At a glance
The 75% problem is the structural reason most US regional and community banks cannot serve fintech treasury needs. The loan-to-deposit ratio for these banks runs above 75%, meaning three out of every four dollars of customer deposits are deployed into the loan book. Fintech deposits, which move in and out the same day, cannot fund loans, making them economically unviable for a bank built on lending.
Key distinction: A bank built on lending requires stable, low-velocity deposits. A treasury services platform built on clearing, custody, and cash management has the opposite characteristics. The mismatch is structural, not operational, and cannot be solved by better partnership terms.
The loan-to-deposit ratio for most US regional and community banks runs above 75%, meaning three out of every four dollars of customer deposits get deployed into the loan book. The remaining quarter sits as reserves and short-duration liquid assets. This single statistic is the reason most community bank to fintech partnerships eventually fail. For a bank whose entire business model rests on lending against deposits, fintech flows are not an opportunity. They are a problem.
How community banks make money
The business model of a community or regional bank is straightforward. Take in deposits at low cost. Lend those deposits out at higher rates. Pocket the spread. This is net interest margin, and it remains the dominant revenue source for most US banks under $250 billion in assets, as the FDIC's quarterly data consistently shows.
The lending side of the balance sheet is where the work happens. Underwriters assess credit. Loan officers build relationships. Mortgage, commercial, and small business credit account for the bulk of new originations. The bank's growth strategy is, almost without exception, about growing the loan book. Deposits exist to fund loans.
This model has been remarkably stable for decades. It also creates a specific kind of bank: one that prioritises stable, low-velocity deposits over high-velocity transactional ones. A small business checking account that maintains a balance for months is more valuable than a fintech account where money moves in and out the same day, even if the headline balance is much larger.
Why fintech deposits don't fit the model
Fintech deposits violate every assumption that community bank lending economics depend on. Money does not sit. A customer of a neobank, a marketplace payouts platform, or a stored value programme moves funds rapidly. Same-day in, same-day out. The bank sees the deposit on its balance sheet for the regulatory snapshot but cannot actually use the money to fund a loan, because the loan term is years and the deposit term is hours.
In banking terms, these deposits have very high deposit beta and low duration. They are not core funding. They are float. The bank can legally count them but cannot economically rely on them. From a treasurer's perspective, they look like a regulatory line item, not a funding source.
The 30% reduction in active correspondent banking relationships documented by the Bank for International Settlements is, in part, a direct consequence of this mismatch. Larger banks have been pricing high-velocity flows out of their networks for years. Community banks are repeating the pattern with fintech partnerships now.
The dead deposits problem
Custody products make the mismatch worse, not better. When a fintech holds large balances at a community bank for safeguarding purposes, the deposits look attractive. The numbers on the balance sheet grow. The reported deposit base expands. But these balances cannot enter the lending pool because they are regulatory safeguarded funds. Commingling them with bank capital would breach the platform's safeguarding obligations.
The same applies to fiduciary balances and segregated custody positions. Funds held in named accounts on behalf of end users sit on the bank's balance sheet but cannot be lent. They count as deposits for reporting purposes. They count against the bank's capital requirements. They produce no net interest margin because they cannot be transformed into loans.
The result is what could be called dead deposits. Money the bank holds but cannot deploy. The larger the fintech partner, the larger the dead deposit problem becomes. A community bank with a billion-dollar fintech custody relationship is carrying a billion dollars of capital cost without the offsetting lending income that justifies the model.
What happens when fintechs scale into community banks
The pattern is now well-documented. A fintech grows rapidly on a community bank partnership. Initial volume is manageable. The bank earns fees, interchange, and some genuine deposit income from lower-velocity balances. Federal Reserve research on community bank profitability shows the economics work best at this stage, when the partnership is small enough not to distort the balance sheet.
Then the fintech scales. Volume grows ten-fold. Internal escalations begin:
- Compliance team. Requests more reporting on transaction volumes and end-user activity.
- Treasury team. Flags the rising share of float and safeguarded balances on the balance sheet.
- Risk team. Raises concerns about concentration in a single fintech partner.
The bank either renegotiates the relationship on much less favourable terms, pulls back on growth, or exits the partnership entirely. This is not hypothetical. The pattern played out across multiple community banks during the rapid expansion of US fintech between 2018 and 2022, and again in the Banking-as-a-Service consolidation that followed. The banks were not the problem. The economics of their business model were the problem.
The structural answer
A bank whose business model is treasury services rather than lending does not have a deposit deployment problem. Deposits do not need to fund loans because there are no loans. Custody balances do not produce dead deposits because the institution is not measured against a loan-to-deposit ratio. Fast-moving fintech flows are not float that cannot be lent; they are exactly the kind of activity the institution exists to handle.
This is the model behind Lorum's multi-currency clearing and named account custody. Lorum is 100% reserve backed, holds no lending book, and operates as neutral infrastructure for fintech platforms and other mid-market FIs. The deposits sit safeguarded, the custody balances stay segregated, and the institution earns its revenue from clearing, custody, and cash management, not from lending against funds that need to stay still.
For platforms reviewing their banking architecture in 2026, the question is not whether a community bank partner is currently working. It is whether the underlying economics can support the next phase of growth. The 75% loan-to-deposit ratio is the answer. Fintech deposits do not fit a bank that earns its living by lending against them.
Lending-based bank vs treasury-focused institution
Two banking models produce two different economic outcomes when serving fintech treasury flows. The differences are structural, not operational.
| Dimension | Lending-based bank | Treasury-focused institution |
|---|---|---|
| Revenue source | Net interest margin (spread) | Fees on clearing, custody, FX |
| Deposit velocity preference | Low (stable funding) | High (built for movement) |
| Loan-to-deposit ratio | Above 75% | 0% |
| Capital treatment of fintech deposits | Dead deposits | Aligned with business model |
| Concentration concern | Yes (single large partner) | No (designed for many) |
| Regulatory direction alignment | Diverging | Converging |
Frequently asked questions
What is the loan-to-deposit ratio of US community banks?
The loan-to-deposit ratio for most US regional and community banks runs above 75%, meaning more than three out of every four dollars of customer deposits are deployed into the loan book. The remaining quarter sits as reserves and short-duration liquid assets.
Why can't community banks serve fintech treasury needs?
Community banks earn their margin by lending deposits at higher rates than they pay to depositors. Fintech deposits move too quickly to be lent against (same-day in, same-day out), and custody balances cannot legally enter the lending pool. The economics do not work.
What are dead deposits?
Dead deposits are balances a bank holds but cannot deploy into loans. Examples include safeguarded customer funds, fiduciary balances, and segregated custody positions. They count as deposits on the balance sheet but consume capital without generating net interest margin.
Why did community bank partnerships with fintechs fail at scale?
As fintechs scaled, the share of high-velocity float and safeguarded balances grew on the community bank's balance sheet. Compliance, risk, and treasury teams flagged the concentration. Banks renegotiated or exited the partnerships because the economics of their lending model did not support the scale of fintech flows.
What kind of bank can serve fintech treasury needs structurally?
A bank whose business model is treasury services rather than lending. These institutions hold 100% reserves, do not operate a loan book, and earn revenue from clearing, custody, and cash management fees. The deposit velocity that breaks the lending model is exactly the activity these institutions are built for.
