Looking for the short version? Read the plain-English summary →
No town in Massachusetts produces a working rental at its typical price. Not one of the 88 we rank. The best town in the state needs a 35% discount before the numbers clear; the median needs 56%. That is not a market you underwrite your way into — it is a market you buy your way into, and the entire return lives in the price you pay rather than in anything you do afterwards.
Run its actual numbers — free, no signup — instead of the town average above.
Each town's typical home value against its typical market rent, run through the offer engine with DSCR financing, real insurance, and a requirement that the first year merely break even — not profit, break even.
| Town | Typical value | Market rent | Offer that works | Discount |
|---|---|---|---|---|
| Springfield | $307,092 | $1,836 | $198,960 | −35% |
| Chicopee | $325,808 | $1,661 | $177,304 | −46% |
| Pittsfield | $313,325 | $1,592 | $168,858 | −46% |
| Taunton | $525,825 | $2,158 | $238,768 | −55% |
| Northampton | $503,242 | $2,454 | $275,294 | −45% |
| Amherst | $551,423 | $2,491 | $279,971 | −49% |
| New Bedford | $446,955 | $1,897 | $206,510 | −54% |
| Brockton | $513,496 | $2,334 | $260,500 | −49% |
The reason is structural. A conventional loan is sized by your down payment, so a bigger cheque buys a bigger house. A DSCR loan is sized by the rent — so the loan stops growing no matter how much cash you have, and every dollar of price past that point comes out of pocket at a hundred cents.
| Town | Conventional · 25% down @ 6.75% | DSCR loan @ 8.00% |
|---|---|---|
| Springfield | −19% | −35% |
| Chicopee | −31% | −46% |
| Pittsfield | −31% | −46% |
| Taunton | −44% | −55% |
| Northampton | −34% | −45% |
| Amherst | −39% | −49% |
This matters because DSCR is what investors actually reach for. It needs no income documentation and does not touch personal debt-to-income, so it is the default product for anyone buying a third or fourth property. It is also the product that punishes overpaying hardest.
Massachusetts legalised accessory dwelling units by right in February 2025 and barred towns from requiring owner-occupancy, so the second rent is genuinely available to investors. It is a real lever, and it is routinely oversold.
What it does is reorder which towns reward you. Rank correlation between the no-ADU and with-ADU orderings is just 0.50 — half the ranking is different. High-rent towns climb hardest: Somerville +15 places, Stoneham +13, Boston +7, because the build costs the same everywhere while the second rent does not. If the ADU is part of your plan, the conventional cheap-town buy-box is the wrong list.
What it does not do is close the gap.
| Build assumption | Median discount still required |
|---|---|
| No ADU | −56% |
| Basement conversion, cheap end · $150/sqft | −47% |
| Basement conversion, lower-mid · ~$175/sqft | −52% |
| Basement conversion, mid-range · $200/sqft | −56% |
The whole strategy lives or dies on landing at the cheap end of the build range, which in practice means an existing shell: a basement or attic with the foundation, framing and roof already paid for. An ADU swings the answer by about nine points in the best case, against a gap of fifty-plus. It changes where you look. It does not change what you pay.
A DSCR lender sizes on PITIA, so the premium sits inside the loan constraint rather than beside it. The exchange rate is exact:
$1 of annual premium costs $11.36 of loan.
A premium that comes back $2,000 over the pro-forma is not a $2,000 problem. It is a $23,000 cash-at-closing problem, arriving two weeks before the close, with the DSCR ratio still reading fine on the worksheet because the lender simply re-sized the loan to hold it. This is the mechanism behind deals that die in underwriting for reasons nobody can quite name afterwards.
Modelled honestly over a ten-year hold, insurance also costs real confidence. A coastal property's probability of positive cash flow in year ten falls from 88.5% to 64.3% once premiums are allowed to step at renewal rather than drift at a flat 2.5%. Twenty-four points of comfort were an artifact of the assumption.
It was, however, a small part of why the discounts above are so deep. Correcting the insurance moved the required discount by 0–6 points, against the 13 points the financing structure contributes. Insurance kills individual deals late. It is not what makes the market hard.
Three separate ceilings set a price, and which one binds changes the advice completely:
| Ceiling | The most you can pay and still… |
|---|---|
| Cash flow | clear your monthly target |
| Cash to close | actually be able to close |
| Return | hit your cash-on-cash target |
Buyers arrive believing they are limited by the first. Past the point where the DSCR loan stops growing with price, they are limited by the second — and there each extra dollar of price is an extra dollar of their own money rather than twenty-five cents of it. Two buyers looking at the same building with different cash positions should be given different advice, and usually are not.
Every operating lever available — the ADU, management efficiency, shopping the insurance, structuring the debt — moves the answer by single-digit percentage points. The gap is forty to sixty points.
Analysis is not the bottleneck. A better spreadsheet does not create a deal that isn't there. What creates one is finding a seller with a reason to sell at 45% off, a building whose rent is materially under market, or a shell an ADU can go into cheaply. That is sourcing and negotiating work, not modelling work.
The conventional buy-box looks in the wrong place. The cheap-town heuristic and the ADU-adjusted ranking are half-uncorrelated. If the second unit is part of the plan, the screen itself has to change.
"It doesn't cash flow" is not the same as "walk away." It means the ask is wrong. Anything over $260,500 in Brockton doesn't work is a negotiating position. Brockton doesn't cash flow is a shrug. The first one is the job.
Reproducible end to end: town_screener.py over Zillow ZHVI/ZORI
for Massachusetts, offer.py for the ceilings, dscr.py
for loan sizing, adu.py for the second unit,
insurance.py for premiums. The offer arithmetic is deterministic
year-one — no simulation, no seed — and is cross-checked against the Monte
Carlo it replaces to within $0.38 per month.
Looking at a specific building, or want help reading a town you don't see above? We'll work the real numbers with you.