As settlement cycles compress, payments become always-on, and AI drives unprecedented data movement, financial institutions can no longer afford infrastructure that was built by accumulation rather than engineered by design.
For decades, financial market infrastructure has evolved in one direction: outward. Connecting trading venues. Adding new data centres. Integrating cloud platforms. Layering third-party providers to meet diverse operating and business requirements.
The outcome is an infrastructure environment built by accumulation, not design. A mélange of carriers, cloud providers, exchanges, co-location facilities and service platforms, knitted together over years of incremental growth and vendor negotiation.
The result? Infrastructural fragmentation that functions well enough under normal operating conditions. The problems arise when conditions are not normal.
“”The real benchmark for modern infrastructure isn’t how it performs under normal operating conditions- it’s how predictably it performs when market volatility, transaction volumes, AI-driven workloads, and operational demands all peak at the same time. That’s where resilient network architecture becomes a business differentiator rather than simply an IT capability.” Pravin Vijay, Principal Solutions Engineer, Financial Markets
The financial institutions and firms most exposed to this fragmentation are those that continue to add layers of complexity without pausing to consider whether the architecture remains fit for purpose for the new and growing demands being placed upon it.
For the most part, the complexity of assembled financial market infrastructure has to date been manageable because the operational demands placed upon it were bounded. Longer settlement cycles. Batch-oriented payments processing. No AI-led workloads. Less prescriptive “demonstrable resilience” obligations.
Today, always-on markets, compressed settlement cycles, accelerating AI adoption and a host of other market drivers are intensifying infrastructural strain and constraints, at precisely the time that resilience and accountability obligations become measurably more onerous.
Regulatory environment changing the conversation
Infrastructure complexity always carries operational risk. However, new regulatory frameworks are more prescriptive about how that risk must be identified, measured and managed, particularly regarding third party providers of digital technology, hardware, software, network systems and data used to process, store and transmit information.
The Digital Operational Resilience Act (DORA) introduces mandatory requirements for ICT risk management, incident reporting within defined windows, and structural resilience testing across interconnected environments. What has been managed traditionally on a “best practice, best efforts” basis is now a legal obligation with board-level accountability.
The UK Critical Third Parties regime extends this logic further. Where a significant portion of financial sector activity depends on shared providers and shared infrastructure, systemic dependency becomes a supervisory concern, as well as an operational one.
Both require that institutions are able to demonstrate that their infrastructure supply chain is visible, governed, and resilient under stress. This is not a procurement decision. It requires a strategic reorientation of the relationship between network environments and operational outcomes.
It means working with an infrastructure partner that understands the operational realities of post-trade settlement, continuous payment processing, AI-driven data movement, and DORA-mandated resilience testing.
Institutions and firms that make this architectural shift earliest will satisfy regulatory requirements more efficiently and operate with an elevated level of infrastructure confidence that becomes a genuine competitive advantage.
Structural shifts accentuating infrastructure fragmentation
Regulatory pressures are arguably sufficient to support a reassessment of accumulated infrastructure models. However, three concurrent market structure changes mean that reassessment is becoming more urgent.
Settlement cycle compression
The EU transition to T+1 settlement, targeted for 11 October 2027, will significantly compress post-trade processing windows. T+1 demands that every component in the post-trade infrastructure stack performss predictably and consistently, in every cycle, without exception.
Infrastructure assembled across multiple providers, with unclear failover governance, cannot reliably provide that assurance.
| 11 Oct 2027 | EU T+1 Settlement Deadline Compressed post-trade windows leave zero tolerance for infrastructure uncertainty in the settlement chain |
Scale of always-on payment infrastructure
The ECB reported a five-fold increase in TARGET Instant Payment Settlement (TIPS) volumes during 2024. No longer batch-oriented, payments systems must operate continuously and at scale. The infrastructure supporting it must do the same, without scheduled maintenance windows.
An environment in which routing decisions pass through multiple carrier layers introduces concentration risks and governance gaps that may be invisible in normal conditions, yet catastrophic when they fail under load.
| 5x volume growth | TARGET Instant Payment Settlement ECB-reported increase in TARGET Instant Payment Settlement volumes in 2024, reflecting the relentless demand placed on always-on payment infrastructure |
AI as an infrastructure stress test
The financial sector’s investment in AI-driven analytics, real-time risk modelling and algorithmic execution is not a software challenge. It is a physical infrastructure problem. Ciena research forecasts at least a six-fold increase in data centre interconnect (DCI) bandwidth demand in the next five years, with 43% of new data centre capacity projected to support AI workloads.
The traffic patterns generated by distributed AI inference bear no relation to the stable, predictable flows that traditional financial infrastructure was designed to carry. Bursty demand, east-west compute traffic between co-located facilities, and variable inference loads create strain profiles that expose routing vulnerabilities in accumulated infrastructure that have never previously been visible.
| 6x bandwidth growth | Data Centre Interconnect Ciena forecasts 6x data centre interconnect bandwidth demand growth over five years, driven by AI workloads that generate bursty, unpredictable traffic profiles |
Together, these three shifts define a new operational baseline. Infrastructure that was adequate for a world of longer settlement cycles, batch payments and predictable data flows is now being asked to support a world that is structurally different in almost every dimension.
Rising costs of accumulated infrastructure
Accumulated infrastructure costs rarely show up as a single line item on a balance sheet. Costs may accumulate, and compound, across operations, business and governance stacks without clear visibility.
When a performance issue occurs in an infrastructure spanning multiple carriers, cloud providers and co-location facilities, troubleshooting specific components means navigating myriad separately managed environments, each with its own operational data, incident processes and contractual obligations. Resolution timescales extend. Accountability diffuses. Governance visibility degrades.
DORA does not ask if institutions have a resilience policy. It asks whether and how institutions can demonstrate resilience across their ICT supply chain. Something that becomes progressively harder as infrastructure complexities and dependencies grow, and holistic visibility diminishes.
ENISA recorded 188 major telecommunications security incidents in 2024, a year-on-year increase of 20.5%. Each of those incidents represents a point at which accumulated infrastructure dependencies, rather than engineered architecture, determined the operational outcome for firms sitting downstream.
The Financial Stability Board explicitly identifies management of third-party and cloud concentration risk as a priority governance challenge for financial institutions. Infrastructure.
| +20.5%x increase in security incidents | Telecom Security ENISA: year-on-year increase in major telecom security incidents in 2024, highlights growing operational fragility of shared infrastructure environments |
Distinction between assembled and engineered infrastructure
Assembled infrastructure reflects a history of reasonable decisions made in sequence. It connects components that exist at each point in time, using available providers and routing models.
Engineered infrastructure starts from an entirely different premise. First, defining specific operational outcomes and working backwards to design the physical and optical architecture needed to deliver them. Route diversity, failover governance, latency consistency, service assurance and regulatory visibility are intrinsic to this intentional design, not features bolted on after the fact.
“Infrastructure built by accumulation creates complexity that functions until it does not. Infrastructure engineered by design creates operational confidence that can be validated before a fault occurs.” Pravin Vijay, Principal Solutions Engineer, Financial Markets
Zayo Europe’s approach is grounded in this distinction. With 3.3 million km of fibre across sixteen European countries, on-net connectivity to over 600 data centres, Global Tier 1 IP status and ownership and operation of eight subsea cable systems connecting the UK to mainland Europe, the physical layer is owned, controlled and engineered from duct to IP.
Infrastructure ownership is not, in itself, the solution. What it enables is a combination of physical control and managed operational outcomes, with real-time service visibility, proactive incident management and proactive government of operational behaviour rather than reacting once an issue arises.
Predictability under pressure, the new performance standard
Financial infrastructure discussions have centered historically on raw latency benchmarks. In latency-critical trading environments, speed matters, and always will.
However, speed measured under ideal conditions in controlled circumstances is not the right metric to determine how a financial institution will perform in a period of acute market volatility, or sudden spike in volumes, or unexpected failure within a shared infrastructure component.
A more compelling metric is path predictability under pressure: ensuring that infrastructure continues to behave in the same way, following the same routes, with the same performance characteristics, in all circumstances and conditions.
A new standard for financial market infrastructure
Financial institutions have spent years optimising applications, trading platforms, and operating models to support faster and smarter markets. The next competitive advantage may, however, lie deeper in the “tech stack”. As markets become increasingly real-time, AI-driven and interconnected, infrastructure itself becomes a strategic differentiator.
This means moving from reactive problem resolution to real-time visibility across the infrastructure stack to surface issues before they become incidents. It means route governance that evidences built-in structural resilience rather than asserting it as a general capability.
The scope (and scale) of this type of approach moves the conversation’ beyond connectivity and bandwidth to treating the infrastructure layer as an essential component of total cost of operations, and active participant in resilience governance.
Firms that continue to assemble infrastructure incrementally will inevitably carry forward the cost and complexity of those decisions. Those that choose to architect infrastructure around predictability, visibility and operational confidence will be much better equipped to meet resilience obligations today, and ready to tackle whatever financial markets evolution requires of it in the future.
Zayo Europe owns and operates 3.3 million km of fibre across sixteen European countries, with on-net connectivity to 600+ data centres and Global Tier 1 IP status. Trusted by 25 of the top 30 global banks, 10 of the world’s leading trading firms and financial firms operating across 400+ global markets.
For further information: https://zayoeurope.com/industries/finance/