The 100-point rubric is implemented and versioned. Rankings remain unpublished until current evidence, normalized pricing, editorial approval, and compliance approval all pass.
Thresholds implemented. Publication approval pending.
The same thresholds apply to every lender. Page placement and advertised starting rates contribute zero points.
DSCR floors, no-ratio paths, credit, experience, and residency.
Purchase/refinance leverage, loan range, reserves, and seasoning.
Property types, STRs, entities, transaction breadth, and structures.
Normalized quotes, lender fees, and prepayment structures.
Sourced closing expectations, locks, and process clarity.
Verified state coverage and licensing identity.
Field representation, current evidence, and source quality.
Missing, stale, varying, or conflicting facts are excluded rather than converted into favorable assumptions. A lender cannot rank until enough of the model is supported by current evidence.
| Category | Weight | Scored criteria |
|---|---|---|
| Qualification flexibility | 25 points | Minimum DSCR (7) · No-ratio path (5) · Minimum FICO (6) · First-time investor (3) · Foreign-national path (4) |
| Leverage and loan terms | 20 points | Purchase LTV (5) · Rate-term LTV (4) · Cash-out LTV (4) · Loan range (4) · Reserves and seasoning (3) |
| Property and borrower eligibility | 15 points | Property breadth (5) · Short-term rentals (3) · Entity vesting (2) · Loan structures (3) · Transaction breadth (2) |
| Pricing and fees | 15 points | Normalized benchmark quote (7) · Published lender fees (4) · Prepayment flexibility (4) |
| Closing experience | 10 points | Closing time (6) · Rate lock (2) · Process transparency (2) |
| Geographic availability | 10 points | Verified state availability (7) · Licensing identity (3) |
| Transparency and documentation | 5 points | Field representation (2) · Current rankable evidence (2) · Primary-source depth (1) |
The strongest available first-party record governs each field.
Third-party roundups and unsourced sales claims do not establish program facts. When official sources disagree, the field is marked conflicting and removed from scoring until resolved.
Headline programs, broker channels, no-ratio options, short-term rentals, and foreign-national programs may have different thresholds. They are stored as distinct variants rather than compressed into one misleading number.
The scoring system fails closed.
Until every gate passes, category scores, badges, overall scores, and “Best Overall” language remain hidden. Objective fields can still be filtered and sorted.
A marketing rate is not a comparable quote.
Property, purpose, LTV, credit band, DSCR, loan amount, payment structure, and borrower profile remain identical.
Record rate, APR, points, lender fees, third-party fees, reserves, prepayment, lock period, and estimated cash to close.
Quotes must be collected inside the same market window. Published starting rates never determine the pricing winner.
Confirms that thresholds were frozen before results and every explanation matches its evidence.
Confirms that no commercial consideration altered facts, thresholds, ordering, or category awards.
Material corrections require an official source, receive a new verification date, and trigger score recalculation.
Each distinction asks one factual question of one published field and names whoever leads it. It is not a ranking, not a score, and not an editorial opinion, and it does not use the weighted rubric described above.
The order lenders appear in across this site follows the DSCR Accessibility Score, model accessibility-1.0, published 2026-08-12. Its thesis, fixed before any score was computed: Weighted toward whether a borrower can qualify and how much they can borrow, rather than how quickly a file closes. The machine-readable result is at /ranking.json.
minimumDscr, noRatio, minimumFico, firstTimeInvestor, foreignNational
purchaseLtv, cashOutLtv, rateTermLtv
maximumLoan, minimumLoan
programs, propertyTypes, shortTermRental, llc
states
closingDays, processTransparency
