A Four-Star And A One-Star Sold Together In Vermont At One Blended Price — What 14,690 Nursing Homes Say That Blend Hides

One price, two very different buildings

Evans Senior Investments reported this week that an independent owner/operator exited the senior care industry by selling two skilled nursing facilities in St. Albans, Vermont for $10.5 million. Franklin County Rehab Center carries 64 beds. The Villa Rehab carries 30. Ninety-four beds, one price, and the arithmetic anyone will quote is $111,702 per bed.

That number is real, and it is also the least informative thing about the transaction.

Pull the federal file on both buildings and the blend comes apart. CMS Nursing Home Provider Information, retrieved September 14, 2026, reports Franklin County Rehab Center at 110 Fairfax Road with 64 certified beds, 58.7 average residents per day and a 4-star overall rating. It reports The Villa Rehab at 7 Forest Hill Drive with 30 certified beds, 22.2 average residents per day and a 1-star overall rating. Convert those to census and you get roughly 91.7 percent against 74.0 percent.

A 17.7-point gap. Sold at one number.

None of this is criticism of the pricing. A two-asset portfolio trades as a portfolio, and a buyer who wanted only the good one would have had to outbid a buyer who wanted both. It is a caution about what happens next: that $111,702 goes onto somebody's comp sheet, and six months from now it gets applied to a single Vermont facility that resembles neither of these two.

What we wanted to know

The obvious instinct is to say the 1-star explains the 74 percent. Bad rating, weak census, discount the beds.

We wanted to test that rather than assume it, so we pulled the CMS overall rating, certified bed count and average daily resident count for every Medicare- and Medicaid-certified nursing home in the country — 14,690 facilities, 1,563,603 certified beds — and computed bed-weighted occupancy by rating band. National bed-weighted occupancy comes to 80.5 percent, on 1,258,146 average daily residents.

By star rating:

• •1-star: 77.1 percent
• •2-star: 80.6 percent
• •3-star: 81.4 percent
• •4-star: 82.0 percent
• •5-star: 82.6 percent

The relationship is real and it is perfectly monotonic — every step up the rating ladder buys census. It is also far narrower than the industry talks about it. The entire distance from the worst rating band to the best is 5.5 percentage points. The gap that matters most, from 1-star to 2-star, is 3.5 points; from 2-star all the way to 5-star is only another 2.0.

Now put St. Albans back against it. If the national gradient explained The Villa Rehab, a 1-star sitting in a market where the 4-star down the road runs at 91.7 percent should be running somewhere in the high 80s — five and a half points below its neighbour, not seventeen and a half.

The rating does not explain that building. Something specific to it does.

That is the useful finding, and it cuts in the buyer's favour

An occupancy gap explained by a star rating is a hard problem. Ratings are sticky. The health inspection component is built from up to three survey cycles, so a genuinely bad survey history takes years to age out no matter how well the new owner operates. If census is depressed because referral sources read the star rating, the fix runs on the survey calendar, not the owner's calendar.

An occupancy gap that is not explained by the star rating is a different animal. Thirty beds at 74 percent is about eight empty beds. Eight beds in a 30-bed building is a referral relationship, a unit taken offline, an admissions process, a staffing constraint that caps admits, a payer mix decision — causes that are diagnosable in diligence and, in several cases, fixable in quarters rather than years.

That is why the expense-side story in this transaction deserves attention. The advisor did not build the case on a census turnaround. It identified $560,000 in immediate administrative expense savings and attributed the premium valuation to that. Set $560,000 against a $10.5 million price and it is 5.3 percent of the purchase price as an annual figure. Capitalise it at any rate a skilled nursing buyer would actually use and it accounts for a substantial share of what was paid — the arithmetic is ours and the cap rate was not disclosed, so treat the conclusion as directional rather than precise.

An expense thesis and a census thesis are not interchangeable. The first is largely within the new owner's control on day one. The second depends on referral sources, surveyors and a local market. Buyers routinely pay for the second while underwriting it like the first.

The variable almost nobody prices

The more surprising result in the national file has nothing to do with occupancy.

Average certified bed count falls, monotonically, as the rating rises: 120.6 beds at 1-star, 115.2 at 2-star, 107.3 at 3-star, 100.2 at 4-star, and 90.0 at 5-star. The typical 1-star facility is 34 percent larger than the typical 5-star facility.

We are not claiming size causes the rating. Large facilities skew urban, skew higher-acuity, skew toward Medicaid-heavy payer mixes and are harder to staff consistently — any of which could drive both the size and the rating. But the correlation has a direct consequence for anyone building a comp set, because it means bed count and quality are not independent variables. A per-bed number drawn from large facilities is disproportionately drawn from 1- and 2-star facilities. A per-bed number from small facilities is disproportionately drawn from 4- and 5-star facilities. Comparing them without adjusting compares two different qualities of asset while appearing to compare two prices.

It also frames how unusual the St. Albans price is. Across our own transaction tape, Fallbrook Rehabilitation and Care Center in Houston printed $74,257 per bed at 202 beds in July 2026 and a two-facility Georgia package printed $78,109 per bed at 201 beds in June 2026, while a 128-bed Wyoming facility set a state record at $156,250 per bed in April 2026 and a 696-bed Illinois supportive living and skilled nursing portfolio printed $129,310 per bed this month. Ninety-four beds at $111,702 sits in the upper half of that range, and the small-facility, high-rating, rural-New-England profile is consistent with why.

One more piece of context worth holding: the quality distribution is close to even. 40.7 percent of certified nursing homes carry a 1- or 2-star overall rating and 39.3 percent carry a 4- or 5-star. Buying into the bottom two bands is not an exotic position. It is where four out of every ten facilities in the country sit.

How we would use this

Three habits follow from the data.

First, never accept a blended per-bed on a multi-facility portfolio without decomposing it. The CMS provider file is free, updates monthly and gives you certified beds, average daily residents, overall rating and the component ratings for every facility in the country. There is no excuse for taking a portfolio average on faith.

Second, when a facility's census sits far below its rating band, stop and find out why. A 1-star at 77 percent is performing to expectation. A 1-star at 74 percent in a market where a nearby competitor runs at 92 is telling you something the rating does not — and that something is where both the risk and the repair live.

Third, be explicit about which side of the income statement your thesis lives on. Expense savings a buyer controls and census recovery a market controls are not the same asset, and they should not clear at the same price.

The CMS figures above are a federal reporting vintage, not the operating numbers the parties to the St. Albans transaction underwrote. Neither the buyer, the allocation of price between the two facilities, nor the transaction occupancy was disclosed. Everything computed here is labelled as computed, and everything not reported is left alone.