lev-os Analyzing Digital Lending Platforms
Evaluates digital lending models across credit underwriting methodology, capital and funding structure, regulatory posture, and technology infrastructure to produce a structured risk-and-opportunity assessment.
When To Use
- Evaluating an online lending platform as a potential investment, acquisition target, or partnership
- Assessing credit model soundness for a marketplace lender, embedded lending product, or BNPL provider
- Reviewing a digital lender's regulatory compliance posture across applicable state and federal frameworks
- Analyzing funding structure resilience (warehouse lines, securitization, balance-sheet lending) under stress scenarios
- Benchmarking a platform's unit economics, default rates, and portfolio performance against industry cohorts
Inputs To Gather
- Business model overview: Lending vertical (consumer unsecured, SMB, point-of-sale, student, auto), origination channel, geographic footprint
- Credit model documentation: Underwriting variables, model type (logistic regression, gradient-boosted trees, neural network), use of alternative data (bank transaction data, employment verification APIs, behavioral signals)
- Portfolio performance data: Vintage curves, delinquency buckets (30/60/90 DPD), charge-off rates, recovery rates, net loss rates by cohort
- Funding structure: Capital sources (warehouse facilities, forward-flow agreements, whole-loan sales, ABS issuance, balance-sheet), advance rates, covenants, maturity profiles
- Regulatory filings and licenses: State lending licenses held, bank partnership or rent-a-charter arrangements, any pending enforcement actions or consent orders [VERIFY]
- Financial statements: Revenue breakdown (interest income, origination fees, servicing fees, gain-on-sale), operating expenses, contribution margin per loan
- Technology stack: Loan origination system, servicing platform, fraud detection tools, API integrations
Workflow
Map the business model — Classify the platform by lending vertical, origination channel (direct-to-consumer, embedded, marketplace), and whether it holds a bank charter or relies on a bank partner. Identify the primary revenue drivers (spread income vs. fee income vs. gain-on-sale).
Evaluate the credit model — Review underwriting methodology and variable selection. Assess model validation practices: backtesting frequency, out-of-sample performance, champion/challenger testing. Check for fair lending risks — disparate impact testing, use of prohibited or proxy variables. Flag reliance on alternative data sources that lack long track records. [VERIFY applicable fair lending standards: ECOA, state-specific requirements]
Analyze portfolio performance — Examine vintage loss curves against stated projections. Compare actual vs. projected default and prepayment rates. Look for signs of credit deterioration: rising early-stage delinquencies, increasing average balance at default, cohort-over-cohort spread compression. Benchmark loss rates against comparable public ABS deals or peer disclosures.
Assess funding structure and liquidity — Map all capital sources and their terms. Evaluate concentration risk (single warehouse lender dependency). Review covenant headroom — minimum tangible net worth, maximum delinquency triggers, borrowing base eligibility criteria. Model liquidity runway under a scenario where one or more facilities become unavailable. For platforms relying on securitization, assess execution risk and market access.
Review regulatory and compliance posture — Identify all required federal and state licenses. For bank-partner models, evaluate the true lender risk and Madden/valid-when-made exposure. Review for compliance with TILA, ECOA, FCRA, UDAP/UDAAP, and applicable state usury caps. Check for CRA implications if a bank partner is involved. Note any pending litigation, CFPB inquiries, or state AG investigations. [VERIFY state-specific usury limits and licensing requirements for each operating jurisdiction]
Evaluate technology and operations — Assess loan origination system capabilities (automation rate, time-to-fund), servicing platform scalability, and fraud detection effectiveness (identity fraud rate, synthetic fraud controls). Review disaster recovery and business continuity posture. Evaluate API reliability for embedded lending integrations.
Synthesize risk-adjusted assessment — Consolidate findings into a structured view: credit risk rating, funding risk rating, regulatory risk rating, and operational risk rating. Identify the top three risks and top three strengths. Provide a forward-looking outlook under base, upside, and stress scenarios.
Output
Produce an Analysis Report containing:
- Executive summary: Platform overview, key conclusion, and overall risk assessment (1–2 paragraphs)
- Business model profile: Lending vertical, origination channel, charter/licensing structure, geographic scope
- Credit model assessment: Model methodology, validation practices, fair lending posture, alternative data reliance, with a risk rating (Low / Moderate / Elevated / High)
- Portfolio performance analysis: Vintage loss curves, delinquency trends, benchmark comparisons, with commentary on trajectory
- Funding structure analysis: Capital sources, concentration, covenant headroom, liquidity runway, with a risk rating
- Regulatory compliance review: License inventory, true lender analysis (if applicable), pending actions, with a risk rating
- Technology and operations review: Automation metrics, scalability assessment, fraud controls
- Consolidated risk matrix: Summary table of risk ratings across all dimensions
- Key risks and mitigants: Top risks with identified mitigating factors or open items
- Recommendations: Specific next steps, diligence items to pursue, or conditions for engagement
Quality Checks
- Verify all loss-rate data cites a specific vintage, cohort, or reporting period — never use undated aggregate figures
- Confirm funding structure details against actual facility agreements or term sheets, not management presentations alone
- Ensure regulatory license inventory is current and cross-referenced against states where the platform actively originates [VERIFY]
- Check that credit model evaluation addresses fair lending compliance, not just predictive accuracy
- Validate that stress scenarios use plausible macroeconomic assumptions (unemployment spike, interest rate shock, funding market closure)
- Flag any data gaps or areas where management representations could not be independently verified
- Confirm unit economics analysis accounts for full cost stack including customer acquisition, servicing, and credit losses