For community-bank executives and directors, the useful question is not whether digital-only banks are better. It is whether a branch-light competitor belongs in the same peer set at all. A digital consumer bank, a specialty SBA lender, a payments bank, and a local relationship lender can all be insured banks, but their funding, credit, fee-income, and control risks can be completely different.
This review uses Q4 2025 public Call Report data surfaced in Deep Digital Ventures Banking. The six digital, specialty, and branch-light banks below were selected as illustrative examples, not as a representative sample of all digital banks. They are useful because they show how similar legal labels can hide very different balance-sheet models.
The Fast Peer Test
Before comparing a digital-only bank with a community bank, check these five items first:
- Loan/deposit ratio: tells you whether deposits are being used mainly to fund loans or whether the bank is holding deposits for payments, cards, or other activity.
- Deposit mix: separates relationship funding from rate-sensitive or program-driven funding.
- ROA: shows earnings power, but only after you understand whether income is loan spread, fee income, gain-on-sale activity, or another model.
- Efficiency ratio: shows operating leverage, but a low branch count is not the same thing as low risk.
- Credit stress and concentration: past-due loans, charge-offs, allowances, and commercial real estate concentration decide whether earnings are durable.
If those five items point in different directions, the banks are probably false peers. Size alone is not enough.
Why These Six Banks Are in the Table
The table includes branch-light, digital, specialty-lending, sponsor-oriented, and payments-oriented banks that are commonly mentioned in fintech or digital-bank comparisons. The point is not to rank them. The point is to show where a community bank comparison breaks down.
Metrics are calculated consistently across Q4 2025 public filings. Loan/deposit equals total loans divided by total deposits. ROA uses quarterly net income and average assets. Efficiency ratio compares noninterest expense with net interest income plus noninterest income. Balance-sheet, income, deposit, loan, capital, credit-quality, charge-off, and allowance fields come from public Call Report schedules available through federal banking data sources.[1][2]
| Bank | State | Assets | Deposits | Loan/deposit | ROA | Efficiency |
|---|---|---|---|---|---|---|
| SoFi Bank, N.A. | UT | $46.57B | $39.71B | 92.5% | 1.96% | 68.1% |
| Axos Bank | CA | $27.21B | $23.51B | 98.9% | 1.77% | 42.9% |
| Live Oak Banking Company | NC | $15.06B | $13.83B | 88.3% | 0.82% | 56.1% |
| Pathward, N.A. | SD | $7.56B | $6.35B | 78.9% | 2.65% | 65.9% |
| Celtic Bank | UT | $4.79B | $3.44B | 114.3% | 4.12% | 32.1% |
| Green Dot Bank | UT | $5.20B | $4.87B | 2.3% | 1.24% | 91.8% |
The biggest lesson is visible in one column. Celtic Bank’s 114.3% loan/deposit ratio says loans exceeded deposits at quarter-end. Green Dot Bank’s 2.3% ratio says the balance sheet is not primarily a loan-funding machine. A traditional community bank with an 80% loan/deposit ratio may be closer to Live Oak or Pathward on funding behavior than to Green Dot, even if Green Dot looks closer by asset size.
How to Compare Without Building a False Peer Set
Start with business model, then size, then geography. That order matters. A $5 billion community bank with local CRE exposure and branch-raised deposits is not automatically comparable to a $5 billion payments bank with very few loans. Both may be insured depositories. They do not create risk the same way.
Use this sequence:
- Confirm the institution. Verify the legal bank name, charter, regulator, holding-company context, and active status before comparing names that may have changed through merger or rebrand.[3][4]
- Use the same filing quarter. This analysis stays on Q4 2025 because most March 31, 2026 Call Reports were due April 30, 2026, so Q1 data would not yet be complete for every filer as of April 23, 2026.[5]
- Classify the model. Label each bank as local relationship lender, digital consumer bank, specialty lender, sponsor-oriented bank, payments bank, or another model before ranking ratios.
- Compare funding quality. Look at total deposits, noninterest-bearing share, brokered deposits, uninsured deposits where reported, loan/deposit ratio, and available liquidity.
- Compare credit durability. Review loan mix, past-due and nonaccrual loans, charge-offs, recoveries, and allowances. CECL makes allowance comparisons more forward-looking, but it does not make them self-explanatory.[6][7]
- Check public supervisory history. Enforcement actions can change the meaning of a strong ratio. Growth limits, remediation requirements, capital demands, or third-party risk findings should affect the peer call.[8][9][10]
- View the spread side by side. Put the candidate bank and local peers into the peer comparison view so the differences are visible in one place.
The Ratios That Change the Story
Loan/deposit ratio is the first divider. Above-average lending intensity can signal strong origination capacity, but it also raises questions about funding cost, liquidity, and growth discipline. Very low lending intensity can be perfectly rational for a payments or stored-value model, but it should not be benchmarked like a relationship lender.
Efficiency ratio needs a model adjustment. Axos Bank’s 42.9% efficiency ratio and Green Dot Bank’s 91.8% ratio are not two versions of the same digital-bank story. One ratio may reflect operating scale and spread economics. The other may reflect a payments platform, program costs, partner servicing, technology expense, or revenue mix. For a community bank, branch expense can be overhead, but it can also be the source of low-cost deposits and local credit knowledge.
ROA is powerful only after income is unpacked. Celtic Bank’s 4.12% ROA is striking. The next question is whether earnings come from recurring spread income, specialty lending, fee income, gain-on-sale activity, or concentrated exposures. A high ROA bank can be an excellent performer and still be a poor peer for a traditional community bank.
Deposit mix decides whether growth is sticky. Digital deposits can be stable, but they can also be rate-sensitive if they arrived through yield search, embedded-finance programs, or partner channels. A community bank with slower growth and a higher noninterest-bearing deposit share may have a stronger franchise than a headline deposit-growth chart suggests.
Credit concentration is where community banks often need the extra screen. For commercial real estate, the 2006 interagency guidance highlights two supervisory indicators: construction, land development, and other land loans at 100% or more of total capital; and total CRE loans at 300% or more of total capital with 50% or more CRE growth over the prior 36 months. These are not hard lending limits. They are prompts for deeper credit review.[11]
False Peers to Avoid
| Bad shortcut | Why it misleads | Better comparison |
|---|---|---|
| Same asset size | A payments bank and a CRE-heavy community bank can have similar assets but unrelated risk drivers. | Compare asset size only after classifying funding, lending, and revenue model. |
| Lower efficiency ratio | A branch-light model may reduce visible expense while adding technology, partner, compliance, or operational risk elsewhere. | Read efficiency with deposit quality, revenue mix, and supervisory history. |
| Higher ROA | High profitability may come from specialty lending, fee income, or concentrated exposures. | Decompose income sources and test credit performance. |
| Digital label | Digital banks are not one model. SoFi, Axos, Celtic, and Green Dot show very different loan/deposit behavior. | Group banks by how they make money and how they fund assets. |
| Clean-looking ratios | Public enforcement orders can reveal control issues that ratios do not show. | Search agency enforcement records before treating a bank as a model peer. |
Red Flags for a Community-Bank Board
A digital or branch-light competitor deserves closer review when deposits are growing faster than the bank can explain, brokered or program deposits are material, loan/deposit ratio is unusually high or unusually low, efficiency looks unusually strong without clear operating evidence, or credit growth is concentrated in a category that can reprice risk quickly.
Sponsor-bank and embedded-finance activity add another layer. Recent public actions involving Cross River Bank, Blue Ridge Bank, Evolve Bank & Trust, and Lineage Bank show that third-party risk management, ledger control, BSA/AML controls, consumer compliance, and growth management can become central supervisory issues.[12][13][14][15] Those actions should not be used to paint every branch-light bank with the same brush. They should remind directors that a peer review must include control capacity, not just balance-sheet ratios.
The Synapse failure makes the same point from the nonbank side. Synapse was not a bank, but the CFPB’s public action describes alleged recordkeeping failures involving the location of consumer funds and matching records with partner banks.[16] For a bank peer review, that translates into practical questions: who owns the ledger, who reconciles it, how often exceptions clear, and whether the bank can identify end-user balances without depending on a failing vendor.
Methodology Note
Call Report schedule codes are useful for auditability, but they should not dominate the analysis. In plain English: use the balance sheet for assets and deposits, the loan schedules for loan mix, deposit schedules for funding mix, income schedules for ROA and efficiency, capital schedules for regulatory capital, credit-quality schedules for past-due and nonaccrual loans, charge-off schedules for losses, and allowance schedules for reserves. The source trail begins with FFIEC and FDIC public filing resources, FDIC BankFind, and the FFIEC National Information Center.[1][2][3][4]
The decision rule is simple: if a branch-light bank beats a community bank on efficiency, ask whether the gap comes from scale, product mix, fee income, lower branch cost, or underbuilt controls. If a community bank looks less efficient but has stronger relationship deposits, lower funding volatility, and better-known local credit, the branch network may be franchise value rather than dead weight.
FAQ
Are digital-only banks more efficient than community banks?
Only when the business model supports it. A low efficiency ratio is meaningful if revenue is durable, deposits are stable, and control costs are fully reflected. It is less useful when the bank’s model depends on partner programs, unusual fee income, or risks that do not show up in branch expense.
Are community banks safer than digital-only banks?
Safety depends on capital, credit quality, liquidity, deposit mix, interest-rate risk, and governance. A local branch network can support stable funding, but it does not offset weak credit discipline or excessive concentration.
What is the best first ratio to compare?
Loan/deposit ratio is usually the best first divider because it tells you whether the bank is mainly using deposits to fund loans. In this sample, the range from Green Dot Bank at 2.3% to Celtic Bank at 114.3% shows why a single digital-bank peer group does not work.
When is a digital bank a fair peer for a community bank?
A digital bank is a fair peer only when the funding model, loan mix, revenue sources, size, and risk controls are close enough to make the comparison useful. If the digital bank is a payments platform, sponsor bank, or specialty lender, compare it with banks that share those economics instead of forcing it into a local community-bank set.
Sources
- FFIEC Central Data Repository Public Data Distribution site: https://cdr.ffiec.gov/public/ManageFacsimiles.aspx
- FDIC current-quarter Call Report forms and instructions: https://www.fdic.gov/bank-financial-reports/current-quarter-call-report-forms-instructions-and-related-materials
- FDIC BankFind Suite: https://banks.data.fdic.gov/bankfind-suite/bankfind
- FFIEC National Information Center: https://www.ffiec.gov/NPW
- FDIC FIL-10-2026, March 31, 2026 Call Report filing deadline: https://www.fdic.gov/news/financial-institution-letters/2026/consolidated-reports-condition-and-income-first-quarter
- FASB ASU 2016-13, Topic 326, credit losses standard: https://storage.fasb.org/ASU_2016-13.pdf
- Federal Reserve interagency policy statement on allowances for credit losses: https://www.federalreserve.gov/frrs/guidance/interagency-policy-statement-on-allowances-for-credit-losses.htm
- FDIC enforcement decisions and orders: https://orders.fdic.gov/s/
- OCC enforcement actions: https://www.occ.gov/topics/laws-and-regulations/enforcement-actions/
- Federal Reserve enforcement actions: https://www.federalreserve.gov/supervisionreg/enforcementactions.htm
- Interagency CRE concentration guidance: https://www.federalreserve.gov/frrs/guidance/interagency-guidance-on-concentrations-in-commercial-real-estate-lending-sound-risk-management-practices.htm
- Cross River Bank FDIC consent order dated March 8, 2023: https://orders.fdic.gov/sfc/servlet.shepherd/document/download/0693d000007xEStAAM?operationContext=S1
- Blue Ridge Bank OCC consent order dated January 24, 2024: https://www.occ.gov/static/enforcement-actions/eaAA-ENF-2023-68.pdf
- Evolve Bank & Trust Federal Reserve enforcement action announced June 14, 2024: https://www.federalreserve.gov/newsevents/pressreleases/enforcement20240614a.htm
- Lineage Bank FDIC consent order tied to January 29, 2024 stipulation: https://orders.fdic.gov/sfc/servlet.shepherd/document/download/0693d00000BrElHAAV?operationContext=S1
- CFPB Synapse Financial Technologies action page: https://www.consumerfinance.gov/enforcement/actions/synapse-financial-technologies-inc/