Great underwriting is a speed game and a risk game. MCA providers and lenders need to decision faster without taking on avoidable losses. In 2025, the winning stack blends live bank transaction data, KYB/KYC identity verification, OFAC screening, and velocity/device intelligence—then ties approvals directly to RTP disbursements and recurring ACH collections. Liftoff Platform delivers all of this in one place and adds AI repayment intelligence that predicts when a customer will pay, flags early hardship, and recommends the best collection action to maximize cure with minimal friction.
Why this underwriting stack matters now
· Speed to decision = win rate: Approve good files same day (often same hour) with fewer back-and-froths.
· Risk compression: Catch identity mismatches, sanctioned parties, and synthetic behavior before a dollar moves.
· Clean handoff to funding & collections: Approve → RTP disbursements → ACH schedules with mandates captured and accounts validated.
· Lower total cost of ownership (all-in): Fewer bad approvals, fewer returns, less manual review, fewer chargebacks on card rails.
The modern underwriting flow (what good looks like)
1. Intake & pre-qual
◦ Business details, intended use, revenue signals, bank-link prompt.
◦ Soft checks to filter obvious mismatches (industry rules, geography, min revenue).
2. Identity & entity verification (KYB/KYC)
◦ Verify legal entity, beneficial owners, and signers.
◦ Crosscheck addresses, EIN/SSN, and corporate status; capture ESIGN consent.
3. Sanctions & watchlists (OFAC, PEP, adverse media)
◦ Screen business, owners, and control persons; store match logs and clearing outcomes.
4. Bank data pull & analysis
◦ 6–12 months of transactions, daily balances, NSFs/returns, deposits mix (card/ ACH/cash), seasonality.
◦ Derived metrics: average daily balance, volatility, MTD trend, return rate, effective DTI for small-biz owners.
5. Velocity, device, and behavioral signals
◦ New device risk, IP mismatch, application editing patterns, repeated declines across identities.
◦ Cohort-level velocity: too many applications from same device or address within 24–48h.
6. Score & policy rules
◦ Rules + model: thresholds for approval/decline/review; industry overlays (e.g., high-risk MCCs).
◦ Explainability: keep reason codes for adverse action letters.
7. Decision → Funding → Collections
◦ RTP disbursement with confirmation.
◦ Capture ACH mandates and set recurring schedules (with account validation) at approval.
8. Ongoing monitoring (post-fund)
◦ Return codes, payment performance, bank data refresh triggers, and velocity anomalies.
The underwriting toolkit (components you actually use)
· KYB/KYC
◦ Entity verification, beneficial ownership (BO), ID doc + liveness, and watchlist checks.
◦ Store artifacts with timestamps; re-verify on material changes.
· OFAC & watchlists
◦ Real-time screening with audit logs; configurable risk scoring for fuzzy matches.
◦ Automated, documented clearance process (who cleared, why, when).
· Bank transaction data
◦ Ingest via secure bank-link; parse deposits/withdrawals, identify recurring obligations, NSFs, payroll cycles.
◦ Flags: thin cash buffers, seasonal cliffs, high-return months, stacked obligations.
· Velocity & device intelligence
◦ Device fingerprint, IP geolocation, proxy/TOR usage, form-fill speed; cross applicant link analysis.
◦ Thresholds that push to manual review or require extra proofs.
· Policy engine & models
◦ Rule-based filters + scorecards; industry and ticket-size verticals; minimum months-in-business and revenue.
◦ Versioned policies with A/B policy experiments (e.g., “plus 10% approval on these segments”).
· Mandates & authorizations
◦ SEC-appropriate ACH authorization (PPD/WEB/TEL/CCD/CTX) with ESIGN consent and logs.
◦ Store retrievable artifacts for audits and disputes.
· Collections readiness
◦ Account validation (instant + micro-deposits), pre-due reminders, intelligent retries timed to deposits, and return-code rules.
Operating playbook (simple rules that work)
· Bank data first: Decline or pause files with <60 days of usable history or excessive NSFs.
· Owner identity tie-out: KYC the signer/guarantor; match to KYB records and bank owner signals.
· Sanctions are a hard stop: Resolve any OFAC/PEP hits before proceeding; log the decision trail.
· Velocity triggers review: New device + high-edit application + multiple attempts in 24h → manual.
· Approve → disburse via RTP: Speed reduces abandonment and support friction.
· Set ACH schedules at approval: Capture mandates right away; validate accounts before first pull.
Risk controls that reduce losses (and busywork)
· Account validation catches R02/R03/R04 before day one.
· Pre-due reminders (T-72h/T-24h) drive on-time payments and fewer NSFs.
· Return-code automation:
◦ Retry-eligible: R01/R09 once, timed to payroll/posting windows.
◦ Fix first: R07/R08/R10/R29 (authorization), R03/R04 (data quality).
· Self-serve changes: Let borrowers update methods or reschedule within policy limits— fewer tickets, better cure rates.
Total Cost of Ownership (TCO): what improves
· Total Cost of Ownership spans more than approval rates. With this toolkit you reduce:
· Acquisition & ops cost: Fewer manual reviews, faster decisions, fewer support contacts.
· Losses & write-offs: Identity/sanctions mismatches and risky cash-flow patterns filtered pre-fund.
· Return & exception costs: Validation + timing cuts NSF loops and reattempt spend.
· Timing costs: RTP accelerates funding; ACH schedules stabilize inflows for planning.
· Compliance cost: Stored mandates, OFAC logs, KYC/KYB artifacts, and immutable events simplify audits.
Why Liftoff Platform is best-in-class for underwriting + payments
· Underwriting data layer: KYB/KYC, OFAC, device/velocity, and bank-link integrations with audit-ready storage.
· Decision → money movement: RTP disbursements (24/7/365) and $0 ACH pulls for recurring collections; $0 debit for payer preference on invoices.
· Invoice automation: One-click e-payments, recurring schedules, reminders, resend, and activity tracking.
· Collections intelligence: Account validation, return-code rules, and intelligent retries tuned to deposit windows.
· Compliance-by-default: ESIGN, mandates, OFAC logs, immutable events, and role based access.
· Portal + API: Launch fast without engineering; automate deeply with webhooks and ledgers when ready.
· Analytics: Approval rates, time-to-fund, return codes, cure rates, blended $/1k moved, and cohort behavior.
AI-Powered Repayment Intelligence (unique to Liftoff)
· Signals we learn from: bank transaction patterns (pay cycles, cash buffers, NSFs), historical return codes, device/velocity risk, communication engagement (email/SMS opens, clicks), and prior promise-to-pay behavior.
· Predictions we generate:
◦ Propensity-to-pay this cycle and best day/time to debit
◦ Probability of NSF (R01/R09) and risk of unauthorized (R07/R08/R10/R29)
◦ Early-hardship likelihood and expected cure time if a payment is missed
· Actions we automate:
◦ Schedule optimization (move due date to predicted cash-in window)
◦ Intelligent retries (one, deposit-window aligned)
◦ Right-channel outreach (email/SMS) with adaptive tone and offers
◦ Hardship flows (partial/deferral plans) when cure probability is higher than brute-force retries
◦ RTP micro-credits to resolve disputes or service interruptions without derailing the main ACH plan
Implementation checklist (one-week pilot)
· Turn on KYB/KYC + OFAC and store artifacts.
· Add bank-link to intake; require 6–12 months of transactions.
· Configure velocity/device thresholds and manual-review queues.
· Define policy rules (approve/decline/review) with reason codes.
· Wire decision to RTP disbursements; capture ACH mandates and validate accounts at approval.
· Enable reminders, intelligent retries, and return-code automation.
· Connect webhooks to your CRM/LOS and GL/ERP; monitor dashboards weekly.
Example targets (first 60–90 days)
· Time-to-decision: same-day approvals for ≥70% of clean files
· Bad-approval rate: −20–30% via bank-data/velocity/OFAC filters
· NSF/return rate: −20–35% with validation + timing
· Ops minutes per file: −30–50% from fewer manual touchpoints
· Time-to-fund: minutes with RTP; fewer abandoned approvals
Summary
A modern MCA/lender stack uses bank transaction data to size real cash flow, KYB/KYC to validate who you’re dealing with, OFAC to stay compliant, and velocity/device to catch synthetic behavior—all tied to instant RTP disbursements and predictable ACH collections. The result is faster approvals, fewer losses, and a lower Total Cost of Ownership across underwriting and payments. Liftoff Platform brings the data, rails, risk controls, and automation together so you can deploy quickly and scale with confidence.
FAQs
Do I really need both KYB and KYC?
Yes—KYB verifies the business; KYC verifies the owners/signers. You need both to mitigate impersonation and synthetic risk.
How much bank data is enough?
Aim for 6–12 months. Shorter histories increase variance and hide seasonality/stacked obligations.
What if a borrower fails OFAC?
Pause immediately, clear the match per policy, and retain logs. No funding until resolved.
Can I still fund instantly?
Yes—once cleared, use RTP to disburse in seconds and capture ACH mandates in the same flow.
How do I reduce NSF cycles post-fund?
Validate accounts at onboarding, send pre-due reminders, and retry once for R01/R09 at deposit windows; don’t auto-retry unauthorized/stop codes.