Personal Loan Decisioning

Cash flow underwriting for the bureau-invisible.

AltFiScore scores personal loan applicants using actual checking account behavior + verified employment — not just bureau history. Approve qualified thin-file borrowers your competitors auto-decline.

Get sandbox access See score methodology

POST /v1/decisions/personal-loan

298ms

Outcome

Approved

Score

762

Grade

Approved amount

11.0% APR · 48mo

$15,000

$388/mo · DTI 28%

Why approved despite thin bureau

Bureau file: 6 months, 1 trade line, 580 FICO

Plaid: 2yr savings discipline, $0 overdrafts

Argyle: $72K verified, 3yr employer

✓ Reg Z disclosure · ✓ Reg B adverse action ready · ✓ State usury verified

The problem

Thin-file borrowers are 70M Americans your underwriting can't see.

Recent immigrants. Younger adults building credit. Gig workers. Subprime customers who paid down debt and stayed clean. Bureaus call them "unscoreable" because they lack 3+ trade lines — but their bank accounts tell a richer story bureaus can't access.

26%

Of personal loan applicants get declined for thin-file reasons despite having stable employment and healthy savings.

Modern alt data decisioning

Bank statements + verified income, scored like a 24-month bureau file.

AltFiScore reads 90 days of Plaid transactions to build a true cash flow profile — net flow, overdraft pattern, savings discipline, recurring debt. Combined with Argyle income verification, this synthetic credit profile outperforms FICO for borrowers under 25 and recent immigrants.

Plaid

90-day cash flow analysis

Net positive cash flow over 90 days predicts personal loan repayment better than a thin bureau file. AltFiScore catches the 26% of borrowers who genuinely have repayment capacity.

Plaid

Savings discipline

Consistent savings transfers — even small amounts — strongly predict on-time repayment. Bureau files can't see this signal at all.

Plaid

Overdraft frequency

Zero overdrafts in 90 days is a stronger positive signal than mid-tier FICO. Heavy overdrafts indicate cash flow stress no bureau report shows.

Argyle

Income consistency

Recurring deposits from the same employer over 12+ months signal repayment capacity. Outperforms self-reported income on a paper application.

AltFiScore

DTI from actual debts

DTI calculated from real recurring debits — credit cards, subscriptions, auto loans, BNPL plans — not bureau-reported minimums.

AltFiScore

Stability composite

Address tenure + employment tenure + bank account age combined into a single signal. For thin-file borrowers, this signal is more predictive than FICO.

Decisioning details

Rate bands and policy thresholds

Configurable rate bands per tenant. Federal compliance caps applied automatically. All policy decisions are versioned, reproducible, and audit-logged.

Grade	Score range	APR
A+	880–1000	6.0%
A	820–879	8.5%
B	740–819	11.0%
C	660–739	15.5%
D	300–659	19.0%

Federal compliance caps

MLA 36% APR cap for active-duty military
State-specific usury ceilings applied automatically
Reg Z (TILA) disclosures attached
ECOA / Reg B compliant adverse action codes

Integration

One endpoint. JSON in, JSON out.

Idempotent. Versioned. TILA disclosures and FCRA codes built in. Sandbox tenant for testing, production tenant when you're ready.

1Sign up for a sandbox tenant
2POST to /v1/decisions/personal-loan with consumer + loan details
3If consumer needs to link bank/employer, response includes consumer_complete_url
4Poll polling_url or receive webhook when consumer completes
5Receive decision with full Reg Z disclosure + reason codes

POST /v1/decisions/personal-loan

curl -X POST 'https://api.altfiscore.com/v1/decisions/personal-loan' \
  -H 'Authorization: Bearer altfi_live_YOUR_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
    "consumer": { /* PII */ },
    "loan": {
      "requested_amount": 15000,
      "term_months": 48,
      "purpose": "debt_consolidation"
    }
  }'

← Response · 200 OK

{
  "decision_id": "9c2b...",
  "status": "complete",
  "outcome": "approved",
  "score_at_decision": 762,
  "risk_grade": "B",
  "confidence_at_decision": 0.91,
  "recommended_amount": "15000.00",
  "recommended_apr": "11.0",
  "recommended_tenure_months": 48,
  "estimated_monthly_installment": "387.95",
  "reason_codes": [],
  "dti_pct": 28.0,
  "evaluation_details": {
    "score_drivers": {
      "plaid_cash_flow": "+85",
      "argyle_employment_stability": "+62",
      "thin_bureau_file": "-12"
    }
  }
}

Compliance

Built for compliance teams.

Every decision is versioned, reproducible, and audit-logged. Adverse action codes follow ECOA Regulation B. Disclosures meet TILA Reg Z.

Reg Z (TILA)

Truth in Lending disclosures generated for every approval — amount financed, finance charge, total of payments, APR.

Reg B (ECOA)

FCRA-compliant adverse action codes returned in every declined or referred decision.

MLA

Active-duty military APR cap (36%) detected and applied automatically.

State usury

Per-state APR ceilings (e.g., 25% MA, 36% CA payday) applied based on consumer residence.

FCRA

AltFiScore reports are not consumer credit reports under FCRA. Adverse action codes returned for downstream FCRA-compliant notice generation.

CFPB UDAAP

Pricing and disclosure language reviewed against Unfair, Deceptive, or Abusive Acts/Practices standards.

Score the personal loan applicants your competitors decline.

Test thin-file approvals using sandbox consumers with realistic Plaid + Argyle profiles. See exactly which signals turn declines into approvals.

Get started free Talk to sales