Every platform that holds client funds faces a foundational infrastructure decision: should those funds sit in pooled accounts tracked by internal ledgers, or in named accounts structurally segregated at the custodian level? The choice affects compliance, operational risk, audit complexity, and ultimately the platform's relationship with regulators. This is not a feature comparison. It is an architecture decision.
Pooled accounts and named accounts represent fundamentally different approaches to custody, each with distinct trade-offs in cost, scalability, regulatory exposure, and resilience under stress. Understanding those trade-offs is increasingly urgent as regulators across the UK, EU, and US simultaneously tighten requirements for how client funds are held. The FCA's Supplementary Regime, PSD3's segregation-at-receipt mandate, and the GENIUS Act's reserve requirements all converge on the same principle: structural separation over ledger-based tracking.
How pooled accounts work
In a pooled account structure, also called an omnibus or For Benefit Of (FBO) account, a single bank account holds funds belonging to multiple end customers. The bank maintains one account with one aggregate balance. The platform maintains a separate internal ledger tracking which portion belongs to which customer. The operational advantage is efficiency: opening one bank account is faster and cheaper than opening thousands, and fund movements within the pool do not require bank processing.
The structural risk is the gap between the bank's view and the customer's expectation. The bank sees one account. The customer assumes they have an individual account. If the platform's internal ledger drifts from the bank's records, or if the platform becomes insolvent, there is no structural mechanism to identify which funds belong to which customer. Recovery depends on the quality of internal recordkeeping, which is exactly the component most likely to fail under stress.
The Yale Journal on Regulation's analysis of the Synapse collapse documented how a single unreconcilable ledger froze $265 million across 100,000 accounts. The FDIC's subsequent proposed rulemaking targets precisely this vulnerability, requiring banks to maintain daily reconciliation of beneficial ownership records in custodial accounts. The FCA found that payment firms that failed between 2018 and 2023 returned an average of 35 pence per pound to customers. Pooled account structures were the common factor.
How named accounts work
In a named account structure, each client receives a distinct account held in their name at the custodial institution. The bank sees individual accounts with individual balances. Ownership is determined by the account structure itself, not by an intermediary's ledger. When a regulator or insolvency practitioner needs to identify who owns which funds, the answer is visible in the account records. There is no dependency on a third-party reconciliation process. The custody architecture is self-documenting.
The operational trade-off, historically, has been speed and cost. Opening individual bank accounts for thousands of customers was slow and expensive. This is changing. API-based account provisioning now allows platforms to open named custody accounts programmatically, maintaining the onboarding speed of a pooled model while providing the structural protection of segregated custody. The speed objection that historically favoured pooled structures no longer holds.
Named accounts also simplify compliance with safeguarding requirements. Under the FCA's Supplementary Regime, effective May 2026, payment firms must perform daily reconciliation and maintain resolution packs. With named accounts, reconciliation is a verification that each account holds the correct balance. With pooled accounts, reconciliation requires matching thousands of ledger entries against a single aggregate bank balance, a process that scales poorly and fails under stress. The audit burden is fundamentally different.
The regulatory direction is clear
Regulators in every major jurisdiction are moving toward prescriptive requirements for client fund segregation. The FCA's PS25/12, effective May 2026, mandates daily reconciliation, resolution packs, named senior manager accountability, and annual audits. While it does not explicitly prohibit pooled accounts, the compliance burden it imposes on pooled structures creates a strong practical incentive toward segregated custody. E-money institutions were safeguarding approximately £26 billion by 2024, more than double the £11 billion held in 2021.
PSD3, which reached provisional political agreement in November 2025, requires funds to go directly into safeguarding accounts at receipt with no commingling. The European Commission's legislative package establishes segregation at receipt as a baseline requirement, not a best practice. The GENIUS Act in the US requires stablecoin issuers to segregate reserve assets and prohibits commingling, with Latham & Watkins noting that custodial services may only be performed by entities under federal or state banking supervision.
The MAS in Singapore and the DFSA in Dubai have both tightened custody requirements for licensed institutions, with emphasis on structural segregation of client assets. Platforms operating on pooled-only architectures face a closing compliance window across all of these jurisdictions simultaneously. The direction is not ambiguous. The timeline is not distant. The infrastructure decision is now.
Making the comparison
Pooled accounts offer lower upfront cost, faster initial deployment, and operational simplicity for internal fund movements. They carry higher regulatory risk, higher reconciliation cost at scale, and structural vulnerability in insolvency. They require increasingly sophisticated compliance processes to satisfy tightening safeguarding requirements, and the cost of those processes grows with the platform's client base and regulatory footprint.
Named accounts offer structural compliance, simpler reconciliation, clearer audit trails, and resilience under stress. They require API-based provisioning infrastructure for scalability and involve per-account costs at the custodian level. They satisfy current and anticipated regulatory requirements by design rather than by process. For platforms at early stage with limited regulatory exposure, pooled accounts may be a pragmatic starting point.
For platforms at scale, operating across regulated jurisdictions, holding significant client fund balances, and subject to safeguarding requirements, named accounts are increasingly not optional. The compliance cost of maintaining pooled structures across the UK, EU, and US simultaneously, with daily reconciliation, annual audits, resolution packs, and monthly regulatory returns, will exceed the infrastructure cost of named accounts for most platforms within the first year.
The infrastructure choice
The custody architecture a platform chooses today determines its regulatory posture, operational resilience, and audit complexity for years to come. Migrating from pooled to named accounts mid-growth is expensive and operationally disruptive. Building on named accounts from the outset, or transitioning proactively before regulatory deadlines arrive, avoids that disruption and positions the platform for compliance across jurisdictions from day one.
Lorum provides named custody accounts across 30+ markets, provisioned via API, with structural segregation by design. For fintech and PSP platforms, payroll and EOR platforms, and marketplace operators, the named account model provides cash management infrastructure that satisfies current requirements and anticipated regulatory expectations across the UK, EU, and US simultaneously. The question is not whether to transition. It is when.
