goes wrong.

SPEAKER_00 7:01

Exactly.

SPEAKER_01 7:02

That sounds great in theory, but is anyone actually pulling this off in the real world without causing a massive disruption?

SPEAKER_00 7:08

Yeah, look at the LP Bank case study highlighted in the interview. They utilized Finastra's modular cloud native solution, which is called Condor. LP Bank completely modernized its treasury operations, meaning the way they manage their core capital, liquidity, and investments in just six months.

SPEAKER_01 7:25

Wait, really? Six months to overhaul a bank's treasury system is practically the blink of an eye.

SPEAKER_00 7:31

It is. They actually won Best Treasury Implementation and Best Risk Management Implementation at the IBS Intelligence Global FinTech Innovation Awards in 2024 for it.

SPEAKER_01 7:41

Oh wow. So it proved that modular modernization actually works.

SPEAKER_00 7:44

It did. And a key part of that success was moving to real-time position management. Under legacy systems, banks often relied on batch processing.

SPEAKER_01 7:53

Right, where they wait until the market closes at 5 p.m., run a massive calculation overnight, and find out the next morning what their actual risk exposure was.

SPEAKER_00 8:00

Aaron Powell, which is completely useless if the market crashes at noon.

SPEAKER_01 8:03

Seriously.

SPEAKER_00 8:04

But with Condor's microservices, every time a trade happens, the entire risk profile of the bank updates instantly.

SPEAKER_01 8:11

Now, the source also stresses that Finastra isn't doing this alone. There's huge emphasis on this being an ecosystem approach, right?

SPEAKER_00 8:18

Aaron Powell Yes. Modernization is totally a team sport now. The era of a single software vendor building every single tool a bank needs from scratch is over.

SPEAKER_01 8:27

Aaron Powell So Finastra provides the core platform, but they rely on partners.

SPEAKER_00 8:31

Exactly. They bring in companies like DXE Luxoft and right-click solutions to handle managed services.

SPEAKER_01 8:38

Let's clarify managed services really quick. That essentially means farming out the IT maintenance, right?

SPEAKER_00 8:42

Pretty much. Yeah.

SPEAKER_01 8:43

So the bank doesn't have to hire a private army of engineers just to keep the servers running. These partners manage the day-to-day operations.

SPEAKER_00 8:49

Aaron Powell, which allows the bank's leadership to focus on actual banking rather than software troubleshooting. And for highly specialized functions, they integrate with external platforms.

SPEAKER_01 9:00

Aaron Powell The interview mentions a partnership with Cloud Margin, right?

SPEAKER_00 9:03

Aaron Powell Yes. Cloud Margin is a SAW software as a service platform dedicated entirely to collateral management.

SPEAKER_01 9:10

Aaron Powell Okay. Explain the mechanism behind collateral management for us, because that's a term that gets thrown around a lot in high finance.

SPEAKER_00 9:17

Aaron Powell Sure. When banks trade complex derivatives, they require the opposing party to post collateral, usually cash or highly liquid bonds, to guarantee they can cover the trade if it goes bad. Okay. But as the market fluctuates minute by minute, the value of that trade changes, which means the amount of collateral required also constantly changes.

SPEAKER_01 9:38

So if you aren't tracking that dynamically, you might suddenly find yourself under-collateralized and exposed to a massive loss.

SPEAKER_00 9:44

Aaron Powell Which is exactly why you need a specialized API integrated tool like Cloud Margin, constantly crunching those specific numbers in the background. It's a whole ecosystem of specialized tools talking to each other instantly.

SPEAKER_01 9:57

Aaron Powell We've talked about how this modular approach solves the speed issue, allowing for quick upgrades. But this architectural shift also solves geography and culture.

SPEAKER_00 10:05

Oh, absolutely.

SPEAKER_01 10:06

There's this detail in the text about Vision Bank in Saudi Arabia. I found it so interesting that this exact same core technology can adapt to an entirely different regulatory framework.

SPEAKER_00 10:16

Yeah, Vision Bank is a purely digital institution, and importantly, it is fully Sharia compliant.

SPEAKER_01 10:22

Meaning it operates under Islamic finance principles, which prohibit the charging or paying of interest. That is a fundamental difference from standard Western banking.

SPEAKER_00 10:32

It's huge. If you look at traditional Western banking software, the calculation of interest rates is baked into the absolute deepest layers of the monolithic code. You can't just flip a switch and turn it off. Right. But because Vision Bank is running fully on Condor's microservices architecture, they literally just removed the standard interest calculation module and snapped in a specialized profit sharing module that complies with Sharia law.

SPEAKER_01 11:00

They didn't have to hire developers to build a custom banking platform from the ground up for the Middle East. They just swapped the Lego blocks.

SPEAKER_00 11:07

Exactly. It proves that high-tech scalability can perfectly marry with Islamic banking requirements.

SPEAKER_01 11:13

But, you know, technology evolves relentlessly. Even with a brilliant microservices set up today, how does a bank prevent this shiny new system from becoming just another obsolete piece of technical debt in five years?

SPEAKER_00 11:25

Well, Carrera introduces this concept called evergreening.

SPEAKER_01 11:28

Evergreening.

SPEAKER_00 11:28

Yeah. It's not just about software updates, it's about continuous life. An evergreen system is designed to remain constantly up to date, secure, and adaptable without the bank having to authorize massive overhaul projects every few years.

SPEAKER_01 11:42

But doesn't constantly updating a banking system reintroduce the exact risk we were trying to avoid, like every update is a chance to break the code.

SPEAKER_00 11:51

It would be if human developers were manually deploying the code and hoping for the best. But the mechanism that makes evergreening possible is automated testing.

SPEAKER_01 12:00

Delivered with partners, right?

SPEAKER_00 12:01

Yes. Through these continuous integration pipelines, every time a new micro update is proposed, the system automatically clones the banking environment, runs millions of simulated transactions against the new code in the background, flags any errors, and only pushes it live if it passes flawlessly.

SPEAKER_01 12:18

Oh wow. So the system tests its own upgrades before humans ever touch them.

SPEAKER_00 12:22

It basically ensures continuous, repeatable upgrades without deployment complexity or risk.

SPEAKER_01 12:27

Aaron Powell Okay, so we've established the foundation. You have this modern, modular, constantly updating, evergreened IT infrastructure. The plumbing is clean, fast, and secure. Right. Here's where it gets really interesting. Once you have that perfect nervous system, what do you do with it? You plug in an entirely new kind of brain.

SPEAKER_00 12:45

Aaron Powell You're talking about the massive shift in AI adoption. Aaron Powell Yeah.

SPEAKER_01 12:49

The numbers in Finastra's 2024 State of the Nation survey are wild. 61% of financial institutions reported recently adopting AI or improving their capabilities.

SPEAKER_00 12:59

That's double the number from 2022.

SPEAKER_01 13:01

Double. And 35% are specifically deploying generative AI.

SPEAKER_00 13:06

Aaron Powell We have to move past the public perception of generative AI as just a tool for writing polite emails or summarizing meeting notes. Right. Gen AI isn't just writing emails for bankers, it's running sophisticated stress tests, predicting liquidity gaps and spotting anomalies across thousands of transactions.

SPEAKER_01 13:22

Give us a concrete example of how a bank actually uses that, though.

SPEAKER_00 13:25

Okay, take the stress testing required by Basil the Fifth that we discussed earlier. Gen AI can ingest thousands of pages of new central bank policy announcements. The unstructured text cross-reference it with the bank's millions of structural data points regarding its current asset holdings and instantly predict potential liquidity gaps.

SPEAKER_01 13:45

So it's basically a really fast research assistant reading global newsfeeds and central bank policies to give human bankers advice.

SPEAKER_00 13:54

If we connect this to the bigger picture, it's actually much more than that. Generative AI is merely the current stepping stone.

SPEAKER_01 13:59

Okay.

SPEAKER_00 14:00

The interview highlights the final frontier of this modernization journey, which is agentic AI.

SPEAKER_01 14:05

Aaron Powell Agentic, meaning it has agency.

SPEAKER_00 14:07

Yes. While Gen AI provides insights and requires a human prompt, agentic AI is capable of autonomous task execution. Wait, really? Yeah. It independently perceives its environment, reasons through a complex problem, acts on its own conclusions, and then learns from the results of that action.

SPEAKER_01 14:23

Aaron Powell So it acts on its own. It doesn't wait for human approval.

SPEAKER_00 14:26

In a world of global volatility, it can make lightning fast decisions on trading activities and hedging strategies without waiting for a human.

SPEAKER_01 14:34

Let's go back to our opening scenario then. A geopolitical event happens. The news hits the wire. The agenic AI perceives the unstructured text of the news report. It uses its risk models to reason that this event will crash a specific currency. And instead of alerting a committee of human risk managers to schedule a meeting, it instantly executes massive hedging trades to protect the bank's balance sheet in fractions of a second.

SPEAKER_00 15:00

It transforms the bank into a highly proactive entity that navigates global uncertainty at a speed humans just cannot match.

SPEAKER_01 15:08

That is awe-inspiring and frankly a little daunting. But there is a brilliant, almost ironic human twist to how this technology is being developed.

SPEAKER_00 15:16

Oh, the customer centricity part.

SPEAKER_01 15:18

Yes. Despite all this talk of autonomous AI and cloud native architecture, Finastra actually figures out what to build next in a really old school way.

SPEAKER_00 15:26

They rely heavily on customer surveys and regional user groups to guide their product development.

SPEAKER_01 15:31

They literally get their clients in a room and ask, what is frustrating you today? They use human operational friction to dictate the direction of this hyperadvanced tech.

SPEAKER_00 15:40

It grounds the theoretical potential of agentic AI into the practical reality of what a treasury manager actually needs.

SPEAKER_01 15:48

It really does. But uh there is one final irony hidden in this document. If you look closely at the very bottom of the interview, there is a tiny footnote.

SPEAKER_00 15:56

Oh, right, the editor's note. It states that the interview took place just prior to the news of the sale of Finastra's Treasury and Capital Markets division.

SPEAKER_01 16:05

The very division Herv Carrer is leading sold right after he gave this expansive interview about creating stability in a turbulent market.

SPEAKER_00 16:13

It perfectly proves his underlying thesis. In the volatile world of finance, massive shifts happen behind the scenes in the blink of an eye.

SPEAKER_01 16:22

So what does this all mean for you? It means that the very foundation of global money is moving from rigid manual vault to agile AI-driven ecosystems.

SPEAKER_00 16:32

Which leaves us with the final question to ponder. If we are entering an era where agentic AI can independently perceive global news, reason through complex market conditions, and autonomously execute high-stakes trades in fractions of a second. Right. What happens to the global market on the day two? Vastly intelligent, highly autonomous banking AIs reach completely opposing conclusions about a massive global event. Who is truly in control of the market then?

SPEAKER_01 17:00

It really makes you wonder what's actually happening behind those glossy, upward trending charts, doesn't it? The invisible nervous system is waking up. Thanks for exploring it with us today. Keep diving deep.