RadNet, Inc. (NASDAQ: RDNT) stock traded sharply lower Tuesday, Dec. 16, 2025, after a newly published short-seller report from Hunterbrook Capital questioned the durability of the company’s artificial intelligence narrative and the way investors interpret several key growth metrics. By the end of the session window covered in market data, RDNT was trading around the high-$60s after printing an intraday low in the low-$60s.
The selloff lands at an awkward time for RadNet: the company has been reporting double-digit revenue growth in its core imaging business, investing heavily in Digital Health/AI, and earning a generally positive analyst posture with many price targets still well above today’s trading range.
What happened to RadNet stock today?
The immediate catalyst for RDNT’s move on Dec. 16 was the Hunterbrook short report, which argues that RadNet’s AI story “doesn’t add up” economically and that the market’s enthusiasm may be outrunning fundamentals. The report was widely recirculated across financial media during U.S. trading hours, with multiple write-ups attributing the day’s weakness to Hunterbrook’s claims. HUNTERBROOK+3Investing.com Nigeria+3Benzin…
Price action varied throughout the day, but the direction was consistent: down, volatile, and on elevated attention. MarketBeat described the stock as gapping lower at the open and trading heavily through the morning.
The Hunterbrook thesis in plain English: “AI is a sideshow,” growth is overstated, and accounting choices matter
Hunterbrook’s report is dense, but its core assertions cluster around four themes:
1) AI/Digital Health is small relative to the whole company
Hunterbrook argues that RadNet’s AI business—marketed through its Digital Health segment and DeepHealth brand—is a modest fraction of total revenue, despite being a major driver of investor excitement. Investing.com’s summary of the report put a number on it: roughly $65 million out of about $1.5 billion over the first nine months of 2025, or less than 5% of total revenue.
That “less than 5%” framing also appears in other coverage and commentary tied to the report. TipRanks+1
2) A meaningful portion of Digital Health revenue may be internal rather than third-party adoption
A major nuance in RadNet’s Digital Health reporting is that the segment’s revenue can include intersegment (internal) revenue—in other words, software and services sold to RadNet’s own imaging centers. Hunterbrook argues that this dynamic can make the AI business look more commercially penetrated than it truly is in the broader market.
This isn’t a minor technicality: if a large share of “AI revenue growth” is effectively internal transfer pricing, skeptics will view it as operational optimization rather than standalone software scale.
3) “Same-center” growth could be flattered by footprint changes
Hunterbrook also targets RadNet’s “same-center” growth disclosures. The short thesis argues that by consolidating reporting units and closing centers located close to other RadNet sites, patient volume may shift rather than expand—creating the appearance of stronger organic growth. In one summary, Hunterbrook’s analysis suggested “organic” growth could be closer to ~2.5%–3% versus reported ranges often cited in the 6%–10% neighborhood. Investing.com Nigeria+2Benzinga+2
4) Adjusted profitability and insider selling deserve scrutiny
The report also questions RadNet’s use of “adjusted” metrics, arguing that excluding items like stock-based compensation and certain R&D can materially change how investors perceive profitability—especially for a business being valued partly as a tech/AI story. Investing.com Nigeria+2HUNTERBROOK+2
Separately, Hunterbrook highlights insider selling; media coverage of the report pointed to substantial insider sales over time and noted claims that RadNet did not respond to requests for comment in connection with the report.
Important context: Hunterbrook disclosed that Hunterbrook Capital was short RDNT at the time of publication—meaning it benefits if the stock falls. That doesn’t make the report wrong, but it does mean readers should treat it as adversarial research and verify claims against filings and company disclosures.
What RadNet’s latest official results say: strong imaging growth, rising Digital Health revenue, and big R&D spend
To evaluate a short thesis, investors usually triangulate it against primary-source disclosure. RadNet’s most recent quarterly release (Q3 2025) provides a lot of the raw material for that check.
Q3 2025 highlights (reported Nov. 9, 2025)
RadNet reported:
- Total company revenue:$522.9 million, up 13.4% year over year
- Digital Health revenue (inclusive of intersegment revenue):$24.8 million, up 51.6% year over year
- Adjusted EBITDA:$84.9 million, up 15.2% year over year, with Adjusted EBITDA margin rising to 16.2%
- Adjusted diluted EPS:$0.20 (after adjusting for items listed in the release)
- Cash balance:$804.7 million as of Sept. 30, 2025, and a Net Debt / Adjusted EBITDA ratio of roughly 1.0x
The same release also details that RadNet incurred non-capitalized R&D expenses related to “DeepHealth Cloud OS and AI solutions” during the quarter, underscoring that the company is actively spending into its AI platform—whether or not investors agree with the valuation it commands. RadNet
Guidance was raised after Q3
In that Q3 release, RadNet also raised multiple 2025 guidance items, including Imaging Center segment revenue/Adjusted EBITDA ranges and Digital Health revenue guidance. The updated ranges included:
- Imaging Center segment total net revenue: raised to $1.900B–$1.930B (revised after Q3)
- Imaging Center segment adjusted EBITDA: raised to $276M–$284M
- Digital Health segment total net revenue (inclusive of intersegment revenue): raised to $85M–$95M
- Digital Health non-capitalized R&D (DeepHealth Cloud OS & Generative AI): guidance raised to $18M–$20M
Those figures matter because they illuminate the debate at the center of today’s selloff: RadNet’s operating story (imaging center scale + improving volumes + expanding joint ventures + tuck-in M&A) has been strong, while the market is still deciding how much “tech multiple” the Digital Health/AI layer deserves. RadNet+1
Analyst forecasts and price targets: still mostly positive, even after today’s controversy
Even as RDNT fell Tuesday, published sell-side posture (as aggregated by MarketBeat) remained broadly constructive:
- MarketBeat lists a “Moderate Buy” consensus and an average target price of $87.60, noting several firms recently lifted targets into the mid-$80s to mid-$90s range. MarketBeat+1
- The same MarketBeat roundups cite examples including B. Riley ($87), Barclays ($86), Truist ($90), and Raymond James ($95) targets (with varying rating language).
- Yahoo Finance’s summary snapshot showed a 1-year target estimate of $91.71 (reflecting a different aggregation set/methodology than MarketBeat, which is common across platforms).
One additional nuance: MarketBeat also references at least one bearish ratings outlet in its compilation, highlighting that not all “analyst” inputs are bullish. MarketBeat+1
How investors typically read this: price targets are not guarantees; they’re conditional models. A credible short report can force analysts to revisit assumptions about segment mix, organic growth rates, and what multiple the market should pay for AI-flavored revenue—especially if management responds with clarifications (or if regulators/earnings disclosures surface additional detail).
Options and positioning: call volume spikes as RDNT swings
One of the more striking “today” datapoints was the options tape. MarketBeat reported that traders bought 9,142 call options in a single session—about a 1,250% jump versus typical call volume cited in its report. MarketBeat
That does not automatically mean “smart money is bullish.” Call spikes can reflect:
- directional dip-buying,
- hedging by short sellers (calls as protection),
- volatility trades,
- or event-driven speculation around a potential company response.
But it does signal that RDNT moved into a higher-attention bucket for derivatives traders on Dec. 16.
Recent corporate developments shaping the bull case (before today’s short report)
To understand why the AI narrative matters so much to RDNT’s valuation, it helps to look at what RadNet and its DeepHealth unit have been announcing in recent months:
- RadNet’s newsroom shows a string of late-2025 DeepHealth updates, including RSNA-related product positioning and workflow/AI announcements (with the most recent newsroom entry dated Dec. 1, 2025—no new press release appears posted for Dec. 16).
- On Nov. 12, 2025, RadNet/DeepHealth and GE HealthCare announced a letter of intent tied to expanding collaboration across AI-powered imaging workflows and remote scanning, including proposed distribution and modality expansion.
This is the strategic “setup” that makes a short report sting: the company is actively building a platform story around AI-enabled imaging informatics, but the bears are arguing the financial contribution hasn’t (yet) earned the hype.
Regulatory and capital markets backdrop: the shelf filing
Another thread investors may see resurfacing amid volatility: on Dec. 4, 2025, RadNet filed a shelf registration statement on Form S-3 and a prospectus supplement registering the potential resale of up to 73,567 shares by certain selling stockholders (per the company’s 8‑K).
Shelf registrations are common for flexibility and don’t automatically signal imminent dilution—but in a nervous tape, anything that smells like potential supply can become part of the trading narrative.
What’s next for RDNT stock: the catalysts investors will watch
In the near term, the market’s focus is likely to narrow to three questions:
1) Does RadNet respond directly to Hunterbrook’s claims?
As of published coverage earlier Tuesday, at least one outlet reported it had contacted RadNet for comment and had not yet received a response, while another summary said Hunterbrook claimed RadNet did not respond to its own requests.
If RadNet issues a detailed rebuttal (especially with segment KPIs like third-party AI customer growth, net retention, and the split between intersegment vs external Digital Health revenue), that could reset sentiment quickly.
2) Will analysts defend their models—or revise assumptions?
Price targets and “Buy” ratings are built on explicit narratives: procedure volume growth, margin expansion, center openings, reimbursement, joint ventures, and an AI upside option. RadNet’s own Q3 release reported strong imaging volume trends and detailed Digital Health revenue growth, while the short report argues those optics can be misleading. RadNet+1
The next wave of notes may focus less on “AI buzzwords” and more on “what is truly recurring, external, and scalable?”
3) When is the next earnings update?
Earnings dates are still estimates on many calendars (companies can confirm later). Zacks and MarketBeat both point to Feb. 26, 2026 as the expected next report date based on historical patterns.
That next earnings cycle is where this debate gets tested with fresh numbers—particularly Digital Health revenue composition, updated guidance, and any additional disclosure around “same-center” definitions and footprint changes.
The RDNT debate after Dec. 16: one stock, two narratives
After today’s news, RadNet stock is likely to trade as a referendum on which narrative investors find more convincing:
Bull narrative: A scaled outpatient imaging leader with improving advanced modality mix, operating leverage, strategic acquisitions, strong liquidity, and an AI-enabled workflow platform that can monetize across the industry over time.
Bear narrative: A solid imaging roll-up whose valuation became inflated by an AI rebrand; the “AI segment” is small and partly internal; organic growth is overstated; and adjusted metrics can flatter profitability, while insiders sell into strength. Investing.com Nigeria+2HUNTERBROOK+2
In the short run, markets are rarely philosophical. They’re practical: they want clean numbers, consistent definitions, and credible rebuttals. Tuesday’s move suggests RDNT holders are demanding more proof.
