The economic engine01Can a New Retail Bank Make Money at a 3.7% Net Interest Margin?
The cleanest way to judge viability is to start with net interest margin, or NIM. In the first quarter of 2026, community banks reported a 3.71% NIM, while community-bank pretax return on assets was 1.42%, according to the FDIC Quarterly Banking Profile. A startup should not underwrite itself to those mature-industry returns on day one. It begins with excess staff, underutilized systems, a small loan book, and deposit-acquisition costs that arrive before spread income does.
The market question is equally local. Before choosing a headquarters or assessment area, organizers should map branch deposits, incumbent market share, deposit growth, household income, small-business density, and competitor pricing. The FDIC Summary of Deposits market-share reports show branch counts, deposits, and share by market. A gap in personal service may be real; a gap in profitable deposits is a different question.
The non-obvious constraint is not finding borrowers; it is funding good loans with durable deposits at a cost that preserves the spread. Buying rate-sensitive deposits can make asset growth look impressive while quietly destroying the economics.
- Prove a deposit niche, not just a lending niche.
- Model NIM at least 30–50 basis points below the optimistic case.
- Expect operating leverage only after the bank passes roughly $200 million in assets under a lean one-office model.
Startup capital02How Much Capital Does It Take to Open a Retail Bank?
There is no universal statutory dollar minimum for a de novo institution. Regulators size capital to the proposed market, growth, concentrations, products, management team, and risk. Recent public approvals show the order of magnitude: Houston Bank and Trust was approved subject to at least $35 million of paid-in capital in 2025, while NOLA Bank's approval required $26.35 million. Those amounts appear in the Houston Bank and Trust order and the NOLA Bank order.
The table below is an organizer planning range, not a regulator quote. It assumes one headquarters or flagship branch, a modern digital channel, outsourced core processing, and a conventional retail-plus-local-business model.
| Startup requirement | Low | High | What the range covers |
|---|---|---|---|
| Opening paid-in capital after organizational costs | $24.0M | $35.0M | Capital cushion, early losses, asset growth, and regulatory minimums |
| Legal, application, feasibility, accounting, and pre-opening payroll | $1.5M | $4.0M | Charter counsel, projections, policies, executive recruitment, audits, and organizer costs |
| Core, digital banking, cards, data, and integrations | $1.0M | $3.5M | Implementation fees, interfaces, cybersecurity controls, testing, and first-year commitments |
| Branch, office, vault, ATM, furniture, and physical security | $0.8M | $2.5M | Leasehold improvements, hardware, surveillance, access control, cash equipment |
| Launch marketing, deposit acquisition, and contingency | $0.5M | $2.0M | Brand launch, community outreach, promotional rates, overruns, and delayed opening |
| Total organizer cash requirement | $27.8M | $47.0M | Rounded headline range: $28M–$47M |
Where a $37.4M midpoint raise goes
Nearly four-fifths remains regulatory capital; the rest gets the institution through approval, build, testing, and opening.
The FDIC organizer handbook says capital is set case by case and generally expects a Tier 1 leverage ratio of at least 8% throughout the first three years. In practice, an approval may require 10% or another tailored level. If the plan needs a $300 million balance sheet at opening maturity, a 10% target alone implies $30 million of Tier 1 capital before absorbing startup losses.
Charter and launch03How Do You Get a Bank Charter and FDIC Insurance?
Opening a retail bank is a coordinated regulatory project, not a normal license filing. Organizers choose a national or state charter, apply for federal deposit insurance, address any holding-company approvals, raise capital, hire an experienced board and executive team, build policies and systems, and pass a pre-opening examination. The OCC Charters licensing manual explains the national-bank route, while the FDIC deposit-insurance application page provides the current federal-insurance resources.
Illustrative 15–24 month opening path
The FDIC aims to act within four months after an application is accepted as substantially complete; getting to “substantially complete” is where most of the calendar risk sits.
What the three-year business plan must prove
- 1Market need: who will move deposits, why they will switch, and how the bank will serve its CRA assessment area.
- 2Management: directors and executives with direct banking, lending, risk, compliance, finance, and technology experience.
- 3Financial resilience: monthly and quarterly projections, interest-rate sensitivity, liquidity, credit losses, capital, and downside cases.
- 4Controls: BSA/AML, consumer compliance, fair lending, internal audit, vendor oversight, information security, and business continuity.
Do not sign long, non-cancelable leases or core contracts before the approval path is clear. The cheapest-looking vendor commitment can become the most expensive line item when opening slips by six months.
Systems and control04Core Banking, BSA/AML, and Cybersecurity Form the Pre-Opening Cost Stack
A startup can outsource processing, but it cannot outsource accountability. The core platform must connect deposits, loans, general ledger, online and mobile banking, debit cards, ATM access, statements, regulatory reporting, fraud monitoring, identity verification, customer service, and data analytics. Each interface adds implementation cost, testing time, vendor risk, and a failure point.
The FDIC technology and cybersecurity resources emphasize third-party risk management and due diligence for community banks. Meanwhile, the FFIEC BSA/AML Examination Manual frames the risk assessment, customer identification, monitoring, suspicious-activity reporting, and independent testing that must be operational—not merely documented.
Core pricing is rarely just a per-account fee. Model minimums, implementation charges, statement and item fees, card processing, network charges, data storage, interfaces, conversion support, annual escalators, and termination rights. The contract architecture matters almost as much as the software.
Run the formula at 50%, 100%, and 150% of planned account volume. A low-volume minimum can make the early cost per account look punishing; a poor variable-price schedule can make growth expensive later.
The correct build decision is usually “small number of proven vendors, tightly governed,” not “cheapest stack.” Every new product should have an owner, policy, risk assessment, reconciliations, complaint path, and backup procedure before launch.
Monthly burn05What Does It Cost to Run the Bank Each Month?
A lean one-office institution can still burn $560,000 to $1.37 million per month before deposit interest and credit-loss provisions. The low case assumes a compact team, outsourced infrastructure, and restrained marketing. The high case reflects a larger executive bench, broader product set, stronger deposit promotions, and heavier compliance or technology support.
| Monthly operating expense | Low | High | Main driver |
|---|---|---|---|
| Payroll, payroll tax, and benefits | $250K | $500K | 20–35 FTE across management, risk, lending, operations, and branch service |
| Core, digital, cards, telecom, and data | $100K | $250K | Contract minimums, accounts, transactions, interfaces, and support |
| Occupancy, facilities, cash handling, and security | $45K | $120K | Market rent, vault/ATM service, guards, utilities, and maintenance |
| Compliance, legal, audit, and board costs | $50K | $140K | Independent testing, counsel, audits, exams, directors, and training |
| Cybersecurity, fraud, insurance, and loss prevention | $35K | $100K | Monitoring, incident response, fidelity, D&O, fraud tools, and losses |
| Marketing and deposit acquisition | $50K | $175K | Community launch, direct response, incentives, and relationship staff |
| Other administration, travel, supplies, and training | $30K | $80K | Reporting, subscriptions, recruitment, travel, postage, and reserves |
| Total monthly noninterest expense | $560K | $1.365M | Annualized: $6.72M–$16.38M |
Midpoint monthly noninterest expense
Payroll dominates the burn rate; the second-order issue is that technology and compliance do not shrink much when account volume is low.
Use local compensation, not national averages, in the hiring plan. The BLS May 2025 industry-specific wage tables are a useful starting point, but a credible application needs actual market quotes for the CEO, chief financial officer, chief credit officer, BSA officer, compliance lead, operations staff, lenders, and branch personnel.
Revenue architecture06How Do Deposits, Loans, and Fees Build Revenue?
A retail bank earns most of its money by gathering deposits, deploying them into loans and securities, and keeping the difference between asset yield and funding cost. Fee income helps, but it rarely rescues a weak spread model. Account fees, debit interchange, treasury management, mortgage gains, safe-deposit rent, and referral revenue should be modeled separately because their cost and volatility differ.
| Illustrative $250M-asset base case | Annual amount | Calculation |
|---|---|---|
| Average earning assets | $225.0M | 90% of average total assets |
| Net interest income | $8.35M | $225.0M × 3.71% NIM |
| Noninterest income | $1.50M | Cards, account services, treasury, mortgage, and other fees |
| Total operating revenue | $9.85M | Net interest income + noninterest income |
| Noninterest expense | ($7.50M) | Lean operating case within the monthly range above |
| Credit-loss provision | ($0.80M) | Illustrative through-the-cycle provision |
| Pretax income | $1.55M | About 0.62% pretax ROA |
| Estimated income tax | ($0.33M) | Illustrative 21% federal rate before state effects and tax attributes |
| Net income | $1.22M | About 0.49% ROA |
Deposit products should be priced as funding instruments, not merely customer offers. A 50-basis-point increase in cost on $200 million of interest-bearing deposits costs $1 million per year. That single change can erase most of the example's net income. The bank should therefore track deposit beta, mix of noninterest-bearing balances, renewal behavior, and concentration by depositor.
Customer trust also shapes funding quality. The standard FDIC coverage is $250,000 per depositor, per insured bank, per ownership category, as explained in the FDIC deposit-insurance guide. Relationship managers should understand coverage structures, but the financial model should still stress uninsured and concentrated deposits for runoff.
A smaller base of operating accounts with payroll, merchant, treasury, and lending relationships is usually more valuable than a larger base of promotional savings accounts. The first group produces sticky balances and fee income; the second leaves when the rate does.
Break-even and ramp07When Does a De Novo Bank Break Even?
Using $7.50M of expense, $0.80M of provision, $1.50M of fees, and a 3.71% NIM: ($7.50M + $0.80M − $1.50M) ÷ 3.71% = $183.3M of average earning assets.
If earning assets equal 90% of total assets, the example breaks even around $204 million of average assets. That is the planning answer—not a guarantee. If NIM compresses to 3.40% and operating expense rises to $9 million, required average earning assets climb to roughly $244 million. Small changes in spread and staffing create large changes in scale required.
Illustrative pretax earnings ramp
Annual profit can turn positive in year three while cumulative launch losses remain unrecovered.
Account count alone is a poor break-even unit. A more useful operating translation is roughly 6,000 funded relationships averaging $35,000, or another mix that supports about $200 million in assets without expensive wholesale funding. The model must distinguish opened accounts, funded accounts, primary relationships, and deposits that remain after promotional pricing ends.
Founder economics08How Much Can a Bank Founder or Executive Owner Make?
Owner income is not bank revenue, and bank profit is not automatically distributable. The FDIC's organizer guidance says de novo plans should not assume dividends during the first three years; dividends should come only from net operating income after the allowance for credit losses and capital are adequate. That restriction is set out in the FDIC organizer handbook.
| Founder position | Salary | Bonus | Illustrative dividends | Total cash |
|---|---|---|---|---|
| Year 1 executive founder | $175K–$250K | $0 | $0 | $175K–$250K |
| Year 3 executive founder | $200K–$325K | $0–$50K | $0 | $200K–$375K |
| Mature executive founder with 10% ownership | $225K–$375K | $25K–$100K | $48K–$150K | $298K–$625K |
| Mature passive investor with 5% ownership | $0 | $0 | $24K–$75K | $24K–$75K |
The mature dividend range assumes $2.4 million to $5 million of net income, a 20% to 30% payout ratio, and the ownership percentages shown. It excludes stock appreciation, option value, taxes, dilution, and any regulatory or board decision to retain all earnings. Salary also must be reasonable, documented, and consistent with the role.
A founder who needs near-term distributions is a poor fit for a de novo bank. The first return is usually salary for real work; the investment return comes later through retained earnings, book-value growth, dividends, or an eventual transaction.
Balance-sheet survival09Liquidity, Capital, and Asset-Liability Management Decide Survival
A bank can report accounting profit and still run into trouble if deposits leave, securities are underwater, loans are illiquid, or capital cannot support planned growth. This is why the opening model needs linked schedules for deposits, loans, securities, borrowings, cash, capital, interest-rate repricing, credit losses, and liquidity—not a single income statement.
At $30M of Tier 1 capital and a 10% internal target, the bank supports about $300M of average assets. If ramp losses reduce capital to $25M, capacity falls to about $250M unless earnings recover or investors add capital.
That is why growing faster than plan can be dangerous. New loans consume capital, liquidity, underwriting time, and servicing capacity. Fast deposit growth can also create rate and concentration risk if balances come from brokers, listing services, one fintech partner, or a handful of large commercial customers.
The FDIC 2026 Risk Review highlights interest-rate, liquidity, credit, operational, and market risks across the industry. For a new institution, the board should run at least three linked stresses: a 100–200 basis-point deposit-cost shock, a 10%–20% deposit runoff, and credit losses materially above plan.
Set growth gates. Do not add a second branch, new lending vertical, or major partner until capital, liquidity, staffing, compliance testing, and service levels are within board-approved bands for several consecutive months.
Board dashboard10Which KPIs Should the Board Watch Every Month?
The dashboard should connect directly to the approved plan. A good board pack does not drown directors in 100 numbers; it shows whether the bank is gathering the right deposits, producing risk-adjusted assets, controlling overhead, protecting capital, and staying liquid.
| KPI | Formula | Planning benchmark | Decision it drives |
|---|---|---|---|
| Net interest margin | Net interest income ÷ average earning assets | Base near 3.5%–3.9%; investigate a 20+ bp miss | Loan pricing, deposit rates, securities mix, hedging |
| Cost of deposits | Deposit interest expense ÷ average deposits | Track by product and channel; rising faster than asset yield is a warning | Promotional pricing and relationship strategy |
| Loan-to-deposit ratio | Net loans ÷ deposits | Often 70%–90% in planning; stress above policy limits | Funding needs and loan-growth pace |
| Efficiency ratio | Noninterest expense ÷ net operating revenue | De novo may exceed 100%; target below 70%, then toward low 60s | Hiring, branch, and vendor spend |
| Pretax ROA | Pretax income ÷ average assets | Negative in ramp; 0.5%–1.0% emerging; 1%+ mature target | Capital generation and investor return |
| Tier 1 leverage ratio | Tier 1 capital ÷ average total assets | Maintain the approval floor plus a board cushion | Asset growth, dividends, and capital raise timing |
| Nonperforming assets | Nonaccrual loans + OREO ÷ total assets | Trend and concentration matter; investigate early migration | Underwriting, reserves, workout staffing |
| Deposit concentration | Top 20 depositors ÷ total deposits | Set internal limits and runoff scenarios | Liquidity buffer and relationship diversification |
| Primary relationship rate | Customers with salary/operating activity + two products ÷ funded customers | Directional target 35%–60%, depending on niche | Product design and retention investment |
The benchmark ranges above are planning targets, not universal regulatory standards. Peer comparisons should match asset size, charter, geography, business mix, and age. The board should also see variance to the approved three-year plan, a rolling 12-month forecast, and the capital and liquidity effect of management's corrective actions.
Downside control11What Can Break the Model—and What Does It Cost?
The biggest failures are linked. Weak deposit traction forces higher rates; higher rates compress NIM; compressed NIM delays profit; delayed profit consumes capital; thinner capital limits growth just when the bank needs scale. Credit, compliance, fraud, and technology problems can accelerate the same spiral.
| Risk | Early trigger | Illustrative financial impact | Mitigation |
|---|---|---|---|
| Deposit pricing war | Cost of deposits rises 50 bp above plan | About $1.0M annual cost on $200M of deposits | Relationship pricing, noninterest-bearing balances, slower asset growth |
| Loan concentration | One property type or industry exceeds policy | A 2% loss on a $75M book equals $1.5M | Limits, participations, independent review, stronger collateral stress |
| Opening delay | Core testing, hiring, or approval slips six months | $3.4M–$8.2M extra burn at stated monthly costs | Cancelable milestones, contingency capital, phased hiring |
| Fraud or cyber event | Account takeover, vendor outage, data loss | Direct loss plus investigation, notification, restitution, and attrition | Layered controls, tested response, vendor recovery, insurance |
| Compliance breakdown | Backlogs, repeat exceptions, weak complaints or monitoring | Remediation staff, consultants, customer restitution, growth restrictions | Independent testing, clear ownership, board escalation, quality assurance |
| Vendor lock-in | Minimums and fees outrun account growth | $500K–$2M annual avoidable cost in a complex stack | Volume bands, exit rights, interface inventory, full TCO model |
Consumer compliance should be built into product design, marketing, underwriting, servicing, complaints, and vendor oversight. The CFPB supervision and examination resources are one reference point for the federal consumer-financial-law environment, alongside the bank's primary prudential regulator and state requirements.
The spreadsheet hides sequencing. A credit problem usually appears after the loan was booked, but its capital and staffing cost arrives when growth is already consuming resources. Build the downside reserve before the growth, not after the first delinquency wave.
Funding and return12Is Starting a Retail Bank Worth It? Payback and the Honest Verdict
This can be an attractive long-duration investment when the group has a genuine deposit franchise, experienced management, patient capital, and a path to several hundred million dollars of assets without buying every dollar of funding. It is a poor fit for founders seeking a low-capital business, quick distributions, or a passive side investment.
Funding is primarily common equity from organizers and outside investors. Holding-company debt may be possible in some structures, but regulators and investors will test the source of repayment, especially when the bank is not expected to pay dividends. SBA financing is not the answer: federal rules list financial businesses primarily engaged in lending, including banks, as ineligible for SBA business loans under 13 CFR 120.110.
This formula is incomplete for a bank because most earnings are retained as capital and remain in tangible book value. Investors should model dividends plus growth in book value plus a conservative exit value.
| Investor case | Initial equity | Year-5 economics | Illustrative realization | Economic payback |
|---|---|---|---|---|
| Conservative | $30M | $250M assets, $1.5M net income | Year 10 at 1.0× assumed tangible book plus limited dividends | 10+ years |
| Base | $30M | $400M assets, $4.0M net income | Year 8 at 1.25× assumed tangible book plus cumulative dividends | About 8 years |
| Upside | $30M | $600M assets, $7.0M net income | Year 7 at 1.5× assumed tangible book plus cumulative dividends | About 7 years |
The exit multiples above are explicit assumptions, not market guarantees. A bank with weak credit, concentrated deposits, compliance issues, or poor growth may trade below book value or have no attractive buyer. A bank with scarce deposits, clean credit, strong management, and durable earnings may command more. The model should therefore show value at 0.9× book1.0× book1.25× book1.5× book rather than relying on one exit price.
- Worth pursuing: experienced team, $28M–$47M available, underserved deposit relationships, disciplined credit, and investors who can wait seven to ten years.
- Do not proceed: the plan depends on promotional deposits, immediate dividends, rapid branch expansion, or a single fintech/partner channel.
- Final test: run a linked financial model with lower NIM, slower deposits, higher payroll, six months of delay, larger credit losses, and no dividends for three years. If capital still holds, the idea is ready for serious pre-filing work.
