Open in interactive viewer → charts, metric popovers & call review

USPH Bought a Rate Upgrade; It Also Bought Somebody Else's Medical Bills

Converting 72 clinics to hospital-rate billing lifted net rate to a record $111 a visit and revenue 13% — yet full-year EBITDA guidance was only reaffirmed, because a $7 million self-insured claims overrun is eating the upside.
USPH · Earnings Call · 2026-11-05

The record that arrived wrapped in a loss

U.S. Physical Therapy's third quarter was, operationally, the best quarter the company has produced in years. Total revenue of $223 million rose 13%, visits of 1.68 million rose 8%, and daily visits per clinic held at 33 through the seasonally soft summer — the 27th consecutive month in which every single month set a record for that month. Adjusted EPS of $0.80 was up $0.14. Chris Reading opened by noting the company has been "as busy as we have ever been." And yet the headline GAAP line was a loss of $1.04 per share against $0.48 of earnings a year earlier. New CFO Nchacha Etta, on his first call, had to spend a paragraph explaining why: under GAAP, changes in the value of redeemable noncontrolling interests — the partner stakes the company does not own — are excluded from net income but included in the EPS denominator. When the business performs better, the partners' stakes are worth more, and that dilutes GAAP EPS. As Etta put it, “stronger operating results are contributing to the GAAP loss per share, which is why we focus on adjusted earnings per share as a clearer view of ongoing performance.” — Nchacha Etta, EVP and Chief Financial Officer · 2026-11-05 It is an accounting artifact of the partnership model, not a crack — but it is the kind of artifact that gets worse as the business gets better, and it now sits permanently next to a growth story.

The real engine is a rate arbitrage, not volume

Strip out the noise and one number explains the quarter: net rate per visit rose more than $5 to $111, and hit $112 in September — all-time records for both the level and the year-over-year increase. The driver is the hospital side of the model. The company finished converting Metro New York into its NYU Langone affiliation and onboarded a second hospital partnership, Gulf Coast, bringing in 10 facilities — 62 and 10, respectively, for 72 converted clinics. Hospital-affiliated revenue went from under 5% of the total in Q2 to over 13% in Q3, exceeding $25 million. That is why mature clinic revenue grew 6% — a number Reading called the best "in some time" — and why the growth is roughly half rate, half visits (3% and 3%, per Etta). Compare that to the mature-clinic disclosure a year ago, when the CFO described “visit growth was 2.2% and then our rate growth in mature clinics... rate decreased 2.0% for the third quarter.” — Carey Hendrickson, Chief Financial Officer · 2025-11-06 The same cohort has gone from flattish revenue to mid-single digits purely on the strength of a revenue per visit reset. What matters more is what comes next. The second half of the arbitrage — taking out the redundant billing and collections infrastructure inside the acquired partnerships — has barely started. Constantine Davides pushed on it, and Reading conceded the process is live: “that process has begun in part here very recently to begin to adjust some admin staff in our first based partnership... some of that will happen in this quarter.” — Christopher Reading, Chairman and CEO · 2026-11-05 On the prior call he was more expansive, promising the original $7.3 million 2027 EBITDA accretion estimate was conservative: “it's going to be bigger than what we originally said.” — Christopher Reading, Chairman and CEO · 2026-08-06 Wednesday he would only say the Metro side is “trending nicely ahead of that.” — Christopher Reading, Chairman and CEO · 2026-11-05 Analysts will get the number in February.

A policy win that quietly graduated from hope to fact

The most consequential item on the call got about ninety seconds. The final physician fee schedule landed the night before, and the base update comes out roughly 0.7% positive — modest, and consistent with the long-running IPSI modifier saga. But the modifier itself, the individual practice cost index adjustment that penalized efficient, low-equipment-cost providers like physical therapy, is being phased out over two years beginning in 2027. This is a genuine change in status, not a rehash. On the August call, Reading described the same issue as a 2028 event: “they've given us an early indication that in 2028, we'll see the beginning of some — what we hope to be not clear yet... resolution of that difference.” — Christopher Reading, Chairman and CEO · 2026-08-06 Now it is a confirmed phase-out a year sooner, achieved through the APTQI lobbying alliance the company has funded for years. Reading's framing: “That phaseout will actually be a very significant benefit to us because that IPSI modifier has hurt us as a very efficient provider.” — Christopher Reading, Chairman and CEO · 2026-11-05 This is the multi-year, compounding version of the story — the thing that could finally bury the Medicare rate cuts Reading declared "behind us."

The self-insured hole nobody can hedge

Which brings us to the tension. Revenue +13%, EBITDA +16%, records everywhere — and full-year EBITDA guidance was reaffirmed at $102–106 million rather than raised. The reason is a line item most investors never look at: the company self-insures its own employee health plan, and it has been unlucky in a very expensive way.

We have a handful of very large complex claims, including several premature infants that have dramatically increased our expense in this area... Year-to-date, employee insurance claims expense is $7 million higher than it was last year. In the third quarter, it was $2 million over budget.

Christopher Reading, Chairman and CEO · 2026-11-05
That cost 150 basis points of physical therapy margin; adjusted PT margin still rose 60 basis points to 19%, implying the underlying business improved by something closer to 300. The company went clinic by clinic and took out "several million dollars worth of labor," mostly administrative, and cut discretionary spend in mid-October. But the warning was explicit and repeated: “this fourth quarter, our employee expenses run hot all year... That may continue into the fourth quarter... don't get too far ahead too far, too fast.” — Christopher Reading, Chairman and CEO · 2026-11-05 Note this is the same phenomenon now squeezing the payers themselves — the global tape has Medical cost trend and Medical cost management among its persistent advancers, with the managed-care complex riding it. USPH's margin headwind is a small, unlucky slice of a system-wide problem, which is exactly why medical cost entered the company's keyword set at the top of the list this quarter. Operating margin of 10.3% in the June quarter was 2.4 points below a year earlier and roughly half the 18.4% peak of early 2021.

What stopped being said — and what the tape already believes

Two former narrative pillars have gone quiet. Cash based programs — the laser, shockwave and dry-needling upsell that occupied prominent airtime through 2025 and early 2026 — went unmentioned. Front-desk virtualization and AI documentation, the 2024–25 efficiency thesis, survived only as a single Q&A aside from Eric Williams about WelcomeWare running in 420 locations. The margin story is now a rate story. And the tape has already voted. USPH is up 36.3% in 90 days, a near-unbroken 17-week climb into the print — a strange position for a stock that sat 34% below its 2019 high. The multiples explain it: at $1.45 billion of market cap the company trades at 1.3x revenue, down 21% year over year and 13.7x operating income, versus 35x in early 2024. Meanwhile debt funded the buying: effective net cash of negative $195 million, with credit facility borrowings up to $247 million from $162 million at year-end and a new swap fixing SOFR at 4.6% on $171 million. The bull case is now unusually clean and unusually concentrated: a confirmed multi-year reimbursement tailwind, a billing-and-collections takeout still largely ahead, and a partner model most small providers cannot match on resources. The bear case is a single line of self-insured claims that management cannot forecast, a Q4 caution against extrapolation, and a stock that has already priced a good deal of the first case. Everything hinges on February, when guidance finally reconciles the two.