How We Evaluate Investor-Property Lenders (and Why We Ignore Commissions)

Most lender rankings on the internet are compensation tables wearing editorial clothing. The sites are paid per funded referral, the payouts differ by lender, and the ordering follows the payouts — invisibly, because nothing on the page says so. We run an affiliate model too, and referral relationships help pay for this site. The difference is structural: compensation data is walled off from the evaluation process, and lender ordering is built only from stated terms and verifiable facts. This page documents the method in enough detail that you can audit it — what we collect, how ordering actually works, what we can verify, what we cannot, and where the money flows. It is a pillar page rather than fine print because on a lending site, the methodology is the product.

The conflict this methodology exists to control

The economics of lending content are straightforward. Lenders pay for borrower referrals; different lenders pay differently; and a site that orders its recommendations by payout earns more than one that does not. The reader cannot detect the difference, because a payout-ordered list and a terms-ordered list look identical on the page. That is the conflict, stated plainly, and it applies to us as much as to anyone.

Our control is payout-blindness. The evaluation and ordering of lenders uses no compensation input: not the payout amount, not the existence of a relationship, not whether a lender has ever paid us anything. Rankings are assembled from the criteria below, and the compensation side of the business is reconciled separately, after the fact. When a lender we have no relationship with fits a comparison, it is listed and ordered by the same rules; the absence of a referral link changes our revenue, not the lender's position.

Why not rely on disclosure alone? Because disclosure without separation merely announces a bias; it does not remove one. A page that says "we may be compensated" while quietly ordering by compensation has disclosed its way into exactly the same result. Separation is the control; disclosure is the receipt. We maintain both because either one alone is weaker than it looks.

Two scope notes. First, everything we evaluate is business-purpose, investor-property lending — DSCR, bridge and hard money, rehab, construction, and portfolio products. We do not evaluate consumer mortgages, and a reader who intends to occupy the property being financed is outside our scope entirely and should be reading consumer-focused resources with consumer-protection expertise. Second, nothing here is individualized advice; it is a research method, documented.

What we collect for every lender

The evaluation file for a lender is built from materials the lender publishes or provides in writing, captured with dates:

• Product boxes: minimum and maximum loan amounts, leverage caps, coverage-ratio floors, credit floors, eligible property types, and eligible borrower structures for each product.
• Stated pricing floors: published rate floors and points ranges, as published — never as extrapolated by us, and never from a phone quote we cannot document.
• Structural terms: prepayment-penalty menus, term and amortization options, interest-accrual method on holdback products, draw fees and mechanics, extension terms where published.
• Footprint: states where the lender actually lends, and any state carve-outs inside product lines.
• Transparency artifacts: whether the lender publishes its pricing grid or adjustment structure at all, whether term sheets are specimen-available, and how current its published materials are.

Captures are dated because the date is part of the fact. A leverage cap without a capture date is a rumor with formatting; the same cap, dated, is a checkable claim — against the lender's current sheet, by anyone, at any time. Where a lender provides materials to us directly, we hold them to the same standard: written, dated, and quotable, or not used.

What we deliberately do not collect into the evaluation file: referral payout schedules, conversion data, testimonials, award badges, and the lender's own superlatives. The first two are walled off in the business ledger; the rest are marketing, with no evidentiary weight.

How the ordering actually works

Ordering runs in a fixed sequence: fit filters first, stated economics second, structure third, transparency as the tiebreaker.

Ties that survive the sequence are broken alphabetically. When a comparison depends on a term a lender does not publish, we say "not published" rather than guessing, and an unpublished term cannot earn ordering credit — which is deliberate: it makes opacity expensive in our rankings instead of convenient.

What we can verify — and what we cannot

Honesty about the limits of the method matters more than the method.

We can verify, at a point in time: what a lender publishes about its own products — boxes, floors, fee schedules, state lists — because we capture and date those materials. We can verify corporate existence and, where applicable, licensing: business-purpose lending is regulated state by state, licensing requirements vary — some states require lender licensing for business-purpose loans, many do not — and where a license is claimed, registries such as NMLS Consumer Access allow a public check.

We cannot verify: the rate you will actually be quoted, because actual pricing is file-specific — leverage, credit, coverage ratio, property type, and prepay election all move it, as explained in /blog/dscr-loans-complete-guide. We cannot verify service quality at scale — how fast a lender really closes, how its draw administration behaves under stress — because we cannot audit other people's loans; where we note operational reputation at all, we frame it as pattern-level and unverified. And we cannot verify that published materials will not change the day after capture, which is why everything we cite carries a date.

Anything on this site that reads like a number is one of three things: a definitional fact (how a ratio is computed), a published figure with a date, or a clearly hedged market-typical range. If you find a fourth kind — a specific claim with no source and no hedge — that is an error; the corrections process below exists for it.

Why we never quote you a rate

Rate expectations are where lending content most reliably misleads, so our rule is absolute: no rate promises, anywhere, ever.

The structural reason: business-purpose pricing is built from a base rate plus a grid of adjustments — leverage tier, credit tier, coverage band, prepay election, property type, loan size — so a lender's "as low as" figure describes the best corner of the grid, not the middle and not your file. Base pricing also moves with the broader market, daily, and a captured floor is stale the moment conditions move. Short-term products add points, draw fees, extension fees, and accrual-method differences that headline rates exclude entirely — the honest comparison there is all-in cost over your realistic hold, a framework laid out in /blog/hard-money-vs-bridge-loans.

Lock policy is the final reason. A quote is not a term until it is locked, lock windows and lock fees differ by lender, and a rate that cannot be locked for your realistic closing timeline is trivia. Any comparison built on indications that cannot be locked inherits their expiration dates.

So our comparisons present floors as floors, ranges as ranges, and always dated. Where arithmetic helps — testing a property's coverage ratio against a payment — we give you the tool rather than the prediction: /tools/dscr-calculator runs the calculation on your numbers, under stated conventions, with no forecast attached.

Disclosure, corrections, and update cadence

Disclosure: pages containing referral links say so, on the page. We may earn a fee when a reader connects with a lender through this site. Compensation never affects ordering, inclusion, or exclusion; the wall described above is the mechanism, and this page is the standing disclosure of how it works.

Updates: lender-published terms are re-checked on a scheduled cadence and whenever a reader or lender reports a change. Every lender page and comparison carries a last-reviewed date. Stale captures are a known failure mode of this model; the date is your defense, and ours.

Corrections: when we get a fact wrong — a misread floor, an outdated box, a state we listed incorrectly — we correct the page and note the correction. Readers who find errors are the best audit mechanism a site like this has; reports are acted on, not filed.

What we ask of lenders is what we ask of ourselves: publish the terms, date the changes, and correct the record when it is wrong. Lenders that operate that way tend to fare well under this method, which is not a coincidence. The method is built to reward being checkable.

Common mistakes when reading lender reviews — including ours

• Treating any ranked list as an answer rather than a filter. A ranking narrows the field; your term sheets, on your file, decide.
• Reading a published floor as an expected rate. The floor is the grid's best corner; your cell depends on your file.
• Shopping the rate while ignoring structure — prepayment penalties, accrual method, draw and extension fees — which commonly decide total cost.
• Weighting testimonials. Anecdotes select for extremes and verify nothing.
• Comparing one lender's locked quote against another's floating indication — a timing mismatch dressed up as a price difference.
• Assuming a reviewer's incentives are aligned with yours. Ours are documented here; demand the same documentation elsewhere, and be suspicious where it is missing.
• Forgetting that reviews age. A comparison without dates is a snapshot of an unknown moment.

How to verify

Verify us, then verify the lender — in that order.

• Check the dates on any comparison you use here, and treat anything stale as a lead, not a fact.
• Spot-check our captures against the lender's current published materials; if they diverge, the lender's current sheet is the truth and we owe a correction.
• Where a lender claims licensing, check the public registry (NMLS Consumer Access) and the state regulator's records.
• Ask any lender for its full pricing grid and a specimen term sheet for your scenario — the two documents that convert marketing into terms. A lender that will not show its grid is asking you to price-shop blind.
• Get every deciding term in writing and locate it in the governing documents — note, loan agreement, riders, guaranty — before closing. No review, including ours, is a term of your loan.
• If you intend to occupy the property, stop here entirely; you are in consumer-mortgage territory, and this site's method does not cover it.

The documents govern. Everything else — this site included — is research.