Kind Lending HMDA 2025: 3,991 Investor Loans, $1.33 Billion, and Zero-Income Loans Up From 498 to 1,185

Which number is “the” number: 3,991, 4,110, 4,088 or 3,924?

Searches for “Kind Lending DSCR” and “Kind Lending investor loans” meet several counts, because HMDA offers more than one way to say “investor loan”. We recounted from the raw FFIEC file for the company's Legal Entity Identifier (the API returns only originated loans, action_taken = 1: 29,401 in 2025 and 18,722 in 2024) and stated each filter.

The 3,991 is what the raw file gives. An earlier extract we had seen gave 4,088 business-purpose loans out of 18,492 originations; that does not reproduce from the file the FFIEC serves today: the current file has 29,401 originated loans, 4,110 of them coded business purpose, and none of the filters we tried gives 4,088 or 18,492. We use only what we could recount, and the recount script and its output are in sources/. We apply no filter on state, term, purpose, lien or units. HMDA does not code a loan as DSCR, so the file cannot tell you which of the 3,991 were underwritten on the property's rent.

What Kind Lending says it is, and what the file shows about the channel

The company website footer reads “©2026 | Kind Lending, LLC | NMLS #3925” with the address 1920 Main Street, Suite 1200, Irvine, CA 92614 and lists a California licence under the Residential Mortgage Lending Act, Hawaii licence HI-3925, a New Jersey Department of Banking and Insurance licence, an Arizona banker licence and a Massachusetts mortgage banker licence. We did not open NMLS Consumer Access, so we have not verified that register; the company itself points to it for licensing.

On DSCR, the company's own page (dated November 4, 2024) says: “With loan amounts up to $2 million, DSCR Max is perfect for investors looking to scale their portfolios quickly.” It also says: “Credit scores as low as 640 are eligible, and there are no ratio options to worry about.” And: “One of the most significant advantages of the DSCR Max program is that personal income is not required to qualify.” Those are the company's statements, quoted as published; we have not checked them against a rate sheet or guideline, and the page does not give a loan-to-value limit. The file is consistent with the first claim as to size: the largest 2025 investor loan is $1,975,000. Credit scores are not public in HMDA, so the file cannot confirm the second.

The file does show how the loans arrive: 3,895 of the 3,991 loans (97.6%) were “not submitted directly to your institution”, and 96 were submitted directly (3,374 of 3,442, 98.0%, in 2024). In plain terms, almost every investor loan reaches the company through a third party such as a broker. For you as a borrower that means the person you speak to is usually not Kind's own loan officer.

The 2025 file against 2024

Investor loans rose 15.9% and investor dollars 11.9%, but the whole book rose 57.0%, so the investor share of loans fell from 18.4% to 13.6% (our arithmetic). The company grew faster in owner-occupied lending than in investor lending, and the median investor loan became smaller, not larger. The largest 2024 loan ($2,725,000) is above the $2 million the DSCR Max page quotes; the largest 2025 loan is under it.

Rates: same median, wider spread

Shares are our arithmetic on the counts. The most common rates in 2025 were 6.875% (353 loans), 6.625% (344), 7.125% (319), 7.375% (307) and 6.99% (291); in 2024 they were 7.125% (480), 7.625% (457), 6.625% (386), 7.5% (324) and 6.99% (290). The lowest rate in both years is 4.875% and the highest 11.375% in 2025 (11.25% in 2024). The mean rate barely moved, 7.188% against 7.195% (our arithmetic). A one-number summary hides what happened: more loans at the low end and more at the high end, and fewer in the 7.5% to 8% band.

A worked example (our arithmetic). A $275,000 loan, 30 years fully amortizing, has a monthly principal and interest payment of $1,827.74 at 6.99% and $1,922.84 at 7.5%: $95.10 a month for half a point. These are note rates on closed loans, not a quote; the rate a borrower gets depends on credit score, loan-to-value, loan size and program.

By purpose in 2025 the median rate was 7.0% on the 2,403 purchase loans (60.2% of the book, against 2,470 and 71.8% in 2024), 7.25% on 705 cash-out refinancings (323 in 2024), 6.875% on 485 other refinancings (307) and 7.375% on 398 loans coded home improvement (342). The mix moved toward refinancing: cash-out loans more than doubled.

The zero-income group: the DSCR-like half of the book

HMDA asks for the gross annual income “relied on in making the credit decision”. A loan underwritten on the property's rental income, as a DSCR loan is, would normally report 0 or no income. That is our reading, and the file does not say it. The company's DSCR Max page says personal income is not required; we do not know which of these loans were DSCR Max, or any other program.

The zero-income group grew 138% in loans (1,185 against 498, our arithmetic) and from 14.6% to 28.3% of investor dollars ($174,540,000 of $1,192,250,000 against $377,085,000 of $1,334,235,000). Its median rate fell a quarter of a point, 7.75% to 7.5%, but it still prices half a point above the loans that report income, and one in five (260 of 1,185, 21.9%) is at 8% or more against 5.6% (156 of 2,806) for the others (our arithmetic). The pricing gap is the useful fact for a DSCR borrower: the same lender's file prices the no-income-relied-on loans higher, and puts them in different hands. None was bought by Fannie Mae or Freddie Mac (0 of 1,185); they went to private securitizers, banks or other buyers, or were not sold in the year.

For a Texas or Florida borrower this matters more than the national median. Texas had 611 investor loans in 2025 and 201 of them (32.9%) report income 0; Florida had 402 and 175 (43.5%); California 773 and 166 (21.5%) (our arithmetic). Texas grew from 451 to 611 loans, a rise of 35.5%.

Who bought the loans

Fannie Mae and Freddie Mac together bought 1,555 of the 3,991 loans (39.0%) in 2025 and 1,947 of 3,442 (56.6%) in 2024 (our arithmetic). The loans that stay on a lender's books for a time, code 0, went from 14 to 392. Private securitizer purchases nearly quadrupled, from 167 to 608, which fits the growth of non-agency loans (our reading). A purchaser code is where the loan went by year-end, not a statement about the program's guidelines.

Fixed-period adjustable features are rare: 9 loans in 2025 have an introductory rate period, none in 2024. 3,931 loans are first liens and 60 are second liens (1 in 2024). By property, 3,312 are one unit (83.0%), 426 two units, 105 three units and 148 four units.

Where the loans are

California fell from 23.7% of loans (817 of 3,442) to 19.4% (773 of 3,991) and Texas rose from 13.1% to 15.3% (our arithmetic). Washington (132 loans in 2024, 99 in 2025) and Virginia (110 and 93) fell. Median loans differ by state: $395,000 in California, $345,000 in Utah, $415,000 in New Jersey, $235,000 in Texas and Georgia, and $225,000 in Pennsylvania. For rates and lenders across a state see our DSCR loan Texas page, DSCR loan Florida page and DSCR loan California page.

Loan-to-value and term

Median loan-to-value fell to 73.65% from 75.0%; the middle half is 60.7% to 75.0% in 2025 against 63.78% to 75.0% in 2024. Loans at exactly 75% were 1,170 (29.3%) against 1,242 (36.1%). The bands moved up at the top: 663 loans above 75% up to 80% against 439, and 65 above 80% against 58. The company page does not state an LTV maximum, and loan-to-value in HMDA is the lender's reported ratio, not a program limit.

Terms are almost all 30 years: 3,909 loans at 360 months, 44 at 480 months, 23 at 180, 13 at 240, and one each at 336 and 120. Interest-only loans were 219 (5.5%) against 66 (1.9%); 197 of the 219 are in the zero-income group.

What the file cannot tell you

• Prepayment penalty. The field prepayment_penalty_term reads NA on all 3,991 loans (and all 3,442 in 2024). The FFIEC data-field list describes it as “The term, in months, of any prepayment penalty”; it does not define NA. Do not read “no penalty” into it. Ask for the clause in writing.
• Cost. Total loan costs, origination charges, discount points and total points and fees are NA on every loan.
• DSCR itself. No field records the coverage ratio, the rent, the program or whether the loan is a DSCR loan.
• Credit score. Not public.
• Current pricing. These are closed loans from 2025, not a rate sheet.

Complaints and records

On November 6, 2026 we queried the CFPB Consumer Complaint Database (index updated November 6, 2026) for “Kind Lending” in the company field and across all fields; both returned 0 results. A control search for “Kiavi” in the company field returned 25 the same day. Zero complaints means none are published under that name; it is not evidence of how borrowers were treated, and a lender that works through brokers draws complaints to the broker or to a state regulator. We did not search state regulators or court dockets, and we did not open NMLS Consumer Access (ID 3925, per the company website), so absence of results is not a clean record. We used no consumer-review sites.

What a reader can do with this

1. Find your loan's row. If your loan is underwritten on the property's rent, compare it with the zero-income group: median rate 7.5%, middle half 7.0% to 7.875%, median loan $255,000. A quote at 8% or more would have been in the top 21.9% of that group (260 of 1,185) and the top 10.4% of all 3,991 loans (our arithmetic).
2. Ask the broker three things in writing: the prepayment term in months and dollars, the interest-only period if any, and whether the loan is meant to be sold or held. The file shows the buyer type, not the contract.
3. Do not use the 2025 median as a quote. 7.125% is the median in both years, but 406 loans were under 6.5% and 416 were at 8% or more. The spread, not the median, is what you will meet.
4. Check the licence for your state. Use NMLS ID 3925 on NMLS Consumer Access and your state regulator, and confirm the legal name on your documents.
5. Compare other lenders on the same filter. Our conventional loan for investment property page shows the agency side of this market and where Kind Lending's 3,924 conventional investor loans sit; our DSCR lenders ranked by loans page ranks lenders by 2025 volume, and our New Wave Lending Group page applies the same method to a wholesale lender. For how lenders price and underwrite the loan type, see DSCR loan rates and DSCR loan requirements.
6. If you want a lender that markets directly to investors, the box below lists purchase, rehab and rental lenders. Kiavi pays us a referral fee when a loan closes through the button. Kind Lending is not an affiliate, and the counts above are the same either way.

FAQ

Related reads: conventional loan for investment property, DSCR lenders ranked by loans, DSCR loan Texas, DSCR loan vs conventional mortgage and New Wave Lending Group HMDA 2025.

Sources: loan-level data from the FFIEC HMDA Data Browser, 2025 and 2024 originated-loan files for LEI 549300MZ8VZJOVC63092, downloaded November 6, 2026 through the public API; the FFIEC public loan application register data-field list (codes for occupancy, purpose, purchaser type, business purpose, submission and prepayment term); the Kind Lending website kindlending.com (home page and DSCR Max page dated November 4, 2024, read November 6, 2026, quoted as the company's own statements); the CFPB Consumer Complaint Database API, queried November 6, 2026. The recount script and its output are in sources/. All percentages, differences and “our arithmetic” figures are calculated by us. We did not access NMLS Consumer Access. This is analysis of public documents, not investment, legal or tax advice. Compiled by CrowdfundedWealth on November 6, 2026.