Globus Medical, Inc. (NYSE: GMED) — The Robot Was Never the Moat

An independent fundamental research note Report date: 26 July 2026 · Coverage: Initiation · Sector: Health Care · Medical Devices (Musculoskeletal / Spine & Enabling Technology) Price reference: $76.49 (close, 2026-07-24) · Shares outstanding: 135,055,223 (Class A + Class B, 2025-12-31) · Market capitalisation: ~$10.33bn · Net cash: ~$684M · Enterprise value: ~$9.65bn

The analysis in the body of this article (Sections 1–15) deliberately carries no investment recommendation and no price target. The single exception is the clearly-labelled Claude's Take block immediately below, which is the author’s own opinion.

⚡ Claude’s Take

This block is the author’s own subjective opinion. It is general information and commentary only — it is not investment advice, and nothing in it should be relied upon as a recommendation to buy or sell any security. Do your own work. The analysis in Sections 1–15 below takes no position and contains no price target.

“You are being sold a robotics platform. You are buying a spine implant company. Fortunately, it is on sale.”

Verdict: HOLD at $76.49 — accumulate on weakness below ~$70. Not a short. Not a table-pounder either.

Directional valuation zone. I would be a buyer below roughly $70 — about 14.5x FY2026E non-GAAP EPS of $4.75 and ~2.0x book — where the net-cash balance sheet and a 6%+ free-cash-flow yield do the work even if nothing improves. I would be trimming above roughly $100–105, which requires paying ~17x a FY2028 EPS figure that only exists if the mid-70s gross-margin bridge substantially lands. At $76.49 the stock sits awkwardly between those posts: genuinely cheap on its own history, only fairly priced against its actual business quality.

What the market is pricing, and what I think it has wrong. The market is pricing GMED as a mid-single-digit grower whose adjusted gross margin stalls at the guided 69–70% and never reaches management’s mid-70s target. That is a defensible assumption and it is why the stock trades at the 5.7th percentile of its own ten-year valuation range — the cheapest ~1% on earnings, ~5% on sales, ~11% on book. Where I think consensus is offside is narrow but real: the mid-70s target is not a blue-sky aspiration, it is a restoration of the 74–77% gross margin standalone Globus actually earned every year from 2016 to 2022, now applied across a doubled revenue base. Six consecutive quarters of sequential adjusted-gross-margin expansion, and a Q1 2026 that held 69.2% flat despite the normal seasonal step-down, say the machine is working. Roughly $170M of gross profit sits in that bridge — about $0.94 of EPS, ~20% of current earnings — and the market is paying nothing for it.

Where I think the bulls are wrong, which matters more. The entire sell-side “surgical robotics platform” framing is contradicted by the company’s own revenue disaggregation. Enabling Technologies — imaging, navigation and robotics, the ExcelsiusGPS franchise — was $141.0M in FY2025, down 8.4% from $154.0M, and is 4.8% of revenue. Management is now explicitly shifting the robot from outright sale to lease-and-rental so it can protect implant pull-through. That is a sensible commercial decision and it is also an admission: you give the razor away more cheaply when you can no longer charge for it. The robot was never the moat. What GMED actually owns is a vertical-integration cost advantage worth several hundred basis points of gross margin, and a large exclusive US sales force that is currently winning share from four separate competitors in disarray. That is a real, second-tier, supply-side moat. It is not a compounder.

Framing: contrarian/value with a capital-cycle tailwind — explicitly NOT momentum and NOT a falling knife. The factor model is unusually clear here. GMED carries zero momentum loading (L1-zeroed in all four nested models), a mild Value tilt (+0.10 to +0.12), a SmallSize tilt and — oddly for a non-payer — a positive DividendYield loading, meaning it trades with the defensive value cohort rather than with growth. Model R² is only ~20% against 45.5% annualised idiosyncratic volatility: roughly 80% of the return variance is stock-specific, so this is an execution story, not a regime story. It is not a falling knife because the fundamentals are improving while the price falls — the most diagnostic fact in this file is that GMED fell -8.4% on 2026-05-08 having just beaten on revenue and raised full-year EPS guidance by $0.30. Markets do not sell beats-and-raises unless they distrust the composition, and the composition on offer was Nevro shrinking -17.1% sequentially and a top line pointedly not raised alongside the bottom.

Conviction: MEDIUM, and capped there deliberately. Two things stop me going higher. First, return on invested capital is ~10% — at, not above, a reasonable cost of capital. Strip goodwill and intangibles and the operating business earns ~21%; the difference is the price paid for NuVasive. Legacy holders financed a doubling of revenue via 36% dilution and, three years on, are back at roughly the return on capital they started with. Second, governance. GMED is a controlled company (David C. Paul holds >50% of voting power; insiders ~66.3%), the annual cash bonus is 90% weighted to revenue alone with no margin, EPS, cash-flow or ROIC gate — precisely the incentive that produces dilutive revenue-buying — and the insider record since the NuVasive close is $50.2M of open-market sales and exactly zero purchases, with the CFO and a director selling at $101.10 on 2026-01-08, the day of the five-year intraday high of $101.40.

What would flip me bullish (one thing): two more quarters of sequential adjusted-gross-margin expansion that carries GMED above ~71% while base-business organic growth holds ≥7% — that would make the mid-70s bridge a schedule rather than a hope, and would justify paying up. What would flip me bearish (one thing): US Spine growth decelerating below ~5% while Enabling Technologies revenue declines again in FY2026 — that would confirm the share gains were a one-time harvest of competitor disruption rather than a durable advantage, and would leave a ~10%-ROIC, controlled, price-deflating implant company with nothing to re-rate on.

📈 Stock Price Action — Five-Year Event Map

GMED has completed a full five-year round trip to nothing. The stock closed at $78.59 on 2021-07-01 and at $76.49 on 2026-07-24 — a negative five-year price return (FactorsToday five-year annualised return: -0.42%). Between those two points it fell to a five-year low of $43.79 (2023-11-13), more than doubled to a five-year closing high of $96.82 (2026-04-20) and an intraday high of $101.40 (2026-01-08), and has since given back -21.0% from the closing high. The 52-week range on closes is $52.63 – $96.82. The path was violent: annualised volatility of 49% over the past year, a maximum drawdown of -47.9%, and 45.5% annualised idiosyncratic volatility against a factor-model R² of only ~20%.

Cycle narrative. (1–2) The February 2023 NuVasive announcement is the origin of everything that followed: an all-stock deal that converted each NuVasive share into 0.75 Globus shares took the count from 100.2M to 136.3M, and the market marked the equity down 18% on the day and 43% over the following nine months. (3–4) The 2024 recovery was earned — the May 2024 +20.2% day was the market conceding that synergy capture was real, and the stock ran to $92.72 by January 2025. (5) May 2025 broke it: a Q1 miss driven by GMED’s own inability to get surgical sets and inventory into the field, compounded by soft Enabling Technologies, cost -23% in a session. (6) Two months later the CEO resigned from both the executive suite and the board effective immediately, disclosed in the same 8-K that furnished preliminary Q2 sales — an unusual pairing that the tape read as bad news, taking the stock to its 52-week low of $52.63. (7) The November 2025 +35.9% single-day move is one of the largest in the stock’s history and marked the true fundamental inflection: sequential gross-margin expansion, US Spine at +9.6%, and Nevro accretive 15 months ahead of guidance. (8) The most recent leg down is the important one for anyone underwriting the stock today. On 2026-05-08 GMED reported revenue of $759.9M (+27%), non-GAAP EPS of $1.12 (+64.7%), and raised full-year EPS guidance from $4.40–4.50 to $4.70–4.80 — and fell 8.4%. The market’s objection was composition, not level: Nevro revenue fell $17.1M sequentially, Enabling Technologies is being deliberately shifted toward lease structures that suppress recognised revenue, and management conspicuously declined to raise the top line. The stock has drifted down a further ~7% in the ten weeks since.

Price moves are facts drawn from the AZI daily price file; attributed causes are interpretation, cross-referenced to 8-K filings, earnings dates and transcripts. No price target or recommendation is expressed or implied in this section.

1. Executive Summary

Globus Medical is a $2.94bn-revenue musculoskeletal device company, built through two acquisitions in thirty-one months, that sells spinal implants, trauma and orthopaedic hardware, biologics, intraoperative neuromonitoring, spinal cord stimulation and — in a much smaller way than its narrative implies — surgical robots and navigation.

The business is better than the industry, and the stock is cheaper than either. Standalone Globus ran 74–77% gross margins and mid-teens returns on invested capital for seven years before the NuVasive merger, on the strength of genuine vertical integration: it manufactures its own implants and assembles its own robots. That cost advantage survives. The problem is that management spent it. The all-stock NuVasive merger (closed 2023-09-01) diluted legacy holders ~36% and cut consolidated return on invested capital from ~10–13% to 4.0–4.7%; three years later ROIC has recovered only to ~10% — approximately where it started, on a business with a third more shares outstanding. Strip goodwill and intangibles and the operating business still earns ~21%; the gap is entirely the price paid.

The robotics story does not survive contact with the revenue disaggregation. Enabling Technologies — the ExcelsiusGPS imaging/navigation/robotics franchise that anchors the bull case — generated $141.0M in FY2025, down 8.4% from $154.0M, and represents 4.8% of revenue. Management is now deliberately shifting the robot from outright sale toward leases and rentals to protect implant pull-through, which will suppress recognised Enabling Technologies revenue further. GMED is an implant company with a robot, not a robotics company with implants.

What is genuinely working is the core. US Spine has grown 10.0%, 9.7%, 9.6% and 7.4% (day-adjusted) across the last four reported quarters against a market growing low-to-mid single digits — 58 consecutive weeks of growth — driven by competitive rep recruiting and a deep expandable-interbody and MIS-screw portfolio. Adjusted gross margin has expanded for six consecutive quarters and held 69.2% flat in Q1 2026 despite the normal Q4→Q1 seasonal step-down. Free cash flow has grown from $104M (FY2022) to $579M (FY2025). The $450M of NuVasive-assumed convertible notes were repaid in cash in March 2025; the company is debt-free with ~$684M of net cash.

Earnings quality requires care. FY2025 GAAP net income of $537.9M includes a $117.7M non-taxable bargain purchase gain on Nevro and benefits from a $46.3M deferred-tax valuation-allowance release — together roughly 30.5% of GAAP net income is non-recurring, and normalised GAAP EPS is closer to $2.70–2.75 than the $3.92 reported. Separately, the headline GAAP gross-margin improvement from 55.6% to 64.3% is roughly 90% purchase-accounting roll-off; the honest underlying move was 67.4% → 68.1%. Management’s non-GAAP EPS of $3.98 correctly excludes the bargain purchase gain and taxes at 24%, and is the right anchor.

The industry is mediocre but the capital cycle is unusually favourable. Spine is fragmented, surgeon-preference-driven, working-capital-intensive and subject to ~1% annual price erosion under hospital/GPO consolidation. But scaled competitors are leaving: Stryker divested its spine business to VB Spine, ZimVie (itself a Zimmer Biomet spin-off) divested spine, and Medtronic’s franchise sits in a segment it treats as structurally capped. Capital and management attention are exiting the category — historically the precondition for improving returns among the committed survivors. GMED is the largest committed pure-play and the natural recipient of every disrupted competitor’s reps and accounts.

Governance is the weakest link and it has not been addressed. GMED is a “controlled company” under NYSE Rule 303A.00 — founder and Executive Chairman David C. Paul holds >50% of voting power through a Class B block, insiders control ~66.3%, and 192.6M further Class B shares remain authorised for issuance. The annual cash bonus for named executives is 90% determined by revenue alone, with the remaining 10% a discretionary compliance/quality assessment; there is no margin, EPS, cash-flow or ROIC gate. That is the incentive structure most likely to produce exactly the dilutive revenue-buying that NuVasive represented, and it is unchanged. Since the merger closed, insiders have executed $50.2M of open-market sales and zero open-market purchases, with the CFO and a director selling at $101.10 on 2026-01-08 — the session in which GMED printed its five-year intraday high of $101.40.

Valuation. At $76.49 the equity is ~$10.33bn and the enterprise value ~$9.65bn: 3.0x FY2026E sales, ~9.3x FY2026E adjusted EBITDA, 15.9–16.3x FY2026E non-GAAP EPS guidance of $4.70–4.80, and a 5.6% trailing free-cash-flow yield. On its own ten-year history GMED sits at the 5.7th percentile composite — 1.2nd on P/E, 5.3rd on P/S, 10.7th on P/B. All three agree, so this is not a single-metric artefact. The embedded expectation is a mid-single-digit grower whose gross margin never progresses beyond the guided 69–70%. The entire valuation debate is whether the mid-70s gross-margin bridge — worth roughly $170M of gross profit and ~$0.94 of EPS — is a schedule or a slogan.

2. Business Overview

Globus Medical was founded in 2003 in Audubon, Pennsylvania by David C. Paul, David D. Davidar and Andrew Iott, and listed on the NYSE on 2012-08-02. It employs approximately 6,000 people. The company describes itself as a musculoskeletal technology company and organises its reporting into two product families and two geographies.

Musculoskeletal Solutions — $2,797.9M of FY2025 revenue (95.2% of total). This is the business. It comprises:

• Spinal implants and instrumentation — the core franchise. Traditional fusion hardware (pedicle screw and rod systems including CREO, Reline, REVERE and REVOLVE; plating systems; interbody spacers; corpectomy devices), expandable interbody devices (SABLE, RISE, ALTERA, TLX, Modulus, CALIBER — a category where GMED holds genuine product leadership), minimally-invasive (MIS) pedicle screws, cervical fixation and 3D-printed titanium lattice spacers (HEDRON).
• Motion preservation — dynamic stabilisation, total disc replacement, interspinous distraction.
• Regenerative biologics — allografts and synthetic alternatives, including the Bone Bank Allografts unit; GMED distributes human cell, tissue and cellular/tissue-based products.
• Trauma and orthopaedics — fracture plates, compression screws, intramedullary nails, external fixation, the Anthem elbow plating system, the PRECICE limb-lengthening portfolio (acquired via NuVasive), and hip and knee arthroplasty implants sold through independent agents.
• Intraoperative neuromonitoring (IONM) — proprietary software-driven nerve detection and avoidance services supporting spinal and cranial surgery, delivered onsite and remotely. A services business acquired with NuVasive.
• Neuromodulation — added April 2025 via Nevro: the HFX spinal cord stimulation platform including the Senza SCS system for chronic trunk and limb pain and painful diabetic neuropathy, plus minimally-invasive treatment for chronic sacroiliac joint pain.

Enabling Technologies — $141.0M of FY2025 revenue (4.8% of total). The imaging, navigation and robotics (“INR”) portfolio: the ExcelsiusGPS robotic guidance and navigation platform (launched 2017, floor-mounted, FDA-cleared for cranial, cervical, sacrum, pelvic and orthopaedic applications with end-effector tracking), the Excelsius Hub navigation system, ExcelsiusXR augmented-reality headset, Excelsius3D intraoperative imaging, and Surgimap surgical planning software (acquired with Nemaris in 2018). Newly cleared in Q2 2026: the Scripps patient-specific lumbar interbody spacers and precision-bent rods, designed in Scripps Studio and integrated with the Excelsius suite.

Geography. United States $2,367.6M (80.6%), International $571.3M (19.4%) in FY2025. International grew 10.0% in FY2025 against US growth of 18.4% (the latter flattered by Nevro, which is predominantly US). International strength in FY2025 was concentrated in EMEA and LATAM — the UK, Germany, Poland, Brazil and Mexico — plus Australia and Japan in APAC.

How the money is made. GMED sells consumable implants and disposables through an exclusive direct US sales force supplemented by distributors, and through a mix of direct and distributor channels internationally. Revenue for implants and disposables is recognised when the product is implanted or used in surgery (for consigned sets) or on transfer of title (for direct sales). Capital equipment (INR systems) is recognised when control transfers, typically on shipment or delivery, with a portion deferred against maintenance and support obligations. Neuromonitoring is a fee-for-service business.

Recurring versus non-recurring. This is a genuinely high-repeat business, though not a contractual-recurring one. Roughly 95% of revenue is consumable implant, disposable and biologic sales that recur with procedure volume, plus fee-for-service neuromonitoring. There is no subscription base and no contractual lock-in; revenue recurs because surgeons keep operating and keep choosing the same tray. The ~4.8% capital-equipment tail is the only genuinely lumpy component, and management is actively shrinking its revenue profile in favour of lease and rental structures precisely to convert it into recurring implant pull-through.

The working-capital reality. Spine is a consigned-set business: complete instrument trays and full implant size ranges must sit in hospitals and in reps’ cars before a single unit is sold. GMED carried $759.3M of inventory against $957.8M of FY2025 cost of sales — approximately 289 days of inventory and a cash conversion cycle of 317 days. Capital expenditure runs 5–6% of sales largely to build these sets. This is not a mismanagement signal; it is the structural cost of the business model, and it is why “revenue growth” in spine consumes cash before it produces it.

Verdict. A conventional, consumable-driven musculoskeletal implant business with a small and shrinking capital-equipment attachment, heavily US-weighted, structurally working-capital-hungry, and materially more diversified across spine, trauma, biologics, neuromonitoring and now neuromodulation than it was three years ago. The diversification is real and reduces single-category risk. It was purchased, not built.

3. Industry Dynamics

Market structure. Global spine is a roughly $11–12bn category growing at low-to-mid single digits, and it is unusually fragmented for a medical device market of that size. GMED’s FY2025 10-K names as competitors Medtronic, DePuy Synthes (Johnson & Johnson), Stryker, Zimmer Biomet, Smith + Nephew and VB Spine, plus Alphatec Holdings, Orthofix, Integra LifeSciences, ZimVie, Boston Scientific and other smaller public and private companies. That list contains three of the largest medical device companies on earth alongside sub-$500M-revenue specialists — a structure that tells you barriers to entry at the product level are low. Spinal hardware is, at the commodity end, machined titanium and PEEK; the regulatory path is predominantly 510(k) rather than PMA; and a surgeon who wants to switch trays can do so between cases.

The demand side is durable. An ageing population with degenerative disc disease, spinal stenosis and deformity produces a reliable low-to-mid-single-digit procedure volume tailwind that is largely insensitive to the economic cycle. Elective deferral is a real short-cycle risk (COVID demonstrated it) but not a structural one. Deformity and revision surgery, in particular, skew toward the complex, high-ASP end where GMED is well represented.

The price side is chronically negative. Hospital and IDN consolidation, GPO centralised purchasing and value-analysis committees exert continuous downward pressure on implant ASPs. The author’s prior published work on Medtronic characterises spine hardware as one of several legacy categories where “price is a chronic low-single-digit annual headwind, so unit volume and favourable mix must run hard merely to hold revenue.” GMED’s own CEO models it explicitly: “you are going to launch new products, but you are going to see price erosion — you might be down that 1%.” The corollary, which he also stated, is that the path to higher gross margin “majority… really comes from our ability to drive costs” — not price. A business that cannot raise price does not have pricing power, and pricing power is the cleanest test of a moat.

The site-of-care shift is a genuine structural change. Spine procedures are migrating from hospital inpatient to hospital outpatient and ambulatory surgery centres. This favours vendors offering capital-light, portable enabling technology and single-vendor full-procedure portfolios — a tailwind for GMED’s floor-mounted (rather than table- or ceiling-mounted) ExcelsiusGPS and its “one-stop shop” positioning, and a headwind for vendors dependent on large hospital capital budgets.

Regulation and reimbursement. The 510(k) pathway keeps product-level entry barriers low. Reimbursement is the more material exposure. Spinal fusion has been subject to recurring coverage scrutiny for indication creep, and — more immediately — spinal cord stimulation, Nevro’s entire market, may fall within the CMS WISeR prior-authorisation pilot. Asked directly about this on the Q1 2026 call, CEO Pfeil said he had “not seen that impacting us or impacting how we go to market” and declined further comment. That is an unresolved exposure, not a cleared one, sitting on top of a business already declining.

The capital cycle — the most important thing in this section. Applying Marathon’s supply-side lens, spine is in an unusually favourable phase. Consider the recent history of scaled participants:

• Stryker divested its spine business to VB Spine (the Viscogliosi Brothers vehicle). The author’s prior published work on SYK describes the company’s growth acceleration as partly achieved by “subtract[ing] a slow one (spine)” — an explicit, deliberate exit by the best operator in orthopaedics.
• ZimVie, itself a March 2022 spin-off of Zimmer Biomet’s dental and spine units, subsequently divested its spine business. Zimmer Biomet therefore exited spine twice: once by spinning it, once by watching the spun entity sell it.
• Medtronic retains a large spine franchise but treats it as part of a structurally growth-capped, price-pressured portfolio segment.
• Notably, VB Spine appears in GMED’s FY2025 10-K competitor list and not in earlier years — GMED is documenting in real time that its competition is being transferred from strategic owners to financial ones.

When capital and senior management attention leave an industry, the remaining committed participants face fewer well-funded rivals, less irrational price competition, and a steady supply of dislocated sales representatives and accounts. That is precisely what “competitive recruiting remains a strategic priority” and “58 weeks of consecutive growth” look like from the inside. GMED is the largest committed pure-play left standing, and it is harvesting.

Verdict: a structurally MEDIOCRE industry in an unusually FAVOURABLE phase of its capital cycle — and the distinction matters enormously. On structure, spine fails most tests one would want: low product-level entry barriers, no pricing power, chronic ASP deflation, fragmented competition, surgeon-portable demand, and a working-capital burden that makes growth cash-consumptive. Nothing about the next decade changes those facts. But on cycle, the supply side is contracting for the first time in years, and the profit pool available to the committed is expanding at the expense of the departing. An investor in GMED is therefore underwriting a cyclical opportunity inside a structurally unattractive category. That can be a perfectly good investment. It is not a franchise one holds indefinitely without watching.

4. Competitive Position

4.1 What the moat actually is

Applying the Greenwald taxonomy honestly, GMED possesses a supply-side (cost) advantage, reinforced by a distribution advantage. It does not possess demand-side customer captivity, network effects, or a meaningful intangible/brand barrier.

The cost advantage is real and financially verifiable. GMED manufactures the substantial majority of its implants in-house and assembles its INR systems in-house — a degree of vertical integration unusual in a category where many competitors outsource machining. The financial fingerprint is unambiguous: standalone Globus posted gross margins of 76.1% (2016), 76.3% (2017), 77.6% (2018), 77.1% (2019), 72.4% (2020), 75.0% (2021) and 74.2% (2022) — a seven-year band of 72–78% in a category where scaled competitors run materially lower and where GMED’s own acquired NuVasive business ran lower still. This satisfies the test that a moat must tie to a financial outcome that would deteriorate without it: remove the in-house manufacturing and several hundred basis points of gross margin go with it.

It is also the mechanism behind the entire forward thesis. Management’s “mid-70s adjusted gross margin” target is not an aspiration to a level GMED has never achieved — it is a commitment to restore the legacy Globus margin structure across the merged revenue base by migrating NuVasive and Nevro production into Globus plants and Globus supply-chain practices. CFO Kline framed it exactly this way: “going back to what NuVasive had done historically from a manufacturing perspective versus what Globus had done — trying to align those sites and the approach within those sites.” A restoration target is far more credible than a de-novo one, and the evidence supports it: six consecutive quarters of sequential adjusted gross margin expansion through Q4 2025, with Q1 2026 holding 69.2% flat sequentially despite the seasonal revenue step-down that normally costs margin. Legacy-Globus standalone adjusted gross margin was 69.3% in Q1 2026.

The distribution advantage is real but symmetric. GMED fields a large, exclusive US direct sales force and has demonstrably out-executed rivals on the operational plumbing that sales forces care about — getting instrument sets and consigned inventory into the field. CEO Pfeil is explicit that this is the weapon: “our execution around rep onboarding with sets and inventories continues to differentiate us, as our supply chain meets the needs of the field by providing high-return capital investments.” Note what this actually is: GMED wins reps by funding their inventory better than the competition can. That is a balance-sheet advantage converted into a commercial one, and it works — but it is symmetric by construction. GMED’s stated growth strategy is to poach competitors’ representatives (“competitive recruiting remains a strategic priority, and the spine leadership team is aligned from the top down to aggressively perform against this objective”). An advantage that consists of taking other people’s salespeople is, definitionally, not a barrier to entry — it is a bidding war GMED is currently winning because its balance sheet is better and its competitors are in disarray.

4.2 Pressure-testing the robotics moat — and rejecting it

The consensus bull case on GMED rests on ExcelsiusGPS as a razor-and-blade engine: place robots, lock in surgeons, harvest implant pull-through. Management encourages this framing. The data refuses it.

Enabling Technologies revenue, as disclosed in the FY2025 10-K:

The franchise that is supposed to be the moat shrank 8.4% in FY2025 and now represents 4.8% of revenue. Q1 2026 revenue of $26.9M annualises near $108M; one covering analyst modelled $104M for FY2026 against a consensus of ~$150M, and management explicitly declined to reconcile the gap.

More telling than the level is the strategic direction. Management is deliberately converting the robot from a product it sells into a placement it subsidises:

“the mix of pipeline deals is shifting with a greater focus on leases and rentals compared to the historical mix of outright sales, which historically resulted in higher upfront revenue recognition… we are looking to drive the implant and the implant procedure to make sure that we are getting more of that case.” — CEO Keith Pfeil, Q1 2026 call

This is a rational commercial decision — implant annuity revenue is worth more than a one-time capital sale — and it is simultaneously an admission of weakening bargaining power. Compare the reference case: Intuitive Surgical’s da Vinci instruments are chip-locked to the system, a technical lock-in that lets Intuitive charge for both razor and blade and delivers ~84% recurring revenue. Stryker’s Mako lock-in is economic and behavioural rather than technical, but Mako has surpassed 2 million cumulative procedures and Stryker out-grows the orthopaedic market by a sustained 200–300bps. ExcelsiusGPS, launched in 2017, has accumulated approximately 130,000 cumulative procedures — roughly 6% of Mako’s base, in a platform that is nine years old.

GMED’s robot lock-in is neither technical nor scaled. It is a workflow preference, and management’s own competitive commentary concedes it is under pressure:

“as more competitors have entered the space, I would say that the speed to closing deals has elongated a bit because hospitals are now requiring everyone to go back and look at all the competitive offerings that are out there.” — CEO Keith Pfeil, Q1 2026 call

When your customers start running competitive bake-offs on a product you previously sold on preference, you do not have a moat around that product. The robot was never the moat. It is a competitive necessity — table stakes that GMED happens to execute well — and its principal economic function today is to defend the implant business, not to earn a return of its own.

4.3 Direct competitive comparison

• Versus Medtronic — GMED’s principal rival by management’s own account (“we are competing mainly against Medtronic. That is what I see routinely”). Medtronic brings the Mazor robot, StealthStation navigation, vastly greater scale and hospital-wide contracting leverage. GMED’s counter is portfolio depth in expandables and MIS, faster product cadence, and a floor-mounted single-unit robot architecture that is genuinely differentiated on workflow. GMED is taking share.
• Versus DePuy Synthes (J&J) and Zimmer Biomet — larger but strategically ambivalent about spine; Zimmer Biomet has now exited twice.
• Versus Stryker / VB Spine — Stryker sold. VB Spine is a financially-owned asset without a strategic parent’s R&D budget. This is where GMED’s reps and accounts are coming from.
• Versus Alphatec (ATEC) — the most aggressive share-taking specialist, at $8.26 a share and structurally unprofitable, competing hard on surgeon relationships and lateral-approach technique but without GMED’s manufacturing cost base or balance sheet.
• Versus Orthofix (OFIX) — $11.25 a share, sub-scale, perennially restructuring.
• Versus Boston Scientific / Abbott / Medtronic in neuromodulation — a concentrated market where Nevro is the sub-scale fourth player and is losing revenue. GMED is the weak hand here, not the strong one.

4.4 The share-stability test

Greenwald’s cleanest empirical test of a moat is stability of market share, not level. On this test GMED gives an ambiguous but currently favourable answer. US Spine grew 10.0%, 9.7%, 9.6% and 7.4% (day-adjusted) across the last four reported quarters against a market growing low-to-mid single digits — clear, sustained share gain. Large share shifts are evidence against a stable, moated oligopoly and for a contestable market. In a genuinely moated industry nobody gains 500bps of relative growth for four straight quarters.

The honest reading is that spine remains contestable, and GMED is currently the one doing the contesting because four of its competitors are distracted, divesting or capital-constrained. That is a real and bankable opportunity. It is not evidence of a barrier.

Verdict: a genuine but second-tier moat — a verifiable vertical-integration cost advantage worth several hundred basis points of gross margin, plus a balance-sheet-funded distribution advantage that is symmetric and therefore not a barrier to entry. There is no durable robotics moat; the Enabling Technologies franchise is 4.8% of revenue, shrank 8.4% in FY2025, and is being repositioned as a subsidy. GMED can protect its margin. It cannot protect its price, and it cannot prevent a determined competitor from recruiting its reps back. This is a good business in a contestable market, not a franchise.

5. Growth History and Forward Opportunities

5.1 The historical record, disaggregated

The critical disaggregation is FY2025. Management disclosed that base business revenue excluding Nevro grew 5.0%, with Nevro contributing $293.6M. The 16.7% headline is approximately twelve points of acquisition and five points of operations. Anyone modelling GMED off reported growth rates is modelling an M&A programme, not a business.

Look inside the year and the picture is more encouraging than that single figure suggests, because the base business accelerated sharply through it. H1 2025 was damaged by GMED’s own supply-chain failures — the company could not get surgical sets and inventory into the field, which cost it the Q1 2025 print and a -23.0% single-day move. H2 2025 base-business organic growth was 8.8%, and Q4 2025 base business grew 10.6%. The five-percent full-year figure is an average of a broken first half and a strong second.

Where the growth is coming from now (Q1 2026):

5.2 Quality assessment of the growth

High-quality elements. US Spine growth of ~10% for three consecutive quarters against a low-to-mid-single-digit market is genuine, broad-based share capture — management notes double-digit growth across standard fixation, MIS pedicle screws, expandable TLIF, ALIF, posterior cervical and cervical plating, i.e. not a single-product spike. Trauma at +34% is a real second leg, helped by the Anthem elbow plating system exceeding expectations and by GMED successfully transferring PRECICE limb-lengthening manufacturing from former NuVasive facilities into Globus plants, where “manufacturing output has now surpassed the historical output at the former facility.” That is vertical integration doing exactly what it is supposed to do. International recovery is real and is lapping self-inflicted 2025 supply problems.

Low-quality elements. The single largest contributor to reported growth — Nevro — is shrinking, and shrinking because of GMED’s own actions: “a decline in revenue, driven by the structural changes made within sales and marketing of the Nevro business at the tail end of 2025.” CFO Kline expects it “will probably get a little bit worse before it gets better.” Enabling Technologies grew 21% in Q1 2026 only against a soft comparable and fell 8.4% for FY2025 as a whole; management has pre-announced that the lease/rental mix shift will suppress it further. And a material portion of US Spine share gain is a harvest of competitor dislocation (Stryker’s exit, ZimVie’s exit, NuVasive customers who stayed) — real, but not obviously repeatable once the dislocation is fully harvested.

5.3 Forward opportunities

1. The gross-margin bridge (the largest single value driver). Moving adjusted gross margin from ~69.5% to the mid-70s target on ~$3.4bn of revenue is roughly $170M of incremental gross profit, or approximately $0.94 of EPS after tax on ~137M diluted shares. This is not a growth opportunity in the revenue sense but it is worth more than several years of revenue growth.
2. Nevro turnaround optionality. GMED paid $252.5M for a business generating ~$390M annualised revenue and produced a $117.7M bargain purchase gain doing it. Even stabilised at a lower revenue base, the asset is cheap. If GMED’s selling model eventually works — management expects a return toward historical run-rate “late in the second half” of 2026 — this is meaningful upside. If it does not, the downside is bounded by a small purchase price.
3. Patient-specific implants (Scripps). Two FDA 510(k) clearances received early in Q2 2026 for patient-specific lumbar interbody spacers and precision-bent rods, designed in Scripps Studio and integrated with ExcelsiusGPS/Hub/XR. Management claims GMED will be “the only company positioned to offer a complete portfolio of patient-specific lumbar interbody spacers and rods integrated with our enabling technology.” If that claim holds, this is the most credible route to the pricing premium the company otherwise lacks — patient-specific implants are the one spine category where ASP erosion is not the default.
4. International penetration. At 19.4% of revenue against US-centric peers running 30–40%, geographic expansion is a multi-year runway, though a lower-margin and slower one.
5. Continued competitive rep recruiting. As long as VB Spine, ZimVie’s divested spine assets and Alphatec remain capital-constrained, the supply of dislocated reps continues.
6. R&D re-investment. Management is guiding R&D up to 5–6% of sales in 2026 (from 5.0% in FY2025 and a synergy-suppressed 4.4–4.8% in recent quarters), explicitly “doubling down in spine R&D and enabling tech R&D.” This is the right decision and a near-term earnings headwind.

Verdict: mixed quality, improving trajectory. The headline growth record is an acquisition record and should be discounted accordingly — base-business organic growth was 5.0% in FY2025, and FY2026 guidance of 8.2–9.6% still embeds a full-year-versus-nine-months Nevro contribution. But the underlying trend is genuinely better than the headline: 8.8% base organic in H2 2025, 13.2% base growth in Q1 2026, share gains that are broad rather than concentrated, and a second leg in trauma that GMED built rather than bought. The most valuable forward opportunity is not revenue at all — it is the gross-margin restoration, which management has now delivered for six consecutive quarters and which the market is not paying for.

6. Financial Quality

6.1 Revenue, margin and the purchase-accounting overlay

Two adjustments must be made before any of the FY2025 improvement is believed.

(1) The gross-margin illusion. GAAP gross margin rose from 55.6% to 64.3% in FY2025 on management’s full-year basis (the annual statements show 58.9% → 64.3% on a slightly different basis) — an apparent expansion of many hundreds of basis points. Adjusted gross margin rose from 67.4% to 68.1% — just 70bps. The difference is the roll-off of inventory step-up amortisation on the NuVasive and Nevro acquisitions, a pure purchase-accounting artefact. CFO Kline stated it plainly for Q4: the GAAP improvement was “driven primarily by lower inventory step-up amortization.” Roughly 90% of the headline GAAP gross-margin expansion is accounting, not operations. An analyst extrapolating the GAAP trend will be badly wrong.

That said, the 70bps of real expansion understates the momentum, because it averages a weak H1. The quarterly sequence — six consecutive quarters of sequential adjusted gross-margin expansion, ending at 69.2% in Q4 2025 and holding 69.2% in Q1 2026 against a normal seasonal decline — is the number that matters, and it is genuinely good.

(2) The FY2025 earnings-quality problem. GAAP net income of $537.9M contains two large non-recurring items:

• A $117.7M non-taxable bargain purchase gain on the Nevro merger, disclosed as a discrete income-statement line and given its own MD&A subsection. This is 4.0% of net sales and 19.5% of pretax income. It is non-cash and will not repeat.
• A $46.3M release of valuation allowances on deferred tax assets (R&D credits), which together with the non-taxable gain drove the GAAP effective tax rate to 11.1% against a non-GAAP rate of 24.0%.

Together these contribute roughly $164M, or ~30.5% of GAAP net income. Normalising — removing the bargain purchase gain and taxing the remainder at 24% — produces GAAP net income of approximately $370M and EPS of ~$2.70, versus the $3.92 reported. GAAP EPS overstates FY2025 economic earnings by roughly 45%.

Management’s non-GAAP EPS of $3.98 does exclude the bargain purchase gain and does tax at 24%; it adds back merger intangible amortisation ($276.8M of total D&A against $174.4M of capex), a $13.4M estimated litigation charge and acquisition/integration costs. It is the appropriate anchor for valuation, with the caveat that adding back all intangible amortisation flatters a company whose growth has been acquisition-driven.

The FY2024 comparison is the reverse distortion. FY2024 GAAP EPS of $0.75 was depressed by peak inventory step-up amortisation, a $43.3M foreign currency transaction loss and $12.6M of acquired IPR&D. Non-GAAP EPS was $3.04. The FY2024→FY2025 GAAP EPS move from $0.75 to $3.92 (+423%) is almost entirely accounting; the non-GAAP move from $3.04 to $3.98 (+30.8%) is the real one.

6.2 Cash generation — the strongest part of the file

Free cash flow has grown 5.5x in three years, from $104.4M in FY2022 to $579.0M in FY2025 — a 19.7% FCF margin. Q1 2026 continued the trend with $202.4M of operating cash flow against $39.6M of capex. Cash conversion is genuine: FY2025 OCF/net income of 1.40x reflects D&A of $276.8M (predominantly non-cash merger intangible amortisation) exceeding capex of $174.4M. Because the $117.7M bargain purchase gain was a non-cash credit to earnings, cash earnings are in fact stronger relative to accrual earnings than the ratio suggests.

This is the most important counterweight to the earnings-quality concerns. Whatever one thinks of GAAP net income, $579M of free cash flow on a $10.33bn market capitalisation is a 5.6% FCF yield, and it is not an accounting construct.

6.3 Balance sheet

• Cash and marketable securities: $799.3M at 2026-03-31 (vs. $629.1M at 2025-12-31 and $956.2M at 2024-12-31).
• Total debt: $115.2M, essentially all finance-lease obligations. The $450M of NuVasive-assumed convertible notes were fully repaid in cash in March 2025. GMED carries no funded debt.
• Net cash: ~$684M.
• Total equity: $4,573.3M at 2025-12-31; book value per share $34.24.
• Goodwill $1,435.0M + other intangibles $745.1M = $2,180.1M, or 47.7% of equity and 41.1% of total assets. Tangible book equity is approximately $2.39bn, or ~$17.70 per share. At $76.49 the stock trades at 2.23x stated book and ~4.3x tangible book.
• Current ratio 4.26x. Liquidity is not a question.

The intangible balance is the balance-sheet expression of the capital-allocation problem. Nearly half of GMED’s stated equity is the price it paid for NuVasive and Nevro, not assets that generate cash. A goodwill impairment is a live risk if the spine outgrowth fades (see Section 9), and it would be a non-cash confirmation that the merger overpaid.

Working capital is heavy and permanent. Inventory of $759.3M against $957.8M of COGS is ~289 days; the cash conversion cycle is 317 days. Receivables of $678.9M on $2,938.9M of revenue are ~84 days. This is structurally normal for consigned-set spine but it is a genuine, permanent economic cost that the income statement does not charge for. Every dollar of incremental spine revenue requires roughly a dollar of incremental working capital plus set capex before it converts.

6.4 Returns on capital — the verdict-determining metric

Hand-verification: EBIT of $537.2M taxed at a normalised 24% gives NOPAT of $408M; invested capital of equity $4,573.3M plus debt $118.7M less cash and securities $557.2M equals $4,135M; ROIC = 9.9%. This corroborates ROIC.ai’s 10.12%.

Excluding goodwill and intangibles, invested capital falls to approximately $1,955M and ROIC rises to approximately 20.9%. The operating business earns excellent returns. The acquisition price is what drags the consolidated figure to ~10%.

That number — 10% — is roughly at, not above, a reasonable cost of capital for a mid-cap medical device company with a 0.93 beta. Three years and 36% dilution after the NuVasive merger, Globus earns approximately the same return on capital it earned as a standalone company in 2022. It is a bigger business, a more diversified business and a better-positioned business. It is not a higher-returning one.

Data integrity note. ROIC.ai’s reported FY2025 return_com_eqy of 42.33% and return_on_cap of 33.65% are arithmetically impossible against $537.9M of net income on $4,573.3M of equity (11.8%), and its bs_tot_cap of $1,506.7M contradicts the balance sheet. Its bs_sh_out of 112,625,126 also contradicts the 10-K’s 135,055,223 shares outstanding at 2025-12-31. These four fields were rejected; every figure in this memo is reconciled to the filing.

Verdict: economics improve with scale at the operating level but not at the capital level — and the distinction is the whole argument. Gross margin is genuinely expanding (six consecutive quarters), operating leverage is genuine (SG&A guided to 38–39% of sales from 40.1%), free cash flow has grown 5.5x in three years to a 19.7% margin, and the balance sheet is net cash and debt-free. Those are the marks of a quality business improving. But ROIC of ~10% is barely at cost of capital, the improvement is recovery to rather than advance beyond the pre-merger level, nearly half of book equity is acquisition goodwill and intangibles, working capital consumes ~317 days of cash, and FY2025 GAAP earnings are ~30% non-recurring. This is a good operating business that has been made mediocre as a capital-allocation vehicle.

7. Capital Allocation

7.1 The NuVasive merger — the decision that defines the company

Announced 2023-02-09, closed 2023-09-01. All-stock: each NuVasive share converted into 0.75 Globus Class A shares. Share count went from 100.2M (2022-12-31) to 136.3M (2023-12-31) — approximately 36% dilution to legacy Globus holders. The market’s immediate verdict was -18.2% on the announcement day and a drawdown to $43.79 by November 2023, -43% from the pre-announcement price.

On management’s own scorecard, the integration succeeded. The stated targets were mid-to-high-single-digit sales growth and a mid-30s adjusted EBITDA margin by the end of year three. Both were substantially delivered on the base business: H2 2025 base organic growth of 8.8%, Q4 2025 base-business adjusted EBITDA of 35.7% and FY2025 of 33.4%. Cost synergies were captured — R&D fell from 6.5% to 5.0% of sales on headcount rationalisation, SG&A leverage emerged, and manufacturing was consolidated into Globus plants with demonstrated output gains (PRECICE now exceeds former NuVasive facility output).

But hitting a synergy target and creating shareholder value are different tests. Consolidated ROIC fell from ~10.3% (FY2022) to 4.7% and 4.0% in FY2023–24, recovering to ~10.1% in FY2025. Three years in, returns on total capital are back to the pre-merger level on a business with 36% more shares outstanding. The generous reading is that GMED bought scale, diversification and a wider product portfolio at a price that has so far proved return-neutral, with the margin bridge still to come. The unkind reading is that legacy holders financed a doubling of revenue and received, so far, the same return on capital they already had. Both readings are defensible. The tie-breaker will be whether the mid-70s gross margin lands — if it does, the merger becomes value-creative in retrospect; if margin stalls at 69–70%, it will have been a large, well-executed, value-neutral transaction.

7.2 The Nevro acquisition — cheap, well-timed, and currently shrinking

Closed 2025-04-03 for $252.5M in cash. The purchase produced a $117.7M non-taxable bargain purchase gain — GMED paid materially less than the fair value of the net assets acquired, the accounting signature of buying a distressed asset. Nevro contributed $293.6M of revenue in nine months of FY2025, implying an annualised purchase multiple near 0.6x sales. On price, this was opportunistic and good.

On execution, the record is genuinely mixed. The cost side has been outstanding: Nevro was made EPS-accretive within nine months, beating the initial 24-month guidance by 15 months, with standalone adjusted EBITDA margin moving from 16.2% (Q3 2025) to 21.2% (Q4 2025). The revenue side has not: Nevro fell from $99.7M in Q4 2025 to $82.7M in Q1 2026 (-17.1% sequentially), and CFO Kline expects it “will probably get a little bit worse before it gets better.” Crucially, management attributes the decline to its own actions — “the structural changes made within sales and marketing of the Nevro business at the tail end of 2025.” Nevro SG&A still runs 49.9–55.7% of Nevro sales.

The candid interpretation: GMED bought a cheap asset in a declining industry-wide category, cut it aggressively to profitability, and in doing so removed enough commercial capacity to accelerate the revenue decline. Management is now rebuilding the sales force it cut and expects a return toward historical run-rate “late in the second half” of 2026. That is a hypothesis, not evidence. The downside is bounded by a $252.5M purchase price; the upside requires a turnaround that has not yet begun.

7.3 Buybacks and dilution management

A $500M authorisation was announced in Q2 2025, of which $110M was used in 2025, leaving $390M remaining at 2026-03-31. Management states that since FY2022 it has deployed >$600M to repurchase >10M shares at an average price “under $60 a share,” offsetting “>25% of the dilution that was created from the NuVasive merger.”

The execution was good; the effect was modest. Repurchasing below $60 against a current $76.49 was value-accretive on its face. But look at the share count: 136.3M (2023) → 137.4M (2024) → 135.1M (2025). Despite $386M of repurchases across 2024–25, net shares outstanding fell by just 1.2M, because ~$50M/year of stock-based compensation and option exercises offset most of the buying. Buybacks at GMED are functionally dilution management, not per-share value creation. Investors should not model buyback-driven EPS accretion.

No dividend has ever been paid.

7.4 Capital intensity and internal investment

Management’s stated priority order is: (1) internal product development, (2) capital spending on surgical sets and manufacturing footprint, (3) share repurchase, (4) complementary M&A. Capital expenditure was $174.4M (5.9% of sales) in FY2025 and is guided to 5–6% of sales in 2026. R&D was 5.0% of sales in FY2025 and is guided up to 5–6% in 2026, with management explicitly “doubling down in spine R&D and enabling tech R&D.”

The R&D increase deserves credit. R&D fell from 10.2% of sales (FY2021) to 5.0% (FY2025) — much of that is legitimate synergy capture and scale leverage, but a portion is under-investment, and the ExcelsiusGPS platform is nine years old in a market with new entrants. Raising R&D is the correct decision even though it caps near-term margin expansion.

7.5 Incentive alignment — the weakest link

The annual cash bonus for named executive officers is 90% determined by a single metric: revenue. The remaining 10% is a discretionary Compliance & Quality assessment by the Compensation Committee. Achievement of 75% or less of the revenue goal pays zero; 100% pays 110% of the base reference amount; 110% pays 145%. There is no margin gate, no EPS gate, no cash-flow gate, and no return-on-capital gate. For FY2025, revenue of ~$2.65bn excluding Nevro was 98% of goal, producing a 102% payout.

Long-term equity for NEOs in 2025 was granted as stock options, not performance share units with operating conditions: David C. Paul 75,000; Daniel Scavilla 140,000 (grant-date fair value $5,006,645); Kelly Huller 40,000 — all struck at $92.42 on 2025-01-23 and all currently underwater at $76.49.

This is the structural explanation for the NuVasive outcome. A compensation scheme that pays on revenue and nothing else rewards management for buying revenue with equity, because the numerator rises and the denominator is not measured. The 36% dilution that halved return on invested capital did not reduce a single executive’s bonus; the revenue it purchased increased them. Three years and one CEO later, the metric has not changed. For an investor, this is the single most important unfixed problem at Globus Medical, and it is a live risk to the next acquisition, not merely a critique of the last one.

7.6 Governance and control

GMED is a “controlled company” under NYSE Rule 303A.00 because founder and Executive Chairman David C. Paul holds more than 50% of the voting power through the dual-class structure. As a controlled company GMED may elect not to comply with requirements that a majority of the board be independent and that director nominations be made by independent directors. At 2025-12-31, executive officers and directors beneficially owned approximately 66.3% of voting power on 135,055,223 Class A and Class B shares outstanding, and a further 192,602,552 Class B shares remained available for issuance — an amount exceeding 5% of outstanding common stock, which the 10-K notes the board could issue without necessarily triggering the charter’s automatic conversion provision. Neither Mr. Paul nor CEO Mr. Pfeil is an independent director. The 2026 annual meeting (2026-06-03) approved an increase in shares authorised under the 2021 Equity Incentive Plan.

Two abrupt senior departures in twenty months warrant note. On 2025-07-18, President and CEO Daniel T. Scavilla resigned from both the executive role and the Board, effective immediately, disclosed in the same 8-K that furnished preliminary Q2 2025 sales results — an unusual pairing. The filing carries the standard “not due to any disagreement” language and no successor search or transition period was disclosed; then-COO/CFO Keith Pfeil, 46, was appointed the same day. Scavilla had received a $5.0M option grant six months earlier. On 2026-03-23, director John A. DeFord, Ph.D. resigned from the Board effective immediately, with a separate press release issued. Neither departure has been explained beyond boilerplate.

7.7 The insider record

Across all 84 Form 4 and 4/A filings since the NuVasive close (2023-09-01 to 2026-07-26), open-market activity totals $50,194,615 of sales and $0 of purchases. There has not been a single code-P open-market purchase by any officer or director in nearly three years — a window that included a drawdown to $43.79 and a subsequent recovery to $101.

Two features make this worse than the aggregate. First, the timing. Scavilla’s largest block — 130,000 shares for $11.05M — was sold at $85.00 on 2024-11-21, near a local peak, eight months before he resigned. CFO Kyle Kline sold at $90.00 (2025-12-01) and at $101.10 on 2026-01-08; director Leslie Norwalk also sold at $101.10 that same day. GMED’s five-year intraday high is $101.3975, set on 2026-01-08. The CFO and a director sold at the top tick, and the stock is 24% lower today. EVP/GC Kelly Huller sold 20,000 shares at $94.50 on 2026-02-25 and has sold in nine separate windows since 2024. Second, the pattern. Non-derivative transaction codes since the merger are S 634,068 shares against M 573,595 (option exercise) — the dominant behaviour is exercise-and-immediately-sell, not accumulate.

The one qualification: David C. Paul, the controlling holder, does not appear as a seller. His Class B control block is intact, and the family’s economic exposure remains very large. That is a genuine offset to the aggregate signal — the person with the most at stake has not sold.

Verdict: capable operators, weak capital allocators, poor governance. Management executed both integrations well — NuVasive synergies were captured, Nevro was made accretive 15 months early, manufacturing was consolidated with demonstrable output gains, the balance sheet was deleveraged to net cash, and buybacks were executed below $60. That is a real record of operational competence. But the NuVasive merger diluted legacy holders 36% for a return on capital that is only now back to where it started; buybacks have offset dilution rather than created per-share value; the annual bonus remains 90% revenue-weighted with no return metric; the company is founder-controlled with a nearly two-thirds voting block and 192.6M authorised Class B shares in reserve; two senior figures have departed abruptly in twenty months; and insiders have sold $50.2M with zero purchases, twice at the exact five-year high. Has management allocated capital intelligently? Competently, yes. Intelligently, not yet — and the incentive structure that produced the problem is unchanged.

8. Changes and Headwinds — Last Two Years

Acquisitions and portfolio.

• 2025-04-03: Nevro Corp. acquired for $252.5M cash, adding the HFX/Senza spinal cord stimulation platform and minimally-invasive sacroiliac joint treatment. Generated a $117.7M non-taxable bargain purchase gain. Subject to a subsequent measurement-period restatement adjustment to the purchase accounting.
• 2024: share acquisition of a biotechnology company focused on hemostasis R&D (Fibriant B.V.), with good-manufacturing-process and FDA-approval milestone consideration; plus a series of individually immaterial business acquisitions. FY2024 R&D included $12.6M of acquired in-process R&D.
• 2023-09-01: NuVasive merger closed — the transaction that reshaped the company.

Leadership and board.

• 2025-07-18: CEO Daniel T. Scavilla resigned from the CEO role and the Board, effective immediately. Keith W. Pfeil, then COO and CFO, appointed President & CEO the same day. Kyle Kline subsequently became CFO. Stock -7.9% on 2025-07-21.
• 2026-03-23: Director John A. DeFord, Ph.D. resigned from the Board, effective immediately.
• 2026-06-03: shareholders approved an increase in shares authorised under the 2021 Equity Incentive Plan.

Balance sheet and capital.

• March 2025: the remaining $450M of NuVasive-assumed convertible notes repaid in cash. GMED became debt-free.
• Q2 2025: new $500M share repurchase authorisation; $110M used in 2025, $390M remaining.
• FY2025: $300.5M deployed repurchasing ~4.3M shares at an average ~$70.

Operational.

• H1 2025 supply-chain failure — GMED could not deliver surgical sets and consigned inventory to the field at the required pace, damaging both US and international spine. This drove the Q1 2025 miss (-23.0% on 2025-05-09). Management states the problem is resolved and Q1 2026 results support that.
• Six consecutive quarters of sequential adjusted gross-margin expansion through Q4 2025, extended by a flat-sequential 69.2% in Q1 2026.
• PRECICE limb-lengthening manufacturing transferred from former NuVasive facilities to Globus in early 2025; output now exceeds the historical facility.
• Strategic shift in Enabling Technologies from outright capital sale toward leases and rentals, announced across the Q3 2025 – Q1 2026 calls. Suppresses recognised revenue; intended to accelerate implant pull-through.
• Q2 2026 FDA 510(k) clearances for the Scripps patient-specific lumbar interbody spacer system (seven systems) and Scripps patient-specific rods.
• 30+ new products launched since FY2022; six spine launches in 2025, four in Q4 alone (CREO Traction, Reline 3D Towers, AMS Freehand, HEDRON C-MIS).

Litigation.

• Moskowitz Family LLC (a non-practising entity) sued in 2019 alleging infringement of six patents across a very broad swath of the core spine portfolio (COALITION MIS, CORBEL, MAGNIFY-S, HEDRON IA, INDEPENDENCE MIS, FORTIFY, XPAND, SABLE, RISE, ELSA, ALTERA, ARIEL, CALIBER and others). A jury returned a defence verdict in favour of Globus on 2023-12-14; Moskowitz appealed on 2024-09-30. No liability recorded beyond counsel fees.
• 4WEB LLC sued NuVasive in 2023 alleging infringement of eleven patents by the Modulus product line; transferred to the Southern District of California in May 2024; ongoing.
• Pimenta litigation — Dr. Luiz Pimenta sued NuVasive in 2018 for breach of a Clinical Advisor Agreement, seeking $97M (later reduced to $82M) in unpaid royalties; NuVasive cross-claimed.
• Q4 2025 included $13.4M of one-time net charges for estimated litigation, adjusted out of non-GAAP results.

Headwinds carried into FY2026.

1. Nevro revenue decline, self-inflicted and expected by management to worsen before improving.
2. Enabling Technologies revenue suppression from the deliberate lease/rental mix shift — with consensus modelling a range of ~$104M to ~$150M and management refusing to guide.
3. Rising R&D to 5–6% of sales, capping margin expansion.
4. Intensifying robotics competition — new entrants have forced hospital bake-offs and elongated deal cycles.
5. CMS WISeR prior-authorisation pilot potentially covering spinal cord stimulation.
6. Chronic ~1% ASP erosion in spine.

Verdict: on balance these changes strengthen the operating thesis and weaken the governance thesis. The balance sheet is transformed (debt-free, net cash, $390M of buyback authorisation remaining), the manufacturing consolidation is delivering measurable results, the H1 2025 supply failure is behind the company, the margin trend is unambiguously positive, and Nevro was bought below the fair value of its net assets. Against that: the CEO left abruptly, a director left abruptly, the acquired neuromodulation business is shrinking under GMED’s management, the robot franchise is being deliberately de-monetised, and the compensation structure that produced the dilutive merger is unchanged. The business got better over these two years. The stewardship did not.

9. Risk Analysis

Concentration of risk. The material risks cluster tightly: risks 1, 2, 5 and 6 are all expressions of the same question — is the current operating momentum durable or is it a harvest? Risk 3 is independent and structural. Everything else is second-order. An investor in GMED is fundamentally underwriting two things: that spine share gains persist long enough for the margin bridge to land, and that management does not spend the resulting cash on another dilutive acquisition.

10. Valuation Discussion

No price target and no recommendation appear in this section. The analysis below describes what the current price implies and what would have to be true to justify it.

10.1 Where the stock trades

10.2 Own-history context — the single highest-signal datum

The AZI valuation index measures each multiple against the stock’s own multi-year range:

GMED has traded more expensively than this on roughly 94% of the days in its recent history, and on earnings, on roughly 99% of them.

Two caveats discipline this. First, it is own-history context only — it says nothing about whether GMED is cheap relative to other medical device companies, and a business whose ROIC fell from 13.6% to ~10% should trade at a lower multiple than its own history. A percentile low can be a permanent re-rating rather than an opportunity. Second, the P/E percentile alone would be suspect here given the FY2025 bargain purchase gain and the depressed FY2024 base. But P/B at the 10.7th percentile and P/S at the 5.3rd corroborate independently, and neither is affected by the tax or bargain-purchase distortions. Three metrics agreeing is meaningfully stronger evidence than one.

The corroborating external datapoint: sell-side price targets have been falling toward the price rather than the reverse — Truist cut to $100 from $115 (2026-07-16), RBC Capital to $115 from $120 (2026-07-13) — yet both remain 31–50% above the market. The gap between analyst targets and the traded price is itself a statement that the marginal buyer disagrees with the sell-side models.

10.3 Embedded expectations — what the market is underwriting

At ~$9.65bn of enterprise value against ~$1,040M of FY2026E adjusted EBITDA (9.3x) and 16.1x the midpoint of management’s non-GAAP EPS guidance, the market is pricing approximately:

• Revenue compounding at the low end of ambition — call it 5–6% rather than management’s “high single-digit” aspiration. Given FY2025 base-business growth of 5.0%, this is not an unreasonable prior.
• Adjusted gross margin stalling at the guided 69–70% and never progressing toward the mid-70s target. This is the crux.
• Nevro stabilising but not recovering — no credit for a turnaround.
• Enabling Technologies remaining a low-single-digit-percentage, non-differentiating revenue line.
• No re-rating of the ~10% ROIC, and by implication a permanent governance/control discount.

Put positively: the market is paying for the business as it exists today and nothing for the bridge management says it is building.

10.4 The margin bridge — the whole argument

Moving adjusted gross margin from ~69.5% (the FY2026 guided range midpoint) to a mid-70s ~74.5% on ~$3.4bn of revenue produces approximately $170M of incremental gross profit. Taxed at 24% and spread over ~137M diluted shares, that is approximately $0.94 of incremental EPS — roughly 20% of the current FY2026E base of $4.75.

Arguments that it lands. It is a restoration, not an invention — standalone Globus earned 72–78% gross margins every year from 2016 to 2022, and legacy-Globus standalone adjusted gross margin was already 69.3% in Q1 2026. Six consecutive quarters of sequential expansion is a track record, not a promise. The mechanism is specific and verifiable: migrating NuVasive and Nevro production into Globus plants, where PRECICE output has already demonstrably exceeded the legacy facility. And management has now raised EPS guidance three times off exactly this driver, most recently by $0.30 on 2026-05-07.

Arguments that it does not. The FY2025 full-year adjusted improvement was only 70bps, and at that pace mid-70s is six to seven years away. Management explicitly refuses a timeline (“about the same cadence”). Rising R&D to 5–6% of sales consumes part of the benefit at the operating line. Nevro dilutes mix. And ~1%/yr price erosion is a permanent drag that must be out-run before any of the cost savings show up.

10.5 Scenario analysis

Each scenario is before accumulated net cash, which compounds at ~$500–600M/year of free cash flow less buyback.

The distribution is asymmetric to the upside, and the asymmetry comes from the balance sheet as much as the business. The bear case is only about -8% from $76.49, because net cash of ~$684M and a 5.6% free-cash-flow yield put a genuine floor under a debt-free, profitable, diversified business. The bull case is +50%. But note the base case: roughly +20% over two-and-a-half years, or ~8%/year — an unremarkable return for accepting a ~10% ROIC, a controlled-company structure, a revenue-only bonus and a documented pattern of insider selling. The upside is real but it is concentrated in the bull scenario, and the bull scenario requires the margin bridge to be a schedule rather than a slogan.

10.6 What multiple is appropriate

A genuinely open question. GMED’s factor-similar peer set — NVST, BAX, STE, MMSI, ITGR, COO, SOLV, LMAT, GEHC — trades broadly in the 13–20x forward earnings range, and it is notable that the model does not pair GMED with SYK, MDT, ZBH or ISRG. The market has already classified GMED as a mid-cap value device name rather than a robotics compounder, and 16x forward earnings sits squarely inside that cohort. Against that classification GMED is fairly priced, not cheap; it is only cheap against its own history and against the sell-side’s robotics framing.

A governance discount of 10–15% relative to otherwise-comparable device names is defensible for a controlled company with a revenue-only bonus, and it may be permanent — control does not expire. Investors should not underwrite multiple expansion driven by governance improvement.

Verdict on valuation: cheap on its own history, fairly valued against its actual business quality, and asymmetric in the buyer’s favour — but the base case is ordinary. The stock is priced for the business it is, with a free option on the business management says it is building. Whether that option is worth owning depends entirely on one’s confidence in the gross-margin bridge, and reasonable people can differ.

11. Variant Perception

11.1 The consensus belief

Sell-side consensus, as expressed in $100–115 price targets against a $76.49 price, holds that GMED is a surgical robotics and enabling-technology platform whose ExcelsiusGPS installed base drives durable implant pull-through, that NuVasive synergies plus manufacturing consolidation will deliver mid-70s gross margins, and that ~10% EPS growth deserves a mid-to-high-teens multiple. Consensus models roughly $150M of FY2026 Enabling Technologies revenue.

The market price says something different. Investors are treating GMED as a mid-cap value medical device company with acquisition-driven growth, questionable capital allocation and a control overhang — hence a 5.7th-percentile own-history valuation and a factor model that groups it with Envista, Baxter, Steris and Integer rather than with Stryker or Intuitive.

11.2 The strongest bull case

Spine’s capital cycle has turned, and GMED is the only committed pure-play left standing. Stryker sold its spine business. Zimmer Biomet exited twice. ZimVie divested spine. Medtronic treats it as a growth-capped legacy category. Capital and management attention are leaving the industry precisely as GMED reaches the scale to serve every account with a single-vendor portfolio spanning implants, biologics, neuromonitoring, robotics and now neuromodulation. The result is visible in the numbers: 58 consecutive weeks of growth, US Spine ~500bps above market for four straight quarters, and trauma compounding at 26–34% off a manufacturing base GMED now controls.

Layered on top is a verifiable, mechanical margin bridge. Standalone Globus earned 72–78% gross margins for seven years; the mid-70s target is restoration, not invention; six consecutive quarters of sequential expansion prove the machinery works; and the bridge is worth ~$0.94 of EPS, or ~20% of current earnings, for which the market pays nothing. Meanwhile the balance sheet is debt-free with $684M of net cash, free cash flow has grown 5.5x in three years to $579M (a 5.6% yield), $390M of buyback authorisation remains, and the stock sits at the 1.2nd percentile of its own ten-year P/E range. You are buying an improving business at the cheapest price in its public life.

11.3 The strongest bear case

You are paying 16x forward earnings for a ~10% ROIC price-taker in a structurally deflationary industry, run by people whose bonus is 90% revenue and who have sold $50M of stock without buying a share.

The robotics narrative — the entire justification for a premium multiple — is false. Enabling Technologies is 4.8% of revenue and fell 8.4% in FY2025, and management is now deliberately giving robots away on lease to protect implants. Strip the narrative and GMED is a spine hardware company facing ~1% annual ASP erosion, whose CEO says the path to higher margin comes from cost, not price — the definition of a business without pricing power.

The growth is bought, not earned. FY2025 base-business growth was 5.0%; the 16.7% headline was Nevro. And Nevro, the growth engine, is now shrinking 17.1% sequentially because GMED cut its sales force. FY2025 GAAP earnings are ~30% non-recurring — a $117.7M bargain purchase gain and a $46.3M tax valuation-allowance release — and the apparent 870bps of GAAP gross-margin expansion is ~90% purchase-accounting roll-off against 70bps of real improvement.

The capital allocation is worse than the operations. NuVasive diluted legacy holders 36% and, three years on, ROIC is merely back to where it started. Buybacks have offset dilution rather than created value — net share count fell 1.2M despite $386M of repurchases. Nearly half of book equity is goodwill and intangibles. And the incentive that produced all of this — a revenue-only cash bonus with no return-on-capital gate — is unchanged, in a controlled company where the founder holds >50% of the votes, 192.6M Class B shares sit authorised, the CEO resigned with no notice in July 2025, a director resigned with no notice in March 2026, and the CFO sold at $101.10 on the day of the five-year high.

The market is not wrong. It is pricing exactly this.

11.4 The 3–5 assumptions that actually matter

1. Does US Spine outgrowth persist beyond the harvest? Everything else is second-order. If ~10% growth is competitor dislocation being consumed, it decelerates to 3–4% within 18 months and the multiple compresses.
2. Does adjusted gross margin cross ~71–72% within four to six quarters? This is the falsifiable checkpoint on the mid-70s bridge, and it is worth ~20% of EPS.
3. Does Nevro stabilise in H2 2026 as management guides? Management has already conceded it gets worse first. If it is still declining entering 2027, the acquisition thesis is broken and management’s operational credibility takes real damage.
4. Does management do another dilutive acquisition? M&A remains an explicit stated capital priority, the bonus still pays on revenue, and the controlled structure means no one can stop it.
5. Does Enabling Technologies revenue decline again in FY2026? The revenue impact is small; the narrative impact is large, because it would strip the last support from the premium-multiple case.

11.5 The factor-positioning read

The empirical positioning evidence sharpens the variant perception rather than merely decorating it. GMED carries zero momentum loading — the factor is L1-zeroed in all four nested models — alongside a mild Value tilt (+0.10 to +0.12), a SmallSize tilt (+0.17 to +0.37) and a positive DividendYield loading (+0.36 to +0.52) that is striking for a company which has never paid a dividend. That combination says GMED trades with the defensive-value cohort, not with growth or with robotics. Model R² is only ~20% against 45.5% annualised idiosyncratic volatility: roughly 80% of GMED’s return variance is stock-specific. This is not a crowded trade and it is not a factor bet — it is a pure execution bet, and it will be resolved by results, not by regime.

Consensus is therefore not offside in the sense of being long a crowded name. Consensus is offside in a subtler and more interesting way: the sell-side is modelling a robotics platform while the marginal price-setter has already re-classified the stock as mid-cap value. The $30–40 gap between analyst targets and the traded price is the measure of that disagreement. The most diagnostic evidence for who is winning that argument is 2026-05-08 — GMED beat on revenue, raised full-year EPS guidance by $0.30, and fell 8.4%. The market rejected the composition: Nevro shrinking, Enabling Technologies being de-monetised, and a top line pointedly not raised. Beats-and-raises are not sold by investors who believe the story.

Where I think the market is most likely wrong: it is extrapolating the quality verdict (correctly negative) into the value verdict (possibly too negative). A ~10% ROIC, controlled, price-taking spine company genuinely does not deserve 25x earnings. But it probably deserves more than the cheapest price in its own history, given a debt-free balance sheet, a 5.6% free-cash-flow yield, a demonstrated six-quarter margin trend and an industry whose competitors are walking away.

12. Fact vs. Interpretation

13. Open Questions

1. What is FY2026 Enabling Technologies revenue? Management refuses to guide the line. Consensus spans ~$104M to ~$150M — a 44% range on a line item that carries the entire premium-multiple narrative. Directly asked to reconcile, CFO Kline said only “we are not going to break out our guidance into the different parts and pieces.”
2. Why did CEO Daniel Scavilla resign from both the CEO role and the Board effective immediately on 2025-07-18, six months after receiving a $5.0M option grant, disclosed in the same 8-K as preliminary Q2 sales results? And why did director John A. DeFord resign effective immediately on 2026-03-23 with a dedicated press release? Neither is explained beyond boilerplate.
3. What is the actual timeline to mid-70s adjusted gross margin? Management will say only “about the same cadence.” At the FY2025 full-year pace of 70bps/year it is six-plus years; at the recent quarterly pace it is two to three. The difference is worth several turns of multiple.
4. What was Nevro’s revenue in Q1 2025 (pre-acquisition)? Management explicitly declined to disclose it, so investors cannot compute a true year-over-year decline rate for the acquired business. This is a meaningful transparency gap on the segment currently causing the most damage.
5. Does spinal cord stimulation fall under the CMS WISeR prior-authorisation pilot, and what is the exposure? The CEO said he had “not seen” an impact and offered no further comment. Unresolved.
6. How much of US Spine’s ~500bps of outgrowth is competitor dislocation versus durable share capture? GMED does not disclose the split between competitive rep recruiting, cross-selling into the NuVasive base, and genuine new-account wins. This is the single most important unanswered question in the file.
7. Will the compensation committee add a margin, EPS, cash-flow or return-on-capital gate to the annual bonus? The 2026 proxy shows no such change. Given the controlled structure, shareholder pressure is unlikely to force one.
8. What is the exposure on the Moskowitz appeal and the 4WEB litigation? GMED won a defence verdict at trial in December 2023 but Moskowitz has appealed on patents reading across a very broad swath of the core interbody portfolio, and no liability is accrued beyond counsel fees — an adverse outcome would be entirely unreserved.
9. What is the return on the ~$174M/year of capital expenditure on surgical sets? Management describes set investment as “high-return capital investments” but discloses no utilisation or payback data. Given 289 days of inventory and a 317-day cash conversion cycle, this is a material unmeasured claim.
10. What is management’s M&A appetite now that the balance sheet is net cash with $684M? M&A is explicitly the fourth capital priority, the bonus pays on revenue, and no shareholder can vote it down.

14. What Must Be True

14.1 For the bull case

14.2 For the bear case

Synthesis. The bull and bear cases are unusually cleanly separable here, because both turn on the same two publicly-reported quarterly series: adjusted gross margin and US Spine growth. An investor does not need private information or channel checks to adjudicate this thesis — they need to watch two numbers for four to six quarters. If adjusted gross margin crosses 71% while US Spine holds ≥7%, the bull case is confirmed and the current multiple is too low. If gross margin stalls at 69–70% while US Spine decelerates below 5%, the bear case is confirmed and even 16x is too much for a ~10%-ROIC controlled price-taker. The falsification is fast, cheap and public — which is itself a reason to find this situation interesting.

15. Source Appendix

The full, itemised source appendix — every filing, transcript, dataset and URL relied upon, with access dates — is reproduced as Appendix B below.

Primary sources (highest authority). Globus Medical FY2025 Form 10-K (filed 2026-02-24); FY2024, FY2023, FY2022 and FY2021 Forms 10-K; Q1 2026 Form 10-Q (filed 2026-05-07) and the trailing 10-Q series; DEF 14A filed 2026-04-24 and the four prior proxies; the 45-filing 8-K series 2021–2026 (notably 2023-02-09 NuVasive announcement, 2025-04-03 Nevro close, 2025-07-21 CEO transition, 2026-03-23 director resignation, 2026-06-04 annual meeting); the S-4/S-4A registration statements for the NuVasive merger; and all 125 Form 3/4/4/A insider filings. All are publicly available on SEC EDGAR.

Management commentary (treated as hypothesis, not evidence). Q4/FY2025 earnings call, 2026-02-24, and Q1 2026 earnings call, 2026-05-07 — both publicly webcast and archived on the company’s investor relations site.

Quantitative data. ROIC.ai (income statement, balance sheet, cash flow, profitability ratios, enterprise value — with four FY2025 fields rejected as erroneous and documented in Section 6.4); AZI daily price history and the AZI valuation_index own-history percentiles; the FactorsToday factor model (stock loadings, leaderboard, stock info, specific volatility, related stocks).

Comparative context. A dated third-party US healthcare industry primer (Morgan Stanley, 2011) was consulted for value-chain and reimbursement framing only. The author’s prior published notes on Medtronic, Stryker, Zimmer Biomet, Intuitive Surgical, Boston Scientific, GE HealthCare and Steris were used for peer and industry cross-read; all underlying facts in those comparisons derive from the respective companies’ own public filings and earnings calls.

Every claim in this article traces to a public primary source listed in Appendix B.

APPENDIX A — Standard Diligence Questionnaire

Globus Medical, Inc. (NYSE: GMED) · 2026-07-26

Supplemental to the main article, and grounded in the same public sources. Labels: [F] Fact · [I] Interpretation · [A] Assumption.

General

What thoughtful questions have other investors asked about this company?

The questions asked on the Q4 2025 and Q1 2026 calls are unusually revealing, because analysts are circling exactly the weak points identified in this memo.

• “Are there any incremental headwinds to revenue we should be thinking about… or is this conservatism at this point?” (Ross Osborne for Larry Biegelsen, Wells Fargo) — [I] Analysts do not believe the reaffirmed top line, because EPS was raised and revenue was not.
• “I see a consensus enabling tech around $150 million for 2026. We are at $104 million. Is there any comment you could give on either of those two numbers?” (Richard Newitter, Truist) — [F] A 44% dispersion in consensus on the line item that carries the entire premium-multiple narrative. Management declined to reconcile.
• “Is the gross margin opportunity purely a cost exercise from your seat, or do you need price to drive your gross margins higher, and can you do that in spine in such a competitive market?” (Brian Zimmerman, BTIG) — [I] The sharpest question asked. It goes directly to whether GMED has pricing power. The answer was no: the CEO conceded ~1% annual price erosion and said the mid-70s path “majority… really comes from our ability to drive costs.”
• “Does Nevro get worse before it gets better?” (Brian Zimmerman, BTIG) — CFO Kline: “My expectation is likely that it will probably get a little bit worse before it gets better.” [F]
• “We did not have last year — they actually never reported Q1 2025 — wondering if you have those numbers and can give them out just so we have some basis for comparison.” — Management refused. [F] Investors cannot compute Nevro’s true year-over-year decline.
• “Can you be a double-digit grower over the next few years?” (David Saxon, Needham) — CFO: “we strive to be that high single-digit top-line grower, and we look to go above and beyond that on the bottom line.” [F]
• “I believe spinal cord stimulation might be captured under the pilot WISeR program that CMS has in place.” (Richard Newitter, Truist) — CEO: “I have really no comment. Because I have not seen it.” [F]
• “Any shift in competitive dynamics… from the traditional larger players in spine?” (Matt Miksic, Barclays) — CEO conceded deal cycles have “elongated a bit because hospitals are now requiring everyone to go back and look at all the competitive offerings.” [F]

The question the sell-side is not asking, and should be: why is 90% of the executive cash bonus determined by revenue alone, and why has that not changed after a merger that diluted holders 36% for no improvement in return on capital?

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? [I] Neither — they are at a post-integration normalisation point, which is a different thing and easy to misread. FY2024 non-GAAP EPS of $3.04 was an artificial trough depressed by peak purchase accounting; FY2025’s $3.98 (+30.8%) is the first clean year. On the GAAP line, FY2025’s $3.92 is an artificial peak — ~30.5% of net income was non-recurring ($117.7M bargain purchase gain plus a $46.3M tax valuation-allowance release). [F] The honest read is that FY2025 non-GAAP EPS of $3.98 is close to a true underlying run-rate, with margin still below its structural potential (68.1% adjusted gross margin versus a 72–78% standalone history) and R&D running below normal (5.0% versus a guided 5–6%).

Driven by the external environment or internal actions? [F] Overwhelmingly internal. The three largest earnings swings of the past three years were all self-generated: purchase-accounting amortisation from the NuVasive merger, the H1 2025 supply-chain failure to deliver surgical sets to the field, and the deliberate restructuring of Nevro’s sales organisation that caused a 17.1% sequential revenue decline. [F] Corroborating this quantitatively: ~80% of GMED’s return variance is idiosyncratic (factor-model R² ~20%, idiosyncratic volatility 45.5% annualised). This is an execution story, not a cycle story.

How stable are revenues? [I] Structurally stable at the demand level, unstable at the reported level. ~95% of revenue is consumable implants, disposables, biologics and fee-for-service neuromonitoring that recur with procedure volume; there is no contractual recurring revenue and no subscription base. Procedure demand is largely non-discretionary and demographically driven — the one demonstrated shock was COVID (FY2020 revenue +0.5%). [F] Reported revenue has nonetheless been extremely unstable (+53.4%, +60.6%, +16.7% in FY2023–25) because it has been reshaped by acquisitions three times in three years.

Outlook for products/services? [F] FY2026 guidance is $3.18–3.22bn (+8.2% to +9.6%) with non-GAAP EPS of $4.70–4.80 (+18.1% to +20.6%). Within that: US Spine growing ~10%, trauma 26–34%, neuromonitoring >30%, international recovering, Nevro declining and Enabling Technologies deliberately suppressed by a lease/rental mix shift. [I] The composition is better than the headline in the core and worse in the acquired.

How big will this market be — growing, shrinking, domestic or international? [I] Global spine is roughly $11–12bn growing low-to-mid single digits — a volume tailwind from demographics net of ~1% annual ASP erosion. It is neither shrinking nor exciting. GMED is 80.6% US ($2,367.6M) and 19.4% international ($571.3M) [F], versus large-cap device peers at 30–40% international, so geographic expansion is a genuine multi-year runway — albeit at lower margin and slower pace. Neuromodulation (Nevro’s market) is more concentrated and has been in industry-wide decline.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? [I] Less competitive in ownership and more competitive in technology — an unusual and important split. On ownership, capital is leaving: [F] Stryker divested spine to VB Spine; ZimVie (itself a Zimmer Biomet spin-off) divested spine; Medtronic treats spine as a growth-capped legacy line. VB Spine newly appears in GMED’s FY2025 10-K competitor list and not in earlier years — GMED is documenting its rivals’ transfer from strategic to financial owners in real time. On technology, robotics has become more contested: [F] CEO Pfeil concedes deal cycles have “elongated a bit because hospitals are now requiring everyone to go back and look at all the competitive offerings.” Net, this is a favourable phase of the capital cycle inside a structurally poor industry.

How profitable is the business (ROIC, ROE)? [F] FY2025 ROIC ~10% (ROIC.ai 10.12%; hand-computed 9.9% from EBIT $537.2M taxed at 24% over invested capital of $4,135M). ROE is ~11.8% ($537.9M over $4,573.3M of equity) — and note that ROIC.ai’s reported 42.33% ROE for FY2025 is arithmetically wrong and was rejected. [F] Historical ROIC: 16.4% (2015), 13.6% (2018), 11.2% (2019), 10.3% (2022), then 4.7% (2023), 4.0% (2024), 10.1% (2025). [I] The critical decomposition: excluding goodwill and intangibles, ROIC is ~20.9%. The operating business earns excellent returns; the acquisition price drags the consolidated figure to roughly cost of capital.

How profitable is the industry — how many competitors, what barriers to entry? [F] Eleven-plus named competitors spanning three of the world’s largest device companies down to sub-scale specialists (Alphatec at $8.26/share, Orthofix at $11.25/share, both structurally unprofitable). [I] Barriers to entry at the product level are low: spinal hardware is machined titanium and PEEK, the regulatory path is predominantly 510(k) rather than PMA, and surgeons can switch trays between cases. Barriers at the scale level are real but modest: a national exclusive direct sales force, hundreds of millions of dollars of consigned instrument sets, and vertically-integrated manufacturing are genuinely hard to replicate — which is why the sub-scale players cannot earn a return.

Can the business be easily understood? [F] Yes. GMED makes and sells spinal implants and related hardware through a direct sales force, and gives away robots to protect the implant business. The complexity in the financials is entirely purchase accounting, not business model.

Can it be undermined by foreign low-cost labour? [I] Not materially. The economics are driven by surgeon relationships, regulatory clearance, product breadth, clinical evidence and the ability to fund consigned inventory — not by unit labour cost. Vertical integration is, if anything, a defence: GMED’s in-house manufacturing is the source of its 500–800bps gross-margin advantage, and offshoring it would surrender that. Tariff exposure on inputs is a second-order cost risk.

Do brands matter? [I] Not to patients, materially to surgeons — but as product-line familiarity rather than brand equity. A surgeon trained on CREO screws and SABLE expandables has real muscle memory and workflow investment. That is switching friction, not a brand moat, and it travels with the sales representative rather than residing with the company — which is precisely why GMED’s growth strategy is to recruit competitors’ reps.

What is the nature of competition? [I] Competition is for surgeons, mediated by sales representatives, and adjudicated increasingly by hospital value-analysis committees. It is fought on product breadth, set availability, service intensity and price. [F] GMED’s stated weapon is operational: “our execution around rep onboarding with sets and inventories continues to differentiate us… by providing high-return capital investments.” This is a balance-sheet advantage converted into a commercial one. It works — and it is symmetric, so it is not a barrier to entry.

Customers’ switching costs? [I] Low-to-moderate and often overstated. A surgeon can switch implant systems between cases; the friction is training, instrument familiarity and preference-card administration, not contract or technology. Robot placement raises friction modestly — a hospital that has installed ExcelsiusGPS, trained its OR staff and integrated the workflow will default to Globus implants — but the lock-in is economic and behavioural, not technical. The contrast that matters: Intuitive’s da Vinci instruments are chip-locked to the system, which is why Intuitive charges for both razor and blade. [F] GMED cannot, which is why it is now moving robots to leases and rentals to protect the implants.

Financial Condition & Balance Sheet

Assets not fully recognised on the balance sheet? [I] Three. (1) The exclusive direct US sales force — the recruiting, training and relationship base is expensed as incurred and is arguably GMED’s most valuable asset. (2) The in-house manufacturing capability and process knowledge underpinning a 72–78% historical gross margin, carried at depreciated cost. (3) Internally-developed IP — ExcelsiusGPS, Surgimap, the expandable interbody portfolio and 30+ products launched since FY2022, all R&D-expensed. Offsetting these, [F] $2,180.1M of acquired goodwill and intangibles sit on the balance sheet at 47.7% of equity — assets that are fully recognised and, on current returns, arguably over-recognised.

Off-balance-sheet liabilities? [F] Nothing material or unusual. Finance-lease obligations of $118.7M are on-balance-sheet. Contingent milestone consideration exists on the 2024 Fibriant B.V. biotechnology acquisition (good-manufacturing-process and FDA-approval milestones). The genuine off-balance-sheet exposure is litigation: the Moskowitz appeal (six patents reading across a very broad swath of the core interbody portfolio) and the 4WEB suit against the Modulus line carry no accrued liability beyond counsel fees — an adverse outcome would be entirely unreserved.

How conservative is the accounting? [I] Mixed, with two genuine flags and one mitigant. Flag one: FY2025 GAAP net income is ~30.5% non-recurring — a $117.7M non-taxable bargain purchase gain and a $46.3M deferred-tax valuation-allowance release [F] — which most casual screens will treat as operating earnings. Flag two: the headline GAAP gross-margin improvement from 55.6% to 64.3% is roughly 90% inventory step-up amortisation roll-off; the real underlying move was 67.4% → 68.1% [F]. Mitigant: management’s own non-GAAP presentation correctly excludes the bargain purchase gain and taxes at 24%, and the disclosure of base-business-versus-Nevro splits, day-adjusted growth and constant-currency figures is genuinely above average. Auditor is Deloitte & Touche LLP. [I] Nothing here is aggressive or improper; it is distorting, and the distortions are disclosed for anyone who reads past the headline.

How CapEx-hungry is the business? [F] Very. Capital expenditure was $174.4M in FY2025 (5.9% of sales) and is guided to 5–6% of sales for 2026 — and this understates the true capital intensity, because it excludes the working capital that surgical sets consume. Inventory of $759.3M against $957.8M of COGS is ~289 days; the cash conversion cycle is 317 days. [I] In practical terms, every incremental dollar of spine revenue requires roughly a dollar of working capital plus set capex before it converts to cash. This is the structural reason spine businesses look better on the income statement than on the cash flow statement, and it is visible in GMED’s own history — FCF margin fell to ~10% in FY2022–23 when growth accelerated.

Capital Allocation & Management

How much FCF does the business generate, how does management use it, what is the philosophy? [F] FY2025 free cash flow was $579.0M (19.7% margin), up from $405.2M (FY2024), $165.2M (FY2023) and $104.4M (FY2022) — a 5.5x increase in three years. Q1 2026 generated $202.4M of operating cash flow against $39.6M of capex.

Management’s stated priority order is explicit and consistent across both recent calls: (1) internal product development; (2) capital spending on surgical sets, facilities and manufacturing footprint; (3) share repurchase to minimise dilution; (4) complementary M&A. [F] FY2025 deployment: $174.4M capex, $300.5M buybacks, $252.5M for Nevro, $450M repaying the NuVasive-assumed convertible notes.

[I] The philosophy is coherent and, in FY2025, well executed — the company deleveraged to net cash, bought a distressed asset below the fair value of its net assets, and repurchased stock below $60. The problem is not the philosophy; it is that priority (4) sits behind an incentive structure that pays for revenue.

Significant acquisitions recently? [F] Two, both transformational. NuVasive — all-stock, closed 2023-09-01, 0.75 Globus shares per NuVasive share, taking the count from 100.2M to 136.3M (~36% dilution). Nevro — $252.5M cash, closed 2025-04-03, generating a $117.7M non-taxable bargain purchase gain and contributing $293.6M of revenue in nine months (~0.6x annualised sales).

[I] The scorecards differ. NuVasive was operationally well integrated — synergy targets of mid-to-high-single-digit growth and mid-30s base-business adjusted EBITDA were met (H2 2025 organic 8.8%; Q4 2025 base adjusted EBITDA 35.7%) — but three years on, consolidated ROIC is merely back to the pre-merger ~10% on a business with 36% more shares. Legacy holders financed a doubling of revenue for no improvement in return on capital. Nevro was bought cheaply and made EPS-accretive within nine months, beating guidance by 15 months — genuinely impressive — but its revenue is now declining 17.1% sequentially because of GMED’s own sales-force restructuring, and management expects it to worsen before improving. [I] Cheap purchase price, excellent cost execution, damaged revenue base. The jury is out.

Buying back shares? [F] Yes, and at good prices — >$600M deployed since FY2022 repurchasing >10M shares at an average “under $60 a share” versus $76.49 today; $300.5M in FY2025 alone (~4.3M shares at ~$70). A $500M authorisation announced in Q2 2025 has $390M remaining. [I] But the effect is modest: shares outstanding went 136.3M (2023) → 137.4M (2024) → 135.1M (2025), a net reduction of just 1.2M despite $386M of repurchases across 2024–25, because ~$50M/year of stock-based compensation and option exercises offset most of it. Buybacks at GMED are dilution management, not per-share value creation. Investors should not model buyback-driven EPS accretion.

Issuing large amounts of new shares to insiders? [F] Stock-based compensation runs ~$50M/year ($49.8M in FY2025, $54.2M FY2024, $52.7M FY2023) — approximately 1.7% of revenue and 8.6% of free cash flow. Not egregious for a medical device company. However, the 2026 annual meeting (2026-06-03) approved an increase in shares authorised under the 2021 Equity Incentive Plan, and 192,602,552 Class B shares remain authorised for issuance — an amount the 10-K notes exceeds 5% of outstanding common stock and which the board could issue without necessarily triggering the charter’s automatic conversion provision.

Compensation policy of directors/management? [F] This is the weakest element of the entire investment case. Under the 2025 Non-Equity Incentive Compensation Plan, 90% of the named-executive annual cash bonus is determined by a single metric: revenue. The remaining 10% is a discretionary Compliance & Quality assessment. Achievement of ≤75% of the revenue goal pays zero; 100% pays 110% of base reference; 110% pays 145%. FY2025 revenue excluding Nevro was 98% of goal, producing a 102% payout. There is no margin gate, no EPS gate, no cash-flow gate and no return-on-capital gate.

Long-term equity was granted as stock options, not performance share units with operating conditions — David C. Paul 75,000, Daniel Scavilla 140,000 (grant-date fair value $5,006,645), Kelly Huller 40,000, all struck at $92.42 on 2025-01-23 and all currently underwater.

[I] A revenue-only bonus is precisely the incentive that rewards buying revenue with equity: the numerator rises and the denominator is not measured. The 36% dilution that halved return on invested capital reduced no executive’s bonus; the revenue it purchased increased them. Three years and one CEO later, the metric is unchanged — which makes this a forward-looking risk to the next acquisition, not merely a critique of the last one.

Motivations of management? [I] The evidence points in an uncomfortable direction, with one genuine offset.

Against: [F] Since the NuVasive close, insiders have executed $50,194,615 of open-market sales and $0 of open-market purchases across 84 Form 4 filings — not a single discretionary code-P purchase in nearly three years, a window that included a drawdown to $43.79 and a recovery to $101. The dominant transaction pattern is exercise-and-immediately-sell (S 634,068 shares against M 573,595). The timing is worse than the aggregate: then-CEO Scavilla sold 130,000 shares for $11.05M at $85.00 in November 2024, eight months before resigning; CFO Kyle Kline and director Leslie Norwalk both sold at $101.10 on 2026-01-08 — the session in which GMED printed its five-year intraday high of $101.3975; EVP/GC Kelly Huller sold 20,000 shares at $94.50 in February 2026 and has sold in nine separate windows since 2024.

Also against: [F] two abrupt, immediate-effect senior departures in twenty months — CEO Scavilla from both the executive role and the Board on 2025-07-18 (disclosed in the same 8-K as preliminary Q2 sales results, six months after a $5.0M option grant), and director John A. DeFord on 2026-03-23 — neither explained beyond boilerplate.

For: [F] David C. Paul, the founder, Executive Chairman and controlling holder, does not appear as a seller in this window. His Class B control block is intact and the family’s economic exposure remains very large. [I] That is a meaningful offset — the person with the most at stake has not sold — and it is the strongest single argument that management believes in the long-term outcome even as the professional managers monetise.

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? [F] None of these. Globus Medical is a Delaware corporation filing standard US domestic forms (10-K, 10-Q, 8-K, DEF 14A) and issuing Form 1099 dividends (of which there are none). Class A common stock trades on the NYSE under GMED; a separate Class B class carries superior voting rights and is held by insiders. [I] The dual-class structure is the single most consequential structural feature for a minority shareholder: it makes GMED a “controlled company” under NYSE Rule 303A.00, permitting it to forgo a majority-independent board and independent director nominations. There is no realistic activist or takeover path.

Dividend policy? [F] No dividend has ever been paid, and none is contemplated. All capital return is via share repurchase ($390M of authorisation remaining at 2026-03-31). [I] Appropriate for the growth profile and capital intensity; a dividend would be a poor use of capital for a business that must fund 289 days of inventory and 5–6% of sales in capex.

How profitable is the business? [F] FY2025: adjusted gross margin 68.1% (GAAP 64.3%), operating margin 18.3%, adjusted EBITDA margin 31.3% (Q4 33.9%; base-business Q4 35.7%), free-cash-flow margin 19.7%, ROIC ~10%, ROE ~11.8%. [I] Highly profitable at the operating line, unremarkably profitable at the capital line — the gap is the acquisition price, and it is the central tension in the investment case.

Is net income diverging from cash from operations? [F] Yes, and favourably — which is the right direction. FY2025 operating cash flow of $753.4M was 1.40x GAAP net income of $537.9M; the multi-year series is 1.85x (2021), 0.94x (2022), 1.98x (2023), 5.06x (2024), 1.40x (2025). [I] The divergence is benign and explicable: D&A of $276.8M — predominantly non-cash merger intangible amortisation — exceeds capex of $174.4M. Critically, the $117.7M bargain purchase gain was a non-cash credit to earnings, so cash earnings are in fact stronger relative to accrual earnings than the headline ratio implies. This is the single strongest counterweight to the earnings-quality concerns raised elsewhere in this appendix: whatever one thinks of GAAP net income, $579M of free cash flow is not an accounting construct.

Valuation summary. [F] At $76.49 (2026-07-24) on 135,055,223 shares: market capitalisation ~$10.33bn, net cash ~$684M, enterprise value ~$9.65bn. EV/FY2026E sales 3.00x; EV/FY2026E adjusted EBITDA ~9.3x; P/FY2026E non-GAAP EPS 15.9–16.3x; P/B 2.23x; FY2025 FCF yield 5.6% on market cap. On its own ~10-year history: composite 5.7th percentile, P/E 1.2nd, P/S 5.3rd, P/B 10.7th. Sell-side targets: Truist $100 (cut from $115, 2026-07-16), RBC Capital $115 (cut from $120, 2026-07-13).

Risks & Downside

What factors would cause the stock to decline? [I] In descending order of probability-weighted impact:

1. US Spine growth decelerating below ~5%, confirming the ~500bps of outgrowth was a one-time harvest of competitor dislocation rather than durable share capture.
2. Adjusted gross margin stalling at the guided 69–70%, killing the mid-70s bridge worth ~$170M of gross profit and ~$0.94 of EPS (~20% of current earnings).
3. Another equity-financed acquisition — the revenue-only bonus still points there, M&A remains an explicit capital priority, and the controlled structure means no shareholder can prevent it.
4. Nevro continuing to decline into 2027, which would break the acquisition thesis and damage management’s operational credibility at the same time.
5. Enabling Technologies declining again in FY2026 — small revenue impact (4.8% of sales), large narrative impact, since it would strip the last support from the premium-multiple case.
6. Goodwill impairment on the $2,180.1M intangible balance (47.7% of equity) — non-cash, but a public confirmation that the merger overpaid.
7. CMS WISeR prior authorisation capturing spinal cord stimulation, landing on an already-declining business.
8. An adverse Moskowitz appeal outcome, which would be entirely unreserved.

Risk of a catastrophic loss? [I] Low. The realistic tail risks are a large adverse patent judgment on the Moskowitz appeal (unreserved, and reading across a broad swath of the core interbody portfolio), a severely value-destructive equity-financed acquisition enabled by the controlled structure, or a systematic product-quality or recall event. None of these is remotely likely to be existential given ~$684M of net cash, no funded debt, ~$579M of annual free cash flow and revenue diversified across five product families with no customer above 10% of sales.

Chance of a total loss? [I] Negligible. GMED is profitable on both GAAP and cash bases, debt-free, holds ~$799M of cash and securities against $115M of finance leases, carries a current ratio of 4.26x, and has a tangible asset base above $2.3bn. A total loss would require a simultaneous, sustained collapse of the entire spine implant market and a catastrophic legal outcome. The practical downside case in this memo is approximately -8% from the current price, not a wipeout — which is precisely why the situation is interesting despite the quality reservations.

Recent News & Events

Has the business environment changed recently? [I] Yes, in three material ways. (1) The competitive landscape has been reshaped by exits — Stryker divested spine to VB Spine and ZimVie divested spine, and VB Spine appears in GMED’s FY2025 10-K competitor list for the first time. Capital is leaving the category, which is favourable for the committed. (2) Robotics has become more contested — CEO Pfeil concedes hospitals now run competitive evaluations that have “elongated” deal cycles, and GMED has responded by shifting from outright robot sales to leases and rentals. (3) GMED’s own capital structure is transformed — from $450M of assumed convertible debt at the start of 2025 to net cash of ~$684M today.

Significant acquisitions? [F] Nevro Corp., $252.5M cash, closed 2025-04-03 — spinal cord stimulation (HFX/Senza) and minimally-invasive sacroiliac joint treatment. Produced a $117.7M non-taxable bargain purchase gain, later subject to a measurement-period restatement adjustment. Earlier: the 2024 Fibriant B.V. hemostasis biotechnology share acquisition with milestone consideration, plus a series of individually immaterial acquisitions. The defining transaction remains NuVasive (closed 2023-09-01).

Change in accounting policies? [F] No changes in accounting policy. What has changed materially is the effect of acquisition accounting: inventory step-up amortisation rolling off (driving GAAP gross margin from 55.6% to 64.3% while adjusted moved only 67.4% → 68.1%), the $117.7M bargain purchase gain, a $46.3M deferred-tax valuation-allowance release taking the GAAP tax rate to 11.1%, and a measurement-period restatement adjustment to the Nevro purchase accounting. [I] None of this is improper and all of it is disclosed — but it makes reported GAAP results for FY2023 through FY2025 essentially unusable for trend analysis without normalisation.

Recent changes — new markets, facilities, management?

• [F] Management: CEO Daniel T. Scavilla resigned from the CEO role and the Board effective immediately on 2025-07-18; Keith W. Pfeil (then COO/CFO) appointed the same day; Kyle Kline subsequently became CFO. Director John A. DeFord resigned effective immediately on 2026-03-23.
• [F] New markets: neuromodulation/chronic pain entered via Nevro (April 2025); trauma scaled materially (+34% in Q1 2026) including the PRECICE limb-lengthening portfolio.
• [F] Facilities: PRECICE manufacturing transferred from former NuVasive facilities to Globus in early 2025, with output now exceeding the historical facility; broader NuVasive and Nevro manufacturing consolidation into Globus plants is the mechanism behind the mid-70s gross-margin target.
• [F] Products: two Q2 2026 FDA 510(k) clearances for the Scripps patient-specific lumbar interbody spacer system (seven systems) and Scripps patient-specific rods, integrated with the Excelsius suite; six spine launches in 2025 (four in Q4: CREO Traction, Reline 3D Towers, AMS Freehand, HEDRON C-MIS); 30+ products launched since FY2022.
• [F] Capital: $450M of convertible notes repaid in cash (March 2025); $500M repurchase authorisation announced Q2 2025 with $390M remaining; 2021 Equity Incentive Plan share increase approved 2026-06-03.
• [F] Upcoming: Q2 2026 results are scheduled for release after market close on 2026-08-06 — after the date of this report.

APPENDIX B — Source Appendix

Globus Medical, Inc. (NYSE: GMED) · Research date 2026-07-26

All sources below were accessed on 2026-07-26 unless otherwise stated. Sources are ordered by evidentiary authority: SEC filings first, then management commentary, then quantitative datasets, then third-party material. Every material claim in the memo and Appendix A traces to a source listed here.

B.1 SEC filings — primary sources (highest authority)

The complete trailing-60-month EDGAR corpus for CIK 0001237831 was reviewed. All filings are publicly available at https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001237831. Form breakdown: 5× 10-K, 1× 10-K/A, 15× 10-Q, 45× 8-K, 1× 8-K/A, 5× DEF 14A, 6× DEFA14A, 122× Form 4, 3× Form 4/A, 4× Form 3, 7× Form 425, 1× S-4, 1× S-4/A, 2× S-8, 5× SD.

Annual reports

https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001237831&type=10-K

Quarterly reports

Proxy statements

Current reports (8-K) — material events relied upon

Registration statements

• S-4 and S-4/A — NuVasive merger registration statements, relied upon for the 0.75x exchange ratio and merger mechanics.
• Form 425 (7 filings) — merger communications.

Insider filings (Form 3 / 4 / 4A)

All 84 Form 4 and Form 4/A filings covering 2023-09-01 (NuVasive close) through 2026-07-26 were retrieved as raw XML from EDGAR and parsed programmatically on 2026-07-26. Base path: https://www.sec.gov/Archives/edgar/data/1237831/

Aggregate result: $50,194,615 of open-market sales (code S with a stated price); $0 of open-market purchases (code P). Non-derivative transaction codes: S 634,068 shares; M 573,595 shares (option exercise); A 83,190 shares (award); G 14,100 shares (gift).

Individual transactions specifically cited in the memo and appendix:

The two 2026-01-08 sales at $101.10 were cross-checked against the AZI daily price file, which records an intraday high of $101.3975 that session — the five-year intraday high.

B.2 Management commentary (treated as hypothesis, not evidence — the standing rule that management commentary is a hypothesis, not evidence)

Speakers: Keith W. Pfeil (President & CEO), Kyle Kline (CFO), Brian Kearns (SVP Business Development & IR). Analyst questioners cited: Ross Osborne for Laurence Biegelsen (Wells Fargo), Shagun Singh (RBC Capital Markets), Matt Taylor (Jefferies), Matt Miksic (Barclays), Richard Newitter (Truist Securities), Brian Zimmerman (BTIG), David Saxon (Needham & Company).

Specific management statements quoted in the memo or appendix:

• FY2026 guidance: revenue $3.18–3.22bn; non-GAAP diluted EPS raised to $4.70–4.80 from $4.40–4.50 (Q1 2026 call)
• FY2025: revenue $2.939bn (+16.7%); non-GAAP EPS $3.98 (+30.8%); base business ex-Nevro +5.0%; Nevro contributed $293.6M; adjusted EBITDA 31.3% (Q4 2025 call)
• FY2025 adjusted gross margin 68.1% vs. 67.4% in FY2024; full-year GAAP 64.3% vs. 55.6%; Q4 GAAP improvement “driven primarily by lower inventory step-up amortization” (Q4 2025 call)
• Six consecutive quarters of sequential adjusted gross-margin expansion; 69.2% in Q4 2025 and Q1 2026; FY2026 guided range 69–70%; long-term “mid-70s” target (both calls)
• Q1 2026: revenue $759.9M (+27.0%, +25.5% cc); GAAP net income $124.3M / $0.90 diluted; non-GAAP $154.9M / $1.12 (+64.7%); adjusted EBITDA margin 32.3%; base business $677.2M (+13.2%); Enabling Technologies $26.9M (+21.1%); Nevro $82.7M (-17.1% sequentially); US $604.9M; International $155.0M (Q1 2026 call)
• Nevro EPS-accretive within nine months, beating initial guidance by 15 months; standalone adjusted EBITDA margin 16.2% (Q3 2025) → 21.2% (Q4 2025) → 11.8% (Q1 2026) (both calls)
• “>$600 million to repurchase over 10 million shares at an average price of under $60 a share… greater than 25% of the dilution that was created from the NuVasive merger” (Q1 2026 call)
• “~130,000 robotic procedures” cumulative on ExcelsiusGPS; “58 weeks of consecutive growth” (Q1 2026 call); “48 weeks” (Q4 2025 call)
• “the mix of pipeline deals is shifting with a greater focus on leases and rentals compared to the historical mix of outright sales, which historically resulted in higher upfront revenue recognition” (Q1 2026 call)
• “we are competing mainly against Medtronic. That is what I see routinely”; deal cycles have “elongated a bit because hospitals are now requiring everyone to go back and look at all the competitive offerings” (Q1 2026 call)
• “you are going to launch new products, but you are going to see price erosion — you might be down that 1%”; the mid-70s path “majority… really comes from our ability to drive costs” (Q1 2026 call)
• Nevro “will probably get a little bit worse before it gets better” (Kyle Kline, Q1 2026 call)
• WISeR/spinal cord stimulation: “I have not seen that impacting us… I have really no comment” (Keith Pfeil, Q1 2026 call)
• FY2026 opex guidance: R&D 5–6% of sales; SG&A 38–39% of sales; capex 5–6% of sales; non-GAAP tax rate 24–25% (Q4 2025 call)
• Q1 2026 balance sheet: cash and securities $799.3M; operating cash flow $202.4M; capex $39.6M (5.2% of sales); $390M of repurchase authorisation remaining (Q1 2026 call)

Background reference material Saved GMED earnings-call transcripts for Q2 2022, Q3 2022, Q4 2022, Q1 2023, Q2 2023, Q3 2023 and Q4 2023 (all last modified 2024-04-22). Reviewed for pre- and immediately-post-merger context; superseded for this memo by the current ROIC.ai transcripts.

B.3 Quantitative datasets

ROIC.ai MCP (third-party aggregated data — NOT primary; reconciled to filings)

Tools called against identifier NYSE:GMED (the bare ticker GMED is rejected): get_company_profile, get_income_statement (annual, 11 periods), get_balance_sheet (annual, 7 periods), get_cash_flow (annual, 7 periods), get_profitability_ratios (annual, 11 periods), get_enterprise_value (ttm), get_earnings_call_transcript, get_company_news.

Fields relied upon: multi-year revenue, gross profit, operating income, net income, EPS, EBITDA, D&A, SG&A, R&D; balance-sheet detail (cash, inventory, receivables, goodwill, intangibles, debt, equity); cash-flow detail (OCF, capex, FCF, buybacks, acquisitions, debt repayment); return_on_inv_capital (10.12% FY2025, corroborated by hand calculation at 9.9%).

Fields REJECTED as erroneous (documented in Section 6.4):

• bs_sh_out = 112,625,126 for FY2025 — contradicted by the 10-K’s 135,055,223 shares outstanding at 2025-12-31 and by the 137,056,000 weighted diluted count. All memo valuation uses the filing figure.
• return_com_eqy = 42.33% and return_on_cap = 33.65% for FY2025 — arithmetically impossible against $537.9M of net income on $4,573.3M of equity (11.8%).
• bs_tot_cap = $1,506,748k for FY2025 — inconsistent with the reported balance sheet.

Tool limitations noted: list_earnings_calls ignores the identifier argument and returns a global cross-ticker list; transcripts must be fetched directly by (identifier, year, quarter). get_company_news returned an empty array for NYSE:GMED, so the recent-events timeline was rebuilt from 8-Ks, the 10-K, transcripts and public trade press.

AZI (azitrading.com)

• Daily price history — https://azitrading.com/controls/download-data.php?t=GMED, downloaded 2026-07-26. 3,512 rows spanning 2012-08-03 to 2026-07-24; adjusted and unadjusted OHLC, volume, dividend and split columns, 21/50/200-day EMAs, beta and alpha. Used for the Five-Year Event Map above, all price levels, the five-year high/low, the 52-week range, monthly closes and the twenty largest single-day moves.
• Valuation index (own-history percentiles) — AZI valuation index for GMED, updated 2026-07-24 22:30:58. Price $76.49; TTM EPS $4.289; book value per share $34.2447; TTM sales per share $22.6615; P/E 17.8338 (1.173rd percentile); P/B 2.2336 (10.72nd percentile); P/S 3.3753 (5.31st percentile); composite percentile 5.735; n_components 3. Used only as own-history context, never cross-sectionally.
• Peer closing prices (2026-07-24) for context: ATEC $8.26, OFIX $11.25, MDT $83.21, SYK $330.25, ZBH $91.35.

FactorsToday factor model (https://www.factorstoday.com/api)

Endpoints called for GMED on 2026-07-26 — third-party statistical estimates, not primary data:

• /stock-loadings/GMED — four nested ElasticNet models. Base (R² 0.163): Market +0.794, DividendYield +0.489, SmallSize +0.173, Value +0.124, BetaFactor -0.033. Base + Sector (R² 0.167): Market +0.796, DividendYield +0.524, SmallSize +0.163, Value +0.092, Sector Health Care +0.085. Base + Sector + Industry (R² 0.245): Market +0.987, Industry Medical Devices +0.638, SmallSize +0.365, DividendYield +0.361, Value +0.105. All Factors (R² 0.206): Market +0.824, Industry Medical Devices +0.481, DividendYield +0.379, SmallSize +0.246, Value +0.070. Momentum is L1-zeroed (absent) in all four models. Betas are read within a single model only, never compared across models.
• /leaderboard/GMED — all returns annualised: y1 +43.5% (Sharpe 0.842, Sortino 1.960, max drawdown -22.1%); y3 +7.4%; y5 -0.42%; y10 +11.9% (Sharpe 0.278, max drawdown -47.9%); m6 -31.7% annualised (≈ -17.3% actual); m3 -57.7% annualised (≈ -19.4% actual). De-annualised and cross-checked against the AZI price file.
• /stock-info/GMED — market cap $10,237,858,816; beta 0.9316; alpha -0.0870; rs_ytd -12.39; rs_6m -18.76; rs_12m +41.54; rs_peak -21; close $76.49 on 2026-07-24.
• /stock-specific-vol/GMED — idiosyncratic volatility 45.48% annualised (2.865% daily) over a 252-day window; model R² 0.206.
• /related-stocks/GMED — factor-similar peers by similarity: NVST 0.918, BAX 0.914, STE 0.912, MMSI 0.911, ITGR 0.903, COO 0.897, TRMB 0.896, SOLV 0.892, LMAT 0.891, ALGN 0.887, RMD 0.883, GEHC 0.882, TMDX 0.888. Notably the model does not pair GMED with SYK, MDT, ZBH or ISRG.

Methodology reference: https://www.factorstoday.com/about. Loadings are L1-sparse (absent = zeroed, not missing); orthogonalisation is hierarchical and level-dependent; all factors are volatility-scaled to 10%; R² is in-sample and mildly overstated.

SEC EDGAR (direct)

• Company CIK 0001237831; full filing history at https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001237831
• The trailing 60-month filing corpus (2021-07-01 onward) was enumerated and reviewed in full
• Form 4 insider filings parsed directly from the raw XML at https://www.sec.gov/Archives/edgar/data/1237831/

B.4 Third-party and market data

• Truist Securities — price target lowered to $100 from $115, 2026-07-16 (via public market-data aggregators, accessed 2026-07-26).
• RBC Capital Markets — price target lowered to $115 from $120, 2026-07-13 (same).
• Globus Medical investor relations — https://www.investors.globusmedical.com/news-events/press-releases; Q2 2026 earnings scheduled for release after market close on 2026-08-06 (GlobeNewswire, 2026-07-16, https://www.globenewswire.com/news-release/2026/07/16/3328822/0/en/globus-medical-schedules-second-quarter-earnings-release-and-conference-call.html).
• Company website — http://www.globusmedical.com.

No analyst price target, and no third-party or aggregator valuation figure, has been adopted as or converted into a price target in this memo (the no-price-target discipline applied throughout this article). Sell-side targets are cited solely as evidence of the gap between consensus modelling and the traded price.

B.5 Comparative context and analytical frameworks

Comparative context. This note draws on the author’s prior published work on Medtronic (June 2026), Stryker (June 2026), Zimmer Biomet (July 2026), Intuitive Surgical (July 2026), Boston Scientific (June 2026), GE HealthCare (June 2026) and Steris (July 2026) for peer and industry cross-read — specifically: spine as a structurally price-pressured legacy device category with low-single-digit annual ASP erosion; Stryker’s divestiture of its spine business and the scale of the Mako installed base (>2 million cumulative procedures); Zimmer Biomet’s ZimVie spin-off and that entity’s subsequent spine divestiture; and Intuitive’s chip-locked da Vinci instruments as the reference case for a technical razor-and-blade lock-in. All underlying facts in those comparisons are sourced from the respective companies’ own public filings and earnings calls.

Background reference material. A dated third-party US healthcare industry primer (Morgan Stanley, 2011) was consulted for healthcare value-chain and reimbursement-mechanics framing only, and explicitly not used as current market data.

Note on ownership. Nothing in this research constitutes evidence of, and nothing herein asserts or implies, any position in Globus Medical securities held by the author or by any other party.

Analytical frameworks applied (.claude/skills/investment-research-frameworks):

• Competition Demystified (Greenwald & Kahn) — moat-type taxonomy used to classify GMED’s advantage as supply-side/cost with a symmetric distribution component, and to reject demand-side captivity, network effects and intangible-brand claims; the market-share-stability test applied in Section 4.4; the ROIC test applied in Section 6.4.
• Capital Returns (Marathon Asset Management / Chancellor) — supply-side capital-cycle analysis applied in Section 3 to the exits of Stryker and ZimVie from spine and the transfer of assets from strategic to financial owners; the asset-growth anomaly applied to the NuVasive and Nevro acquisitions in Section 7.