Buried in the public remarks of a four bedroom colonial in Methuen, listed at $724,400, is a sentence in all capitals: “ASK ABOUT 2.75% ASSUMABLE MORTGAGE, QUICK CLOSING POSSIBLE!!” That agent is doing something almost nobody else in this state is doing. I pulled our MLS PIN feed on September 30. Across 39,094 live Massachusetts listings, 42 say anything at all about an assumable loan. That is about one in a thousand.
Meanwhile the 30 year fixed just printed a fresh high. Freddie Mac’s weekly survey came in at 7.03% on September 24, the first 7 handle in twenty months. By September 29 the daily trackers had run past it, with Mortgage News Daily at 7.58%, its highest since November 2023, and Fortune’s index at 7.455% on September 30. Half the mortgages in the country are still sitting at 4% or less.
So here is my position. Assuming a seller’s low rate loan is a real tool, it is badly underused here, and the single belief doing the most damage is the idea that only a veteran can take over a VA loan. That is false, and it is false in the statute. But I am not going to sell you the version the assumption marketplaces sell. Once you price the equity gap honestly, this works beautifully on some houses and loses money on others, and the arithmetic that separates the two is not complicated. Let me show you both halves.
A buyer does not need to be a veteran to assume a VA loan
This is the myth, and it costs Massachusetts buyers real money every month it survives.
The governing statute is 38 U.S.C. 3714. For loans committed on or after March 1, 1988, it says the assumption must be approved when the loan is current and the purchaser is contractually obligated to buy the property, assume full liability for the balance, and “qualifies from a credit standpoint, to the same extent as if the purchaser were a veteran eligible under section 3710.” Read that clause again. The standard is that you qualify as if you were an eligible veteran. It is a credit test, not a service test. Nowhere does the statute require the buyer to have served.
The VA’s own servicing guidance says the same thing. Circular 26-23-10 lists three conditions for approval, and creditworthiness under VA underwriting standards is the only one touching the buyer’s qualifications.
There is a real catch, and it belongs to the seller, not the buyer. An assumption can close with or without a Substitution of Entitlement. If the buyer is not VA eligible, the veteran seller’s entitlement stays tied to that property until the loan is paid off, which limits their ability to use a VA loan on the next house. If the buyer is VA eligible and substitutes their own entitlement, the seller’s is restored. So the buyer’s status does not decide whether the deal can happen. It decides what it costs the seller. Any seller with a VA loan needs that explained before they agree to anything.
The costs are small and they are capped. The VA circular caps the servicer’s assumption processing fee at $300 where the holder has automatic authority, or $250 where VA prior approval is needed, and that fee is meant to cover all underwriting, processing and closing of the assumption. The funding fee is 0.5% of the loan balance, it must be collected at closing, and it may not be financed into the balance. Compare that to originating a new loan.
The rate gap is real, and it is smaller than the headline
Most writing on this subject compares a 3% assumed loan to the 7.5% headline rate. That comparison is wrong, and it inflates the case by about two thirds of a point.
A buyer who can qualify to assume an FHA loan can also just go get a new FHA loan, and government backed money is cheaper than conventional money right now. On September 30, Fortune’s daily survey had the 30 year FHA rate at 6.814% and the 30 year VA rate at 6.947%, against 7.455% for conventional and 7.603% for jumbo. Bankrate’s Tuesday averages had FHA near 7.02% and VA near 7.04%, so call the honest benchmark somewhere between 6.8% and 7.0%.
That is the number the assumption has to beat. Not 7.58%. The spread on a 3.00% assumed FHA loan is roughly 3.8 points, not 4.6, and every dollar of that difference matters once you finance the gap.
MLS PIN has an assumable checkbox. Nineteen listings use it.
I want to correct something I have heard repeated by good agents, including a version of it I believed myself: that MLS PIN has no way to flag an assumable loan. It does. The Listing Terms field carries an “Assumable” value, and it has for years. The problem is worse than a missing field, because a missing field is an honest excuse.
Here is the September 30 count from our feed. Of 39,094 live listings, 34,353 leave Listing Terms completely blank. That is 87.9% of the market declining to fill in the box at all. Nineteen listings carry the Assumable value. Five of those nineteen are a single to be built subdivision in Attleboro whose remarks never mention a loan, which reads like a data entry habit rather than a financing disclosure.
Now the part that actually explains why buyers never find these. Of the 19 listings that set the flag, exactly one also describes the assumption in its public remarks. Twenty four other listings describe an assumable loan in the remarks without setting the flag. The two groups overlap by a single listing.
| Live listings in the feed | 39,094 |
| Listing Terms left entirely blank | 34,353 (87.9%) |
| Flagged “Assumable” in Listing Terms | 19 |
| Describe an assumption in public remarks | 24 |
| Do both | 1 |
That is the actual failure. It is not that the tool is missing. It is that the flag is machine readable but unexplained, and the explanation is human readable but unindexed, and a buyer’s agent running a normal search catches neither. If you are a listing agent with a low rate assumable loan on your hands, do both. Set the field and write the rate and the balance into the remarks. The Methuen listing that opens this piece did the second one and it is the reason I found it.
Where the assumable loans actually are
The intuition I hear most often is that the VA loans cluster in the western suburbs near Hanscom Air Force Base, so Bedford, Concord, Lexington and Sudbury. I went and checked that against our closed records, because it is the kind of claim that sounds obviously right.
It is backwards. I pulled every Massachusetts residential closing from 2020 through 2022 where the financing type was disclosed, which is the origination window that produced today’s sub 4% loans. Across those four Hanscom area towns, over three full years, there were 16 VA financed closings in total. Bedford had 4, Concord 4, Lexington 1, Sudbury 7. Haverhill alone had 55. Lowell had 42, Lynn 41, Methuen 37, and Boston 195.
The reason is not mysterious once you see it. FHA and VA are the tools that get used where prices sit inside the program limits and buyers are stretching on down payment. That is the Gateway Cities, not the towns with the highest medians in the state.
Two program facts frame the map. FHA’s 2026 one unit limit is $962,550 in Suffolk, Middlesex, Norfolk, Essex and Plymouth, against a national floor of $541,287 and a ceiling of $1,249,125 per HUD Mortgagee Letter 2025-23. Worcester County sits at $545,100 and Bristol at $787,750, per the published county schedule. And USDA, which is also assumable, is a rounding error here. There were 34 USDA and RHS financed closings in the entire state across those three years.
The equity gap is the whole ballgame
Here is the number that decides everything, and it comes out of our own closing records rather than a national average.
The median Massachusetts FHA or VA financed purchase in 2020 through 2022 closed at $475,000, across 5,497 closings. The median Massachusetts residential closing so far in 2026 is $650,000. So the typical seller sitting on one of these loans has roughly $175,000 of price appreciation above what they paid, plus four to six years of principal paid down. All of that is equity, and a buyer assuming the loan has to hand it over in cash or borrow it.
Work it through on the median. A $475,000 purchase with 3.5% down is a $458,375 FHA loan. At 3.00% over thirty years the payment is $1,933, and after about 64 payments the balance is roughly $404,000 with 24.7 years left. Sell that house today at $650,000 and the gap is $246,144, which is 37.9% of the price.
Financing that gap is where the savings go. Fortune’s September 30 survey put the 15 year home equity loan at 8.498% and the HELOC at 8.193%, while Bankrate’s late September averages ran closer to 7.4% and 7.1%. Trackers disagree by more than a point on this product, and a high combined loan to value purchase money second behind an assumed government loan prices at the top of that range, not the bottom. Roam, which brokers assumptions and lists Massachusetts among its markets, pairs the assumed first with a second through Spring EQ up to 85% combined loan to value, and charges buyers 1% of the sale price.
At that 85% ceiling our median buyer needs $97,500 down, takes a $148,644 second, and pays $1,933 plus $1,464, so $3,396 a month. A brand new FHA loan on the same house at 6.814% with 3.5% down runs $4,095 and needs $22,750 down.
| Assume the 3.00% | New FHA at 6.814% | |
| Cash at closing | $97,500 | $22,750 |
| Upfront MIP at 1.75% | none | $10,977 |
| First lien payment | $1,933 | $4,095 |
| Gap second at 8.498% | $1,464 | none |
| Total monthly, principal and interest | $3,396 | $4,095 |
| Blended rate | 4.48% | 6.814% |
Seven and a half years to earn back the extra cash. That is the real answer on the median deal, and it is neither a gimmick nor a free lunch. It is a good trade for a buyer who has the cash and intends to stay, and a bad one for a buyer who is stretching to close and might move in four years.
Two FHA rules that close the obvious escape hatches
When I explain the cash requirement, the first two ideas anyone has are to have the seller credit the buyer the difference, or to borrow the gap and not count it. HUD’s assumption chapter closes both doors, and almost no consumer article mentions either.
First, on seller help: “Cash contributions made by the seller to facilitate an assumption are not acceptable. The existing mortgage balance must be reduced by the amount of the contribution.” A seller credit does not bridge the gap. It shrinks the loan you are assuming, dollar for dollar, which defeats the point. The seller may still pay your normal closing costs, including the processing and credit report fees, with no reduction.
Second, on the gap loan: secondary financing is allowed, but “the repayment terms are clearly defined, and included in the underwriting analysis.” That $1,464 second lien payment lands in your debt to income ratio. You do not get to qualify on the 3% payment alone.
Two more constraints worth knowing. Loans closed on or after December 15, 1989 require a creditworthiness review for the life of the mortgage under the HUD Reform Act of 1989, so every loan from the 2020 to 2022 window is credit qualifying. And private investors are barred from assuming those loans, whether or not the seller is released. FHA assumption is an owner occupant tool. If you are shopping two and three families to hold as rentals, this is not your path.
The blended rate lies. The payment tells the truth.
The marketplaces lead with the blended rate, and the blended rate is the most misleading number in this entire subject. Our median deal blends to 4.48% against a 6.814% new FHA loan, which sounds decisive. It is not, because the second lien amortizes over 15 years while the new first would amortize over 30. A blended rate averages the interest and ignores the schedule.
So I ran the payment instead, across the range of loans a buyer will actually encounter. Rows are the assumed balance as a share of today’s price, which is the thing you can observe. Columns are what the gap second costs.
| Assumed balance | Gap second | 7.5% | 8.5% | 9.5% | 10.5% |
| 70% of price | $97,500 | +$1,014 | +$958 | +$900 | +$840 |
| 62% of price | $149,500 | +$781 | +$694 | +$606 | +$514 |
| 55% of price | $195,000 | +$577 | +$464 | +$348 | +$229 |
| 45% of price | $260,000 | +$285 | +$135 | −$20 | −$179 |
The pattern is simple once you see it. A loan that is still large relative to today’s price, meaning a recent purchase in a town that has not run away, is where this works. A well seasoned loan on a house that doubled is where it quietly stops working, which is the opposite of what most people assume.
The timeline, and why Massachusetts adds a step
Both programs set a 45 day clock. HUD requires the creditworthiness review to be completed “within 45 days from the date the lender receives all necessary documents.” The VA requires the holder to decide and notify buyer and seller “within 45 calendar days of receipt of a complete underwriting package.” Where the holder lacks automatic authority, it has 35 days to get the package to VA, and then 30 days from VA’s decision to close.
Those clocks start when the file is complete, which is the loophole. The CFPB’s issue spotlight on servicer conduct documents servicers repeatedly requesting documents already provided, failing to process assumptions, and asserting that delays past federal timelines are acceptable because of volume. I will be straight about that source: those complaints come mostly from successors in interest after a death or divorce, and the spotlight itself notes successors are treated differently from arms length purchasers. But it is the same assumption desk at the same servicers. Complaints about assumption delays rose from 67 in 2021 to 149 in 2023, and industry reporting puts real world processing at 90 to 120 days.
The incentive explains it. By that same Scotsman Guide analysis, a servicer earns 25 to 35 basis points a year to service a loan, and the servicing rights themselves carry 150 to 200 basis points of value. An assumption keeps a cheap old loan alive and pays the servicer $300. A refinance retires it and generates origination revenue. Nobody at the servicer is rooting for your assumption.
Then Massachusetts adds a layer other states do not have. Under REBA v. NREIS, our Supreme Judicial Court held that closing a real estate transaction here requires “not only the presence but the substantive participation of an attorney.” Certifying title, confirming the deed validly conveys, and holding and disbursing the funds are the practice of law. Witness only closings are out. I wrote about why Massachusetts closings require an attorney separately, and it is normally a protection worth having. On an assumption it is also a coordination problem, because a national servicing operation that handles assumptions by mail has to schedule its document delivery and payoff figures around a Massachusetts closing attorney’s table. Build the extra weeks into your offer. Do not write a 30 day close on an assumption.
Rates tripled the incentive. Assumption volume did not move.
If this tool worked the way the marketplaces describe, we would see it in the closing records. We do not.
Our archive codes the financing on closed listings, including assumptions. In 2019, when Freddie Mac’s survey averaged near 3.9% for the year, Massachusetts recorded 116 assumption coded residential closings. In 2026 so far, with rates at 7.5%, we have 96. The incentive to assume went from almost nothing to more than three points of spread, and the volume went down.
|
2019
30 year rates near 3.9%
116
assumption coded closings
|
2026 year to date
30 year rates near 7.5%
96
assumption coded closings, out of 39,177
|
Nationally the picture is the same shape. Scotsman Guide puts the assumable universe at roughly 12 million loans, about 23% of active mortgages, against only about 6,400 assumptions completed in 2023. That is one source rather than two, so treat the magnitude rather than the decimal. Meanwhile FHFA’s National Mortgage Database has 49.9% of outstanding mortgages at 4% or less as of the first quarter of 2026, down from a 65.1% peak in early 2022. The pool is enormous, it is shrinking every quarter, and it is barely being touched.
What I would actually do
If you are buying. Ask your agent to run the two searches nobody runs: Listing Terms containing Assumable, and a remarks keyword search on “assumable” and “assume the loan.” Those catch different listings, and together they caught 42 in our feed. Check Roam and Assumable.io alongside them, and treat what they show as a supplement to the MLS rather than a replacement. Then, before you fall in love with a rate, get the remaining balance and the remaining term from the seller in writing and do the payment math against a new FHA loan at today’s rate. If the balance is below roughly half of the purchase price, expect the gap financing to eat most of the benefit.
If you are selling and you have a sub 4% FHA or VA loan, this is an asset and you should say so. Set the Listing Terms field and write the rate, the approximate balance and the loan type into the remarks. Sixteen of the 19 sellers who set that flag in Massachusetts right now are not explaining it to anyone. If it is a VA loan, ask your attorney and your loan officer about entitlement before you accept an offer from a buyer who is not VA eligible, because your next purchase depends on it. Two Lynn listings priced at $479,900 went pending advertising “3% fixed interest available for total monthly payment of $2,500 or so.” That is what it looks like when it works.
If you are a buyer’s agent. Surface this for cash heavy buyers and for anyone whose payment is the binding constraint rather than their down payment. Write a longer closing timeline into the offer, name the servicer’s 45 day clock in your contingency language, and get the Massachusetts closing attorney talking to the servicer in week one rather than week six.
My honest read is that assumable mortgages deserve to be a normal part of the conversation in Lynn, Lawrence, Methuen, Lowell, Haverhill and Revere, where the 2020 to 2022 government backed share ran between 12% and 34% of all closings. They will almost never matter in Lexington or Concord, and the data is unambiguous about that. And even in the right town, this is a tool for a specific buyer: someone with real cash, a long horizon, and the patience to wait out a servicer who has no reason to hurry.
If you want to pressure test a specific house, the buyer closing cost calculator will get you to a realistic cash to close, and our buyer representation page explains how we work. Sellers wondering what their own low rate loan is worth as a selling point can start with a home value estimate. More on financing and strategy sits in our buying properties archive. Reach out anytime with questions.
Sources
- 38 U.S.C. 3714, Assumptions and release from liability
- VA Circular 26-23-10, VA Assumption Updates
- HUD Handbook 4155.1, Chapter 7: Assumptions
- HUD Mortgagee Letter 2025-23, 2026 Nationwide Forward Mortgage Limits
- 2026 FHA loan limits by Massachusetts county
- Freddie Mac Primary Mortgage Market Survey
- Freddie Mac, Mortgage Rates Average 7.03%, September 24, 2026
- Mortgage News Daily, Mortgage Rates Rise to 7.58%, September 29, 2026
- Fortune, mortgage rates for September 30, 2026
- Fortune, home equity loan and HELOC rates for September 30, 2026
- NPR, mortgage rates surpass 7% for the first time in well over a year
- FHFA National Mortgage Database, Q1 2026 outstanding mortgage rates
- FHFA NMDB Outstanding Residential Mortgage Statistics dashboard
- CFPB Issue Spotlight on servicer handling of assumptions
- Scotsman Guide, Red Tape Slows Mortgage Assumptions
- HousingWire on Roam’s down payment partnership
- REBA v. National Real Estate Information Services, 459 Mass. 512 (2011)
- USDA HB-1-3555, Single Family Housing Guaranteed Loan Program handbook
Proprietary figures in this article come from BMN Boston’s MLS PIN IDX feed and closed listing archive, pulled September 30, 2026. Rates and payments are principal and interest only and exclude taxes, insurance and mortgage insurance. Nothing here is loan advice. Confirm any assumption’s terms with the servicer in writing before you rely on them.
