Marketplace lending and fintech-originated

Does your product fit here?

This asset class covers loans originated through digital-first platforms that use technology-enabled underwriting. The defining characteristic isn’t the collateral type (which can span personal loans, small business, and sometimes real estate) but the origination model: platform-based distribution, algorithmic credit decisioning, and often a bank partner funding arrangement.

The three origination models you’ll encounter:

Marketplace model: The platform matches borrowers with investors but doesn’t hold credit risk. The platform earns an origination fee. LendingClub (pre-2020), Prosper’s investor program, and Funding Circle work this way.

Balance sheet model: The fintech originates loans and holds them on its own balance sheet, using warehouse facilities and term ABS for funding. SoFi, Upstart (whole loans), Affirm, and Avant follow this model.

Bank partner model: The fintech handles marketing, underwriting, and servicing while a partner bank (Cross River, WebBank) is the lender of record. The bank sells the loans to the fintech shortly after origination. This model carries true lender considerations that capital providers will scrutinize.

Products that fit here:

Marketplace consumer loans: Loans originated through platforms that match investors with borrowers
Fintech-originated personal loans: Personal installment loans from digital-native originators regardless of whether they use a balance sheet or marketplace model
Fintech small business loans: Digital platform-originated term loans and lines of credit to small businesses (Fundbox, OnDeck, Kabbage)
Embedded lending: Loans originated at point-of-sale through fintech rails, merchant-integrated but underwritten with fintech credit models

What does NOT fit here:

Traditional bank consumer loans: Even with a modern tech stack, traditional banks are covered in Bank Balance Sheet
BNPL with less than 12 month tenor: See BNPL Receivables
Payday and title loans: Regulatory profile is too different
Fintech-facilitated mortgage: Falls under Non-Agency RMBS or Home Equity and HELOCs depending on product

How lenders will classify you

Platform Stage	Characteristics	Available Structures
Established fintech	5+ years operations, $500M+ cumulative originations, through-cycle performance data	Full range: warehouse, term ABS, forward flow
Growth fintech	2-5 years, $100M-$500M originations	Warehouse available; term ABS emerging
Early-stage fintech	< 2 years, < $100M	Forward flow or whole loan sale only

Your access to capital depends heavily on track record. If you’ve been originating for 18 months and want a warehouse facility, be prepared for a difficult conversation. Most capital providers want to see at least 24 months of static pool data before extending a revolving commitment.

Market benchmarks and comps

Performance benchmarks: fintech consumer (personal loans)

Metric	Prime (680+)	Near-Prime (620-679)	Subprime (<620)
CDR (annualized)	2-5%	6-12%	15-30%
CNL (life of pool)	4-8%	10-18%	20-40%
CPR (annualized)	20-35%	15-25%	10-18%
Loss severity	85-95%	85-95%	90-98%
WAL	1.5-2.5 yrs	1.5-2.5 yrs	1.2-2.0 yrs

Performance benchmarks: fintech small business

Metric	Strong Credit	Moderate Credit
CDR (annualized)	3-8%	10-20%
Severity	60-80%	70-85%
WAL	1.0-2.0 yrs	0.8-1.5 yrs

Fintech vs. traditional originator performance

Fintech platforms typically show higher vintage volatility than traditional lenders. There are structural reasons for this:

Underwriting model iteration: Fintechs update their models frequently (sometimes quarterly). Each model version creates a potential basis between vintages. A capital provider financing your 2023 originations needs to understand whether your 2022 static pool data is still relevant to loans originated under a different model version.

Prepayment dynamics: Digital refinancing makes prepayment frictionless. CPR for fintech personal loans runs 20-35% for prime borrowers, higher than traditional bank-originated product. This reduces duration risk but also compresses interest income.

Early delinquency predictiveness: Fintech collections practices vary widely. A 30 DPD rate at a fintech with aggressive early-stage contact may mean something different than the same metric at a traditional servicer. Roll rate analysis needs calibration to the specific platform’s patterns.

What “good” performance looks like

Prime portfolio CDR below 4% at month 12 for each vintage cohort
Vintage-over-vintage consistency: each cohort at the same months-on-book should look similar to prior cohorts
CPR within +/- 25% of base case projections
No material correlation between model version changes and credit box expansion
Approval rate changes justified by model improvements with supporting backtest data

Red flag performance benchmarks

Vintage-over-vintage CDR deterioration exceeding 20% at the same seasoning point
Model version change coinciding with credit box expansion (more borrowers approved at same or lower scores)
Approval rate increases above 15% without documented explanation
Loan volume growth exceeding 3x year-over-year (you may be outrunning your underwriting capacity)
Any vintage where month 6 CDR exceeds 1.5x the prior vintage at the same point

What lenders and investors focus on

1. Platform quality and track record

Capital providers evaluate the originator as much as the collateral. For fintech platforms, the key questions are:

Operating history: How long have you been originating? Capital providers want to see at least 24-36 months of operations before extending significant credit. Through-cycle performance data (did you exist before 2020? How did your portfolio perform during COVID?) is increasingly important.

Management team: Prior lending or fintech experience matters. A team with consumer credit backgrounds at Capital One, Discover, or established fintechs gets more benefit of the doubt than first-time founders.

Financial condition: What’s your runway? Capital providers will pull your financials and calculate burn rate. If you have less than 12 months of cash without a clear path to profitability or additional funding, expect pushback on commitment size and term.

Regulatory standing: Any regulatory inquiries, consent orders, or state AG investigations? These need to be disclosed and will affect terms or kill deals outright.

2. Underwriting model integrity

Your credit model is central to the diligence process. Capital providers will want to see:

Model validation documentation: Third-party model validation from a recognized firm (Moody’s Analytics, FICO, Aite-Novarica) showing that your model’s predictions correlate with actual default outcomes.

Backtesting results: How did the model perform out-of-sample? What’s the Kolmogorov-Smirnov (KS) statistic or Gini coefficient? How stable are variable importance rankings across time?

Version control: What version of your model originated the loans in your current portfolio? When did you last update? What changed? Capital providers track this because a new model version without sufficient seasoned performance data creates uncertainty.

Alternative data compliance: If you use non-traditional variables (bank account data, education history, employment verification through payroll APIs), you need fair lending testing documentation showing no disparate impact.

3. Bank partner considerations

If you use a bank partner model (Cross River, WebBank, Celtic), capital providers will dig into:

True lender risk: The legal question is whether the fintech, not the bank, is the “true lender” despite the bank’s name on loan documents. If a court determines the fintech is the true lender, state usury caps may apply to loans that exceed those rates.

Current litigation landscape: Madden v. Midland and subsequent cases created uncertainty. State attorneys general have been active. Capital providers will want a true lender analysis memorandum from qualified counsel.

Bank partner concentration: If you use a single bank partner, what happens if they exit? Cross River has regulatory scrutiny. WebBank is Utah-based (ILC charter) with its own dynamics. What’s your backup plan?

Contractual protections: Loan purchase agreements should include representations about valid origination, bank partner commitment to continue the program, and notice periods if the relationship changes.

4. Data quality and servicing capability

Loan tape completeness: Capital providers will audit your data. Missing fields, inconsistent formatting, and data quality issues signal operational immaturity. Common problem areas: income verification documentation, bank partner loan numbers, payment history completeness.

Static pool availability: You need monthly static pool data going back at least 24 months showing CDR, CPR, and CNL by origination vintage. If you can’t produce this in a standardized format quickly, you’re not ready for institutional capital.

Servicing infrastructure: Do you service in-house or outsource? What’s your collections strategy? Days to first contact, escalation procedures, charge-off timing. Capital providers want to see a documented, consistent process.

Backup servicer: For warehouse facilities, you’ll need at least a cold backup servicer arrangement. For term ABS, warm or hot standby is typically required.

5. Funding diversification

Capital providers assess concentration risk:

Current capital structure: What’s your debt-to-equity mix? How many funding sources? Single-warehouse concentration is a risk factor.

Funding runway: If your primary capital provider exits, how long can you operate? What’s your contingency plan?

Path to diversification: Are you building toward multiple warehouses? Term ABS eligibility? The ability to access public markets reduces capital provider risk.

Typical structures used

Forward flow

The entry point for most early-stage fintechs.

How it works: A capital provider (typically a credit fund) commits to purchase your loan production on an ongoing basis at agreed terms. You originate, they buy, usually within days of origination.

Economics:

Pricing is yield-based. Capital provider targets 12-20% net yield depending on credit quality.
You receive par minus a discount, typically 2-6 points depending on credit tier and expected losses.
Volume commitment minimums are common (e.g., $5M/month minimum).

Best for: Originators under $75M/year who want simplicity and predictability. No facility management, no compliance testing, no trigger monitoring.

Transition path: Most fintechs use forward flow to build track record, then graduate to a warehouse facility at 18-36 months as economics improve with scale.

Whole loan sale

Periodic bulk purchases rather than continuous flow.

How it works: You accumulate loans on your balance sheet (or in a warehouse), then sell portfolios periodically to a buyer.

Economics:

Pricing: discount to par based on expected loss and buyer’s yield requirements
Prime portfolios: par minus 2-4 points
Near-prime: par minus 4-8 points
Subprime: par minus 8-15+ points

Best for: Portfolio cleanup, capital management, or if forward flow economics don’t work for your production profile.

Buyers: Credit funds, insurance companies, regional banks, and specialty whole loan investors.

Warehouse facility

The standard structure for growth-stage fintechs with established track records.

Typical terms:

Credit Tier	Advance Rate	Pricing
Prime (680+)	80-90%	SOFR + 175-350 bps
Near-prime (620-679)	70-82%	SOFR + 275-400 bps
Subprime (<620)	55-70%	SOFR + 350-500 bps

Illustrative pricing. See pricing disclaimer.

Facility size: $25M-$300M typical for growth-stage; $300M-$1B+ for established platforms

Revolving period: 12-24 months

Key structural features specific to fintech:

Bank partner compliance reps: If you use a bank partner, reps confirming valid origination, proper documentation, and bank partner standing
Model change notification: You must notify the lender of material changes to your credit model, typically with 30-day advance notice
Originator financial covenants: Minimum liquidity (often $5-15M or 3-6 months operating expenses), minimum tangible net worth, maximum leverage
Servicing standards: Detailed requirements around collections timing, modification limits, and reporting
Backup servicer: Cold standby at minimum; warm for larger facilities

Term ABS (144A and Reg d)

Available once you have sufficient track record and scale.

Requirements:

3+ years of operations
$200M+ annual origination run rate
24-36 months of static pool data
Audited financials
Third-party model validation

Major fintech ABS issuers: LendingClub (LCT), SoFi (SOFI Trust), Upstart (UPST Trust), Avant (AVNT), Marlette (Best Egg), Oportun (OPRT)

Typical deal size: $150M-$750M

Rating agencies: S&P and Moody’s for larger deals; KBRA and DBRS are active in the space and may be more constructive for newer issuers.

Asset-class-specific structural features

True lender considerations

If you use a bank partner, this is the central legal and regulatory issue.

The problem: Your loans are originated with the bank as lender of record, allowing you to benefit from federal preemption of state usury laws. But regulators and courts have questioned whether the fintech, not the bank, is the “true lender” based on which party bears the economic risk and makes credit decisions.

Why it matters: If a court determines you’re the true lender, state rate caps apply. Loans above those caps may be unenforceable or subject to penalties. Your portfolio may include substantial exposure to states where your rates exceed caps.

Structural protections capital providers look for:

Clear documentation that bank retains meaningful economic interest and decision-making
Loan purchase agreement representations about valid origination
Bank partner commitment letters
Rate caps or state stratification showing limited exposure to high-risk jurisdictions
True lender analysis memorandum from qualified counsel

Bank partner risk mitigation

The major bank partners (Cross River, WebBank, Celtic) each have their own dynamics:

Cross River: Largest fintech partner by volume. Has faced regulatory scrutiny. Well-capitalized but any consent order could affect your program.

WebBank: Utah ILC charter. Long track record in fintech partnerships. Less regulatory friction historically.

Celtic Bank: Active in personal loan space. Smaller than Cross River and WebBank.

What happens if your bank partner exits? This is a real risk. Capital providers want to know:

Do you have a backup partner arrangement?
Do you hold state licenses that would allow you to originate directly (at capped rates)?
What’s the transition timeline if you need to switch?

Alternative data and model governance

Many fintechs differentiate through alternative data: bank account transaction history, education credentials, employment verification through payroll APIs, rent payment history.

Capital provider concerns:

Fair lending compliance: Alternative variables must not create disparate impact on protected classes. You need testing documentation showing your model doesn’t discriminate.

Regulatory uncertainty: The CFPB and state regulators have taken varying positions on alternative data use. Model variables that seem predictive today may face future regulatory challenge.

Model documentation requirements: Capital providers want to understand what variables you use, how they’re weighted, and how predictions translate to approval/pricing decisions.

Usury and state rate caps

This connects directly to the bank partner and true lender issues.

The bank partner advantage: Banks can export their home-state rates nationwide. A loan originated by WebBank (Utah) at 29.99% APR is valid in California even though California caps non-bank lender rates at 36% for larger loans.

The risk: If a court determines you’re the true lender, state caps apply. Your 29.99% APR loan in a state with a 25% cap becomes problematic.

Portfolio stratification: Capital providers will stratify your portfolio by state and interest rate to understand exposure. High concentration in states with active true lender litigation (e.g., California, Colorado, Illinois) is a concern.

Rating agency treatment

S&P approach

S&P evaluates fintech originators through an operational review framework:

Platform assessment: Management quality, funding stability, operational infrastructure, regulatory compliance history

Loss curve derivation: S&P builds expected loss curves from your static pool data, applying stress multiples to reach investment-grade scenarios

Stress assumptions:

AAA: 2.5-4.0x base case CDR
Prepayment stress at both fast (extension risk) and slow (excess spread erosion) scenarios
Model change sensitivity analysis for platforms with recent underwriting changes

Moody’s approach

Moody’s uses an Originator Assessment framework (OA1 through OA5):

OA1-OA2: Established platforms with through-cycle data, strong financial condition, proven management. Most favorable loss assumptions.

OA3: Adequate platforms with limited cycle experience. Moderate haircuts to static pool data.

OA4-OA5: Weak platforms, limited track record, financial concerns. Significant loss assumption increases.

For fintech originators, Moody’s applies higher loss assumptions to platforms without full credit cycle data (pre-2020 performance).

KBRA and DBRS

Both agencies are active in the fintech space and may offer more constructive views for emerging platforms:

Lower minimum deal size requirements
Potentially more credit for alternative data model performance
Faster turnaround on new issuer engagements

Typical enhancement levels

Rating	Prime	Near-Prime	Subprime
AAA/Aaa	18-25%	30-40%	45-60%
A/A2	10-15%	18-25%	28-38%
BBB/Baa2	5-9%	10-15%	18-25%

These ranges assume established platforms with adequate data. Newer issuers or platforms with recent model changes will face higher enhancement requirements.

Diligence focus areas

Platform/originator diligence

On-site visit: Capital providers will want to see your operations. Headquarters visit, technology infrastructure review, team meetings.

Management team: Background checks, reference calls, prior experience verification. If your CEO was at a lender that had compliance issues, expect questions.

Financial condition: Audited financials for 2+ years, interim financials, cap table, funding runway analysis, path to profitability or next funding round.

Regulatory review: Litigation search, regulatory inquiry disclosure, consent order history, state licensing status.

Credit model diligence

Third-party validation: Model validation report from a recognized firm (Moody’s Analytics, FICO, Aite-Novarica). Shows model methodology, backtesting results, variable stability.

Performance vs. prediction: How well did the model’s predicted default rates match actual defaults across vintages?

Fair lending testing: Disparate impact analysis documentation. If you use alternative data, this is especially important.

Model change log: Version history, what changed and why, approval process, board or risk committee sign-off.

Data and static pool analysis

Loan tape audit: Field completeness, formatting consistency, reconciliation to financial statements

Static pool construction: Vintage-level CDR, CPR, CNL tracked monthly. Minimum 24 months history preferred.

Stratification: FICO distribution, DTI distribution, state concentration, loan size distribution, term distribution

Roll rate analysis: 30-60-90-charge-off transitions by vintage and model version

Pre/post model change comparison: If you changed your model, how do loans originated under the new model compare to prior vintages at the same seasoning?

Servicing and operations

Platform walkthrough: Servicing system demonstration, collections process documentation, default management procedures

Backup servicer: Status (cold/warm/hot), readiness assessment, data transfer protocols

Borrower communication: Contact frequency, complaint handling, regulatory complaint history (check CFPB database)

Business continuity: Disaster recovery documentation, key person redundancy

Legal and compliance

Bank partner documentation: Loan purchase agreement, bank partner commitment letter, program agreement

True lender memorandum: Legal analysis from qualified counsel addressing true lender risk

State licensing: If you hold direct lending licenses, status verification

Marketing and disclosure review: Compliance with TILA, ECOA, state disclosure requirements

Active participants

Major fintech platforms by segment

Consumer (Prime/Super-prime):

SoFi (balance sheet model, now a bank)
LendingClub (acquired Radius Bank, now a bank)
LightStream (SunTrust/Truist subsidiary)

Consumer (Near-prime):

Avant
Upstart (and Upstart partner banks)
Best Egg (Marlette Funding)
Prosper

Consumer (Subprime):

Oportun
OppFi
Possible Finance

Small Business:

Fundbox
Kabbage (now part of American Express)
Bluevine
OnDeck
Lendio (marketplace)

Capital providers active in fintech space

Banks (warehouse lenders):

Goldman Sachs
Jefferies
Deutsche Bank
Barclays
Credit Suisse (legacy positions)
Regional banks (Customers Bank, Cross River for some programs)

Credit Funds:

Ares Capital
Varde Partners
Waterfall Asset Management
Blue Owl
Castlelake
Cerberus

Insurance (rated notes):

Athene
MassMutual
Principal

Bank partners

Cross River Bank (largest by volume)
WebBank (Utah ILC)
Celtic Bank
Stride Bank
Coastal Community Bank

Law firms with fintech expertise

Mayer Brown
Chapman and Cutler
Latham & Watkins
Paul Hastings
K&L Gates
Morrison & Foerster
Sidley Austin

Red flags and off-market characteristics

Originator-level red flags

Limited operating history: Platform less than 2 years old without significant management track record from prior fintech or consumer lending roles
Single bank partner with no backup: What’s your plan if they exit?
Model changes aligned with credit expansion: If you updated your model and approval rates jumped 20% at the same time, that’s a flag
Approval rate growth exceeding volume growth: You may be loosening credit standards faster than you’re growing
Funding runway below 12 months: If you can’t make it to your next funding round, capital providers bear the transition risk
Active regulatory issues: State AG investigation, CFPB inquiry, consent order in progress

Portfolio-level red flags

Rapid vintage deterioration: Each cohort performing worse than the prior one at the same months-on-book
High concentration in true lender risk states: Heavy exposure to California, Colorado, Illinois, or other states with active litigation
Loans above state usury thresholds: Large portion of portfolio above state caps if preemption doesn’t apply
Limited static pool history: Less than 24 months of data makes loss projection unreliable
Model version gaps: Insufficient performance data for current model version
Thin file concentration above 25%: Borrowers with limited credit history are harder to underwrite accurately

Structural red flags

No backup servicer: Especially problematic for warehouse facilities with single servicer
Weak model change provisions: If you can change your model without notice, capital providers can’t manage risk
Bank partner regulatory issues: Partner bank under consent order or facing examination pressure
Insufficient financial covenants: TNW, liquidity, and leverage floors that don’t provide adequate cushion
Missing or inadequate true lender analysis: No legal memorandum, or analysis that doesn’t address current case law

What will kill a deal

Active regulatory enforcement action against the platform
Pending true lender litigation with adverse facts
Model validation showing significant model decay or discrimination
Financial distress (less than 6 months runway with no clear path forward)
Vintage performance materially worse than static pool data provided during screening
Undisclosed regulatory inquiries discovered during diligence

Worked example: fintech consumer loan warehouse economics

Platform profile:

Growth-stage fintech, 3 years of operations
$200M annual origination
Near-prime personal loans
Average balance: $8,000
Average term: 36 months
Average APR: 22%
Expected CNL: 12%
Expected CPR: 18%

Warehouse facility terms:

Facility size: $75M
Advance rate: 75%
Pricing: SOFR + 325 bps
Assume SOFR at 5.0%, so all-in rate is 8.25%
Commitment fee: 50 bps on unused

Capital structure:

$75M borrowing capacity requires $100M eligible collateral
Originator equity contribution: $25M (25% of pool)
Monthly utilization target: 90% ($67.5M drawn)

Return analysis for the originator:

Component	Rate
Gross yield (APR to borrower)	22.0%
Less: Financing cost (75% x 8.25%)	(6.2%)
Less: Expected credit losses	(12.0%)
Less: Servicing cost	(1.0%)
Less: Platform overhead allocation	(2.0%)
Net originator margin	~0.8%

Illustrative pricing. See pricing disclaimer.

On $100M of originated volume, that’s roughly $800K in annual margin before G&A, customer acquisition, and corporate overhead.

Return analysis for the capital provider:

Yield on funded amount: 8.25% on the 75% advance rate portion
First loss protection: 25% equity cushion (more than covers 12% expected CNL)
Effective position: Senior secured, with 2x+ loss coverage
Target return: SOFR + 325 bps for what is effectively a low-single-digit expected loss position

This is why fintech originators are economically motivated to graduate from warehouses to term ABS. The same portfolio priced at term ABS levels (SOFR + 150-200 for AAA notes) dramatically improves originator economics.

Note: These economics assume steady-state operations. Early-stage fintechs typically run negative unit economics as customer acquisition costs exceed lifetime value until scale improves.