The Infrastructure Concentration Trap: Finance Is Safer on Paper—and More Dependent in Practice
Zeeshan · 2026-10-01
Shared financial infrastructure can improve resilience inside firms while concentrating failure risk at common nodes.
By Zeeshan | YouYaa Intelligence | 1 October 2026
The safer the plumbing looks, the more people may depend on the same pipes.
The controversial thesis
Modern finance is moving risk into shared infrastructure.
Central clearing can reduce bilateral counterparty exposure. Payment systems can make money move quickly. Securities settlement can reduce operational friction. Common standards can make markets easier to connect.
But every improvement creates a second question:
How many firms now depend on the same node, provider, settlement process, or recovery plan?
The Bank of England’s 2025–26 Financial Market Infrastructure Annual Report makes the issue clear. The Bank supervises financial market infrastructures (FMIs) because they support financial stability. It also focuses on cyber risk, third-party disruption, outsourcing, payment systems, central counterparties, settlement, and future tokenised infrastructure.[1]
The controversial conclusion is not that FMIs are unsafe. It is that a system can become safer at each firm while becoming more consequential at the shared infrastructure layer.
The £400 billion dependency
The Bank says the £400 billion figure for recognised payment systems represents average daily figures for CHAPS, Bacs, Faster Payments, Link, Mastercard Europe, and Visa Europe. CHAPS is not technically a recognised payment-system operator, but the Bank supervises it to an equivalent standard.[1]
That number is not a claim that every payment would stop during an outage. It is a scale indicator. When payment infrastructure carries hundreds of billions of pounds of average daily activity, operational resilience becomes a financial-stability issue, not just an IT issue.
| Signal | Verified evidence | Why it matters |
|---|---|---|
| Average daily payment-system figure | £400bn | Infrastructure disruption can affect many firms at once |
| Reported period | 26 June 2025–28 February 2026 | The evidence comes from a defined supervisory period |
| Cyber stress dimension | High severity | Recovery must work beyond normal incidents |
| Third-party stress dimension | Extreme but plausible prolonged outage | Outsourcing can create common failure points |
| Settlement transition | UK move toward T+1 | Faster settlement needs new controls and procedures |
| Future settlement scope | Tokenised and non-tokenised models | New rails may add choice and new dependencies |
The number is large enough to change the board conversation. Resilience is not merely about whether one bank’s server remains online. It is about whether the wider financial system can continue to pay, clear, settle, and reconcile when a shared service is impaired.
Why centralisation looks attractive
Financial market infrastructure exists for good reasons.
A central counterparty can stand between buyers and sellers. A settlement system can standardise delivery and payment. A payment rail can reduce friction. A shared service provider can lower costs and improve technical capability.
These benefits are real. The Bank’s report describes supervision designed to support financial stability and resilient market structures.[1]
| Benefit of shared infrastructure | Potential gain |
|---|---|
| Netting | Less duplicated exposure between participants |
| Standardisation | Fewer bespoke processes and errors |
| Central risk management | More consistent margin and default procedures |
| Faster settlement | Less time between trade and finality |
| Common connectivity | Easier access to markets and customers |
| Specialist providers | Better capability than each firm building alone |
The problem begins when the same benefit becomes a common dependency.
The concentration paradox
At firm level, outsourcing can reduce operational risk. A specialist provider may be more secure, better staffed, and more resilient than a small internal team.
At system level, many firms using the same provider may create concentration risk.
The Bank explicitly notes that recognised payment systems may outsource critical parts of their operations to service providers, and that the system’s ability to deliver its responsibilities may depend on those providers.[1]
That is the concentration paradox:
Outsourcing may reduce the probability of a failure inside one institution while increasing the number of institutions exposed to the same failure.
| Firm view | System view |
|---|---|
| The provider is stronger than our internal team | Many firms may share the same provider |
| The contract includes service levels | A contract cannot create capacity during a systemic event |
| The provider has a backup site | The backup may share the same cloud, network, or staff dependency |
| The service is efficient | Efficiency can increase common exposure |
| The incident is outside our perimeter | Customers still experience the failure through our brand |
This is why a vendor risk review cannot stop at the vendor’s own security certificate.
Extreme but plausible is the new normal
The Bank’s operational-resilience focus includes high-severity cyber events and third-party extreme-but-plausible scenarios. FMIs are expected to meet impact tolerances and provide a minimum level of service even in severe conditions.[1]
This is a higher standard than “we have a disaster-recovery plan.”
A plan is a document. An impact tolerance is a limit: how much disruption can occur before financial stability, customers, or market integrity are materially harmed?
| Question | Weak answer | Stronger answer |
|---|---|---|
| Can payments continue? | “We have a backup.” | “We can process the minimum service within the tested tolerance.” |
| Can trades settle? | “Our provider is resilient.” | “We tested provider failure and participant workarounds.” |
| Can the service recover? | “The contract sets an RTO.” | “The RTO works under shared-provider stress.” |
| Can data be reconciled? | “The database is replicated.” | “We tested integrity, sequence, and reconciliation after recovery.” |
| Can customers be informed? | “Communications are prepared.” | “We can explain status, balances, and exceptions in real time.” |
The test is not whether the system works on an ordinary day. It is whether the system can deliver its minimum purpose on a very bad day.
T+1: less settlement time, more operational pressure
The United Kingdom is moving toward T+1 settlement. The Bank says it will continue engaging FMIs to ensure that their systems and procedures support a smooth and orderly transition.[1]
Faster settlement can reduce counterparty exposure and free capital sooner. It can also compress the time available for trade affirmation, allocation, funding, exception management, and reconciliation.
| T+1 benefit | T+1 pressure |
|---|---|
| Shorter open exposure | Less time to fix trade breaks |
| Faster finality | Greater need for accurate data before cutoff |
| Potentially lower counterparty risk | More intraday operational pressure |
| Less trapped capital | Greater dependence on connectivity and automation |
| Clearer settlement cycle | More severe consequences from late exceptions |
The controversial question is whether the market is becoming faster faster than it is becoming recoverable.
Tokenised settlement does not remove dependency
The Bank is also considering future settlement models and consistent treatment across tokenised and non-tokenised settlement.[1]
Tokenisation can improve programmability, visibility, and settlement speed. It may also add new code, wallets, custody arrangements, or technology providers to the dependency map.
A new rail is not automatically a safer rail. Its resilience depends on governance, access, recovery, legal finality, cyber controls, and the ability to operate when part of the network is unavailable.
| New settlement feature | New dependency question |
|---|---|
| Programmable settlement | Who can change the rules or code? |
| Shared ledger | Who controls access and recovery? |
| Faster finality | How are errors corrected? |
| Digital custody | Who can freeze, move, or restore assets? |
| API connectivity | What happens when the API is unavailable? |
| Automated reconciliation | Can exceptions be reviewed by humans? |
Technology changes the shape of risk. It does not abolish it.
What this means for fintech operators
Fintechs often buy infrastructure to move quickly. That is rational. But the board should know whether the product is dependent on one payment rail, one cloud region, one identity provider, one custodian, one clearing member, or one data-reconciliation service.
A fintech’s own uptime may look excellent while its customer experience depends on an external system it cannot control.
The right question is not only “Is our vendor resilient?” It is also:
How many of our critical services fail together if this vendor, connection, or settlement node fails?
What this means for CFOs
CFOs should treat payment and settlement concentration as a balance-sheet and liquidity issue.
A payment outage can delay payroll, supplier payments, tax transfers, customer refunds, and treasury movements. A settlement outage can leave positions open, collateral trapped, and reconciliations incomplete.
| CFO question | Why it matters |
|---|---|
| Which payment rails do we rely on? | A “diversified” bank account may use the same rail |
| Which provider handles reconciliation? | A payment can move while the ledger remains uncertain |
| What is the true fallback? | A second vendor may share the same infrastructure |
| How much liquidity is trapped during an outage? | Recovery time can become a funding problem |
| Which services are outsourced? | Contractual responsibility is not operational independence |
| What is the impact tolerance? | The board needs a measurable limit, not a general promise |
| Can customers be paid manually? | Manual workarounds may not scale during a systemic event |
What this means for HNWIs and family offices
HNWIs and family offices may have diversified assets but concentrated infrastructure.
Several custodians may connect to the same settlement system. Multiple investment platforms may rely on the same prime broker, cloud provider, market-data vendor, or payment processor. Diversification of accounts does not guarantee diversification of rails.
The practical question is not only where the assets sit. It is how those assets move, settle, reconcile, and become usable during disruption.
A better resilience map
Every critical financial flow should be mapped from instruction to final settlement.
| Step | Dependency to identify |
|---|---|
| Instruction | Bank, platform, API, identity, approval |
| Execution | Venue, broker, liquidity, risk engine |
| Clearing | CCP, clearing member, collateral manager |
| Settlement | CSD, payment rail, custodian, wallet or ledger |
| Reconciliation | Data provider, ledger, operations team |
| Recovery | Backup system, provider, staff, communications |
The map should include common dependencies. If three vendors use the same cloud region, the map should show one shared node, not three independent options.
Conclusion
The Bank of England’s 2025–26 report shows why financial infrastructure now belongs in board-level risk discussions. The supervised ecosystem supports hundreds of billions of pounds in average daily payment activity, must prepare for severe cyber and third-party disruption, and is adapting to T+1 and possible tokenised settlement.[1]
The controversial conclusion is:
Finance may be becoming safer inside each institution while becoming more dependent on a smaller number of shared infrastructure nodes.
That is not an argument against centralisation, outsourcing, faster settlement, or innovation. It is an argument for measuring concentration honestly.
A resilient system is not one with the most technology. It is one that can continue its minimum critical service when the shared technology fails.
FAQ
What is financial-market infrastructure?
It is the systems and institutions that help finance clear, settle, pay, reconcile, and manage risk. Examples include central counterparties, settlement systems, payment systems, and specified service providers.
Why does FMI concentration matter?
A shared provider can improve efficiency and resilience for each firm while creating a common failure point for many firms.
What is the £400 billion figure?
The Bank of England says it represents average daily figures for CHAPS, Bacs, Faster Payments, Link, Mastercard Europe, and Visa Europe in the report’s payment-system context.[1]
Does outsourcing make finance safer?
It can improve specialist capability, but it can also create concentration. The answer depends on shared dependencies, recovery capacity, access, and tested workarounds.
What are extreme-but-plausible scenarios?
They are severe but credible events, including high-severity cyber incidents and prolonged third-party outages. FMIs are expected to meet their impact tolerances under such scenarios.[1]
How does T+1 change risk?
It shortens the settlement cycle. That can reduce open exposure but leaves less time for allocation, funding, exception management, and reconciliation.
Does tokenised settlement eliminate operational risk?
No. It may improve speed and programmability, but it creates questions about code, access, custody, recovery, legal finality, and service-provider dependence.
What should CFOs test first?
Map critical payment and settlement flows, identify common providers, measure liquidity trapped during disruption, test fallback routes, and define an impact tolerance.
Is this investment advice?
No. This is general analysis of financial-market infrastructure and operational resilience. Obtain appropriate advice for specific circumstances.
References
