Somnigroup International, Inc. (NYSE: SGI) — Rolling Up the Whole Mattress at the Trough, on Negative Tangible Equity

Independent fundamental research. Report date: July 3, 2026.

The main body of this article carries no investment recommendation and no price target; valuation is discussed only as embedded expectations and scenarios. The single, deliberate exception is the “Author’s Take” block immediately below, which is explicitly labeled and fenced off from the analysis.

⚡ Author’s Take

This block is the author’s own subjective opinion and general information only — not investment advice. Everything from the Executive Summary onward is position-free and carries no price target.

Verdict: HOLD / not-a-short — a good brand portfolio and a coherent counter-cyclical roll-up, but priced (base case ≈ the current $78) for a recovery it still needs, on a balance sheet with no margin for error. I would accumulate only well below the recent low — a low-$60s / high-$50s entry (roughly ≤14–15x normalized mid-cycle EPS of ~$4) — and I would not chase it here. Conviction: medium.

The bull and bear are both right about different things, and the price sits exactly where they meet. The bull is right that Somnigroup owns the only genuine premium brand in bedding (Tempur-Pedic), the largest manufacturing and advertising scale, and now the #1 retailer — and that it is consolidating a field that has been cleared by the Serta bankruptcy and the DTC bust, at the bottom of a three-plus-year mattress-unit recession. That is a real, ownable franchise and a defensible counter-cyclical logic. The bear is right that the market is paying a re-rated ~19x EV/EBITDA and ~32x trough earnings for it — a quality-compounder multiple on a business whose own factor signature screams housing-linked building-products cyclical (its statistical peers are Griffon, Owens Corning and Fortune Brands, not staples), whose ROIC has fallen every year from 24% to 8%, whose tangible common equity is deeply negative, and which is funding a second debt-cycle-timed roll-up (Leggett & Platt) with fresh equity after buying back stock at half today’s price. The honest read is that scale is diluting returns, not improving them, and that the “cheap” 7.6th-percentile P/S is a Mattress-Firm-consolidation mirage.

Framing: a cyclical, not a compounder — mis-shelved at a compounder’s multiple. This is neither a falling knife (it made an all-time high five months ago and is only ~20% off it) nor a momentum name (6-month relative strength is negative and rolling over). It is a fairly-priced bet on a mattress-cycle recovery plus flawless integration of two large, dissimilar acquisitions, wrapped in more financial leverage than the category’s cyclicality warrants. At $78 you are underwriting the good outcome and getting the leverage for free on the downside — the same one-two (earnings down and multiple down) that took the stock −59% in 2022.

Conviction, and the two things that would move me. Turns me bullish: US mattress unit volumes inflect positively and consolidated ROIC climbs back toward the mid-teens — proof the trough is cyclical, not structural, and that the roll-up is accretive to returns rather than just to revenue. Turns me bearish (toward an actual avoid): a fourth consecutive down year in category units, a DOJ/FTC Second Request that drags the LEG deal into contested litigation, or ROIC stuck below 10% into 2027 — any of which exposes the negative-tangible-equity balance sheet to a cycle that won’t cooperate. Tag: “Bought the whole bed at the bottom — on the house’s credit card.”

📈 Stock Price Action — Five-Year Event Map

Factual price history; no recommendation, no price target. Prices are split-adjusted for the 4:1 split of November 24, 2020, so all figures are on today’s share basis.

Over the trailing five years SGI (as Tempur Sealy / TPX until the February-2025 rename) has traveled a full cyclical round-trip and then made new highs: from a post-COVID-boom peak near ~$50 (Sep 2021), down roughly 59% to a ~$20 demand-bust trough (Jun 2022), then a multi-year grind to a genuine all-time high of ~$98 on Feb 10, 2026, before pulling back to $78.43 (Jul 2, 2026). The stock now sits about −20% off its all-time high, near the upper end of a 52-week range of $61.49–$97.99. The key framing correction: SGI is not a fallen 2021 high-flyer — post-split, its all-time high was set five months ago in the Somnigroup/Mattress Firm era, and today’s price is a pullback from that peak, not the wreckage of a bygone bubble. (Any “$140s” print exists only on a non-split-adjusted pre-2020 chart and is irrelevant to the current share base.)

Cycle narrative. (1) The 2021 run captured the COVID/stimulus demand and margin peak. (2) The 2022 collapse was a textbook cyclical + rate shock — mattresses are a housing-linked, big-ticket discretionary purchase, and Fed tightening plus the stimulus roll-off hit demand and the multiple at once. (3) The 2022–24 recovery paired earnings stabilization with the strategic re-rating that followed the May-2023 Mattress Firm agreement. (4) 2024 was dead money on the FTC’s suit to block that deal. (5) Clearing the FTC (deal closed Feb 5, 2025 with a shelf-space remedy), the rename, and the subsequent Leggett & Platt proposal drove the stock to a genuine all-time high near $98 by February 2026. (6) The Feb–May 2026 drawdown reflects a still-soft category and digestion of the LEG deal. (7) The recent bounce coincides with the June-3-2026 US HSR clearance on LEG. Price moves are Fact (AZI 5-year CSV); attributed causes are Interpretation, cross-referenced to 8-K events, deal dates, and category demand data.

1. Executive Summary

Somnigroup International — the entity formerly called Tempur Sealy International (TPX), renamed in February 2025 — is the world’s largest bedding company and, since the Mattress Firm acquisition closed February 5, 2025, a vertically integrated, three-layer roll-up of the US mattress industry: branded manufacturing (Tempur-Pedic, Sealy, Stearns & Foster), the #1 specialty retailer (Mattress Firm, ~2,300 US doors), and — via the pending all-stock acquisition of Leggett & Platt (NYSE: LEG), signed April 13, 2026 — the upstream components layer. No US competitor spans all three.

The strategy is coherent and the timing is defensible: management is consolidating a structurally weak category at the bottom of a three-plus-year demand recession, after the field has been culled by the Serta Simmons bankruptcy (2023) and the collapse of the direct-to-consumer challengers (Casper dead, Purple and Sleep Number loss-making). Antidumping duties protect the domestic profit pool. The brands are real, and Tempur-Pedic is the closest thing bedding has to a genuine franchise.

But the execution has come at a steep cost to financial quality, and the price already reflects the good outcome. FY2025 revenue jumped 51.6% to $7.48B, but that is almost entirely the Mattress Firm consolidation — organic category demand is flat-to-down. GAAP diluted EPS was $1.84 (adjusted $2.70; 2026 guidance $3.00–$3.40), and the $0.86/share GAAP-to-adjusted gap is built largely from recurring cash deal and integration costs that the LEG deal guarantees will keep recurring. ROIC has fallen every year, 24% → 8%, toward the cost of capital, as $7.2B of goodwill and intangibles and ~$2B of Mattress Firm leases have swamped NOPAT. Tangible common equity is deeply negative. Net debt is $4.55B (pro-forma ~$5.5B post-LEG). Interest expense doubled to $268M. The competitive advantage — one premium brand plus manufacturing/advertising scale — is real but narrow, sitting in a category with zero consumer switching costs and no secular unit growth; the Mattress Firm/LEG vertical integration is a scale-and-channel-control play, capped by an antitrust remedy (43% of premium shelf reserved for rivals), not a wide moat.

At $78.43 (market cap ~$16.2B, EV ~$21.9B) the stock trades at ~19x EV/EBITDA and ~32x trough earnings — the upper half of its own five-year range and roughly a 2x premium to every solvent furniture/bedding peer. The “cheap” 7.6th-percentile P/S is a Mattress-Firm-gross-up artifact. On our scenario work the base case — a mid-cycle earnings recovery to ~$4.00–4.50 with partial synergies at a normalized ~17–18x — lands near the current price, leaving little margin of safety if the trough persists and requiring near-flawless execution to reach the bull. The market is pricing SGI as a quality compounder; its own factor loadings price it as a rate-sensitive housing cyclical. That gap is the central variant perception.

2. Business Overview

What SGI is. Somnigroup styles itself “the world’s largest bedding company,” serving 100+ countries through a combined footprint of over 2,800 retail stores, ~30 e-commerce platforms, 70+ manufacturing plants, four R&D centers and 20,000+ employees. It markets its opportunity as the “$120 billion global sleep industry” — a deliberately expansive TAM. The relevant profit pool is far smaller: the US mattress market, ~$10B at wholesale / ~$11–12B at retail.

The three-layer value chain. The defining feature of the post-2025 company is vertical integration across the bedding stack:

1. Components (pending) — the all-stock Leggett & Platt acquisition adds springs, innersprings, specialty foam and adjustable bases — the upstream layer that today supplies the “substantial majority” of Tempur Sealy’s innersprings under a long-term agreement.
2. Branded manufacturing — Tempur-Pedic (premium specialty foam, highest ASP, est. 1991), Stearns & Foster (premium innerspring, 1846), Sealy / Sealy Posturepedic (value-to-mid volume, 1881/1950), plus Sleepy’s private label and unbranded OEM output.
3. Owned retail — Mattress Firm’s ~2,300 US doors, plus Dreams (UK), Tempur-Pedic boutiques, and SOVA (Sweden).

The strategic logic is control of the shelf and capture of the retail margin that historically leaked to third-party distributors. The strategic cost is a levered, low-return, lease-heavy retail business bolted onto a formerly high-return manufacturer.

Segments and the numbers. SGI reports three segments — Tempur Sealy North America, Tempur Sealy International, and Mattress Firm. FY2025 consolidated net sales of $7,476.5M were up 51.6%, but the growth is the Mattress Firm consolidation, not organic demand:

The optical −28.7% “decline” at Tempur Sealy North America is an accounting artifact: ~$976M of wholesale shipments that used to go to Mattress Firm as a third-party customer are now intercompany and eliminated in consolidation. The important read is the margin structure — manufacturing carries ~50% gross margins, retail only ~33%. Consolidating Mattress Firm therefore grosses up revenue while diluting blended margin (consolidated operating margin fell to ~10% from ~13%), and mechanically compresses the P/S ratio — which is why the “cheap” 7.6th-percentile P/S is misleading versus the honest EV/EBITDA (~18.7x) and P/E (~32x).

How it makes money — and the channel flip. SGI sells through two channels. Direct (company-owned stores, e-commerce, call centers) jumped to 63.5% of sales ($4,745.9M) from just 24.9% in 2024 — the entire shift is Mattress Firm. Wholesale (shipments to third-party retailers, hospitality, healthcare) is the remaining 36.5%. Notably, Mattress Firm is multi-branded — it also sells direct rivals Purple, Serta, Simmons, Beautyrest, Nectar and Tuft & Needle — a fact both blunting channel-conflict claims and reinforced by an antitrust remedy: to close the deal, SGI committed to reserve, on average, 43% of premium ($1,500+) floor slots for third-party premium mattresses, independently monitored.

Recurring vs. one-time revenue. Bedding is a poor recurring-revenue business. A mattress is a big-ticket durable replaced roughly every 8–10 years; there is no subscription, no consumable, no lock-in. The only genuinely recurring line is licensing/franchise royalty, at $39.5M in 2025 (~0.5% of sales). Essentially 100% of revenue must be re-won every purchase cycle — a structural fact that governs the moat discussion below.

Verdict: SGI is a vertically integrated, three-layer bedding conglomerate — components (pending), branded manufacturing, and the #1 US mattress retailer — assembled via a debt-and-stock roll-up at a demand trough. It owns the best brand portfolio and the largest distribution reach in the category, but the model is now half low-margin retail, revenue is almost entirely non-recurring, and reported growth is acquisition-manufactured. Bigger and broader than old Tempur Sealy; also lower-margin, more levered, and lower-return.

3. Industry Dynamics

A genuine multi-year demand recession, in a no-growth category. The relevant pool — the US mattress market — has been shrinking for three-plus years. After a 2021 COVID-stimulus peak (~$10.9B wholesale), ISPA data show 2023 unit shipments of 39.7 million, down 8.0%, on $9.8B wholesale (−6.8%), followed by 2024 units −5.7% and dollar sales −5.9%. This is among the worst multi-year contractions the category has recorded — driven by the collapse in housing turnover, big-ticket discretionary weakness, and payback from the pandemic pull-forward. Critically, there is no secular unit growth to fall back on: long-run demand tracks household formation and replacement — low-single-digit at best. On the Q1 2026 call management pegged global bedding demand down mid-single digits, worse than its own flat-to-slightly-positive expectation, and set full-year 2026 industry assumptions at “flat to slightly down.”

Structure and profit pools. The FY2025 10-K describes it plainly: the domestic manufacturing market is concentrated, the domestic retail market is fragmented, and international is highly fragmented and regional. Even at the top, concentration is modest — Tempur Sealy holds ~13% brand share, with Spring Air and Simmons near 10% each, and the top-five manufacturers only ~22% combined. Specialty stores are ~43% of the distribution channel, where Mattress Firm dominates. Post-merger, SGI’s combined manufacturing-plus-captive-retail position is cited above ~32% — a genuine shift from fragmentation toward a dominant vertically-integrated player, which is the whole strategic thesis.

The Serta Simmons collapse and the DTC bust. Two forces reshaped the field. First, Serta Simmons Bedding — historically the #1/#2 volume player — filed Chapter 11 in January 2023, cutting funded debt from ~$1.9B to ~$300M (its notorious “uptier” was later ruled a credit-agreement violation by the Fifth Circuit in 2025). A former leader emerged smaller and weakened. Second, the direct-to-consumer disruption has largely broken on its own economics: Casper was taken private at ~$300M in 2022 (versus a $1.1B IPO) and faded; Purple runs chronic losses (Q2’25 net loss $17.3M, adjusted EBITDA still negative); Sleep Number saw sales fall ~20% in 2025 with net losses and $80–100M of emergency cost cuts. The DTC playbook — high customer-acquisition cost, generous returns, no store to touch-and-feel — could not out-earn incumbents’ scale advertising and omnichannel reach. The lesson: online lowered barriers to entry but not to profitability.

Regulation — antidumping duties as a real but porous tailwind. In 2021 the US imposed AD/CVD orders on mattresses from eight countries, with China antidumping margins as high as ~1,732% and CVD ~98%. This materially raised the landed cost of imports and protects domestic manufacturers like Tempur Sealy — a genuine structural positive. But it is whack-a-mole: sourcing simply migrated to non-covered geographies (Indonesia became the #1 exporter in 2021; then Mexico/Brazil), so the protection is porous and requires repeated petitions to sustain. It helps the domestic profit pool at the margin; it does not wall the category off.

Marathon capital-cycle read. The supply side looks like a classic bust/consolidation trough: capacity is exiting (Serta’s bankruptcy and plant closures, Casper’s demise, the DTC shakeout), the weakest players are failing, and disciplined survivors are consolidating (SGI’s roll-up). Under Marathon’s capital-returns framework, a rationalizing supply side into a demand trough is precisely the setup that precedes a returns recovery — if demand normalizes. The crucial caveat: this is a cyclical demand trough in a no-growth category, not a secularly expanding profit pool. Capital is exiting because the economics are poor, not merely because the cycle is down. And SGI’s own behavior — a debt-and-equity-funded acquisition binge — is exactly the asset-growth signal Marathon flags as a headwind to forward returns, even as it consolidates the field.

Verdict: structurally below-average industry. Bedding is fragmented at the brand level, has zero consumer switching costs, a commoditizing value tier (private label, Costco, Amazon, Walmart), tight cyclicality tied to housing, no secular unit growth, and thin through-cycle returns. Consolidation and AD/CVD protection modestly improve the structure at the trough and create a cyclical-recovery opportunity — but they do not convert bedding into a good industry. It is a mediocre neighborhood in which one house can still be better-built than the rest.

4. Competitive Position

The two Greenwald tests, applied. Market-share stability — the category fails the strong-barrier test. Over 5–8 years shares moved well beyond Greenwald’s 5-point “no-barriers” threshold: Serta collapsed into bankruptcy, DTC brands surged then faded, and Tempur Sealy gained. Only the very top of the price ladder (Tempur-Pedic’s premium ASP tier) shows durable stickiness. Profitability — SGI’s ROIC has fallen every year, 24.3% (2021) → 16.2% → 13.2% → 11.0% → 8.1% (2025), converging on its cost of capital. Some of that is the trough and some is deliberate margin-dilutive M&A, but the trailing path fails Greenwald’s “sustained 15–25%+” franchise bar. Pre-Mattress Firm, Tempur Sealy earned high-teens-to-low-20s ROIC; the roll-up of low-return, lease-heavy retail has diluted it.

Moat mechanisms — naming each and pressure-testing it.

1. Tempur-Pedic brand (intangible / demand advantage) — real but narrow. Tempur-Pedic commands genuine pricing power and consumer recall in the premium specialty-foam niche; it is the closest thing in bedding to a true franchise. But the advantage is narrow and structurally leaky: mattresses are bought roughly once a decade, so habit — Greenwald’s most durable demand mechanism — barely operates, and there are no switching costs to retain the customer between purchases. It must be re-won at every purchase, in a category where the shopper actively cross-shops. Powerful at the top; it does not extend down the ladder to Sealy (a value/volume brand competing largely on price).

2. Manufacturing and advertising scale (cost advantage) — real, but requires captivity it doesn’t have. SGI is the largest US manufacturer and spends at “industry-leading levels” on national media ($692M advertising in 2025). Scale in production and marketing is real. But Greenwald is emphatic that scale is only a barrier when paired with customer captivity — and this category has little. The absence of captivity is exactly how lightly-capitalized DTC entrants reached national scale cheaply a decade ago. One genuine mitigant, unusual to this cycle: a shrinking market raises fixed-cost-per-unit for subscale rivals, so scale bites harder in a downturn — here, market contraction is the incumbent’s friend, squeezing out weaker competitors.

3. Mattress Firm distribution control (the contested “vertical moat”) — more reach than moat. Owning the #1 specialty retailer (~43% of the specialty channel, 2,300 doors) guarantees shelf and slot access and captures retail margin. The skeptical read is that this is scale/reach and defensive channel control, not a durable barrier, for three reasons: (a) the antitrust remedy forces 43% of premium slots to third-party rivals, deliberately self-limiting the exclusivity that would make it a moat; (b) it creates channel-conflict and antitrust exposure — competitors still sell through a competitor-owned retailer, and the LEG deal invites vertical-foreclosure scrutiny; and © e-commerce structurally erodes the value of controlling physical doors. Vertical integration here grosses up revenue at low retail margins and defends the shelf; it does not manufacture pricing power.

Switching costs — essentially zero. This is the crux. There is no ecosystem, subscription, data lock-in, or network effect. Each purchase is a fresh, price-and-feel-driven decision made once every ~decade. No amount of vertical integration changes the fact that the category’s customer captivity is close to nil — the structural ceiling on any bedding moat.

Versus the field. SGI is unambiguously the strongest player: Sleep Number (differentiated smart-bed and vertically integrated DTC, but sales −20% and loss-making in 2025); Purple (niche gel-grid tech, chronic losses); Serta Simmons (post-bankruptcy, weakened); Casper (effectively dead); and the value tier increasingly commoditized by private label, Costco, Amazon and Walmart. SGI has the best brand set, the strongest balance sheet in the category, and the only owned national retail footprint. That relative strength is real and durable enough to keep it #1.

Verdict: a narrow, eroding advantage in a low-moat category — best house on a mediocre street. SGI possesses one genuine premium brand (Tempur-Pedic) and real manufacturing/advertising scale, but both sit inside a no-switching-cost, no-growth, cyclically-depressed category, and the Mattress Firm/LEG vertical integration is a scale-and-channel-defense play — capped by antitrust remedies — rather than a wide moat. The five-year ROIC decline from 24% to 8% is the market’s honest verdict: the advantage is narrow and under pressure. Durable enough to stay the category leader; not durable enough to earn franchise-level returns through the cycle.

5. Growth History and Forward Opportunities

The headline growth is an accounting event, not a demand event. FY2025 consolidated revenue rose 51.6% to $7,476.5M — but $3,505.4M is eleven months of Mattress Firm retail, a pure gross-up of a business SGI now owns instead of merely supplying. Reading real organic demand requires unwinding two consolidation distortions at once. First, the gross-up: MF adds ~$3.5B of retail revenue. Second, and subtler, the elimination: Tempur Sealy’s wholesale sales to Mattress Firm (management sizes intercompany at ~23% of global Tempur Sealy sales) are now netted out. That is why reported Tempur Sealy segment sales fell — North America $2,701.2M + International $1,269.9M = $3,971M in 2025 versus $4,930M for the whole pre-MF company in 2024. That decline is an artifact of intercompany elimination, not lost demand.

The clean organic read is mixed-to-soft. Management’s like-for-like disclosure for Q1 2026:

True organic SGI growth is roughly flat-to-modestly-positive, achieved by taking share (advertising share-of-voice, a stronger balance sheet than distressed competitors, price discipline) in a down market — not by a growing end market. The −12% direct/e-commerce and −4% third-party-retail lines show demand is genuinely weak where SGI can’t lean on its own captive retail.

Forward drivers — real but incremental, and partly self-cannibalizing. (1) Mattress Firm optimization: a $150M store-refresh/brand-wall program through 2027 ($40M spent), migrating MF’s assortment toward Tempur Sealy brands/private label (guided to low-60% of MF sales, ~$40M incremental EBITDA in 2026) — margin mix, capped by MF’s own flat-to-down traffic. (2) Stearns & Foster relaunch (H2 2026) into higher price points — but management concedes it “cannibalized” Sealy Posturepedic to make room, so net addressable-market gain is smaller than the launch noise implies. (3) Pricing: a ~4% list increase (~$100M annualized lift) to offset ~$100M of commodity inflation — explicitly dollar-neutral to earnings, so it lifts reported sales without adding profit. (4) International: the genuinely highest-quality growth line — mid-single-digit constant-currency, disciplined distribution expansion, Dreams (UK) share gains. (5) LEG synergies: only ~$50M run-rate (~$10M year one) on a ~$1.7B combined-EBITDA base — immaterial near term.

Verdict: low-quality growth. The 50%+ revenue growth is consolidation optics; underlying organic demand is flat-to-down in a category in a multi-year unit recession. What growth is real (International, share gains, profit-neutral pricing) is incremental and partly offset by cannibalization and dilution — diluted share count rose from 173.6M to 209.2M (2025) to a guided ~213M (2026) on the MF stock consideration, so 51.6% dollar growth translated to only +5.9% adjusted EPS in 2025. SGI is buying revenue and share at a cycle trough; it is not compounding organic demand.

6. Financial Quality

The GAAP-to-adjusted gap is unusually wide, and it is made of cash costs — the single most important QoE flag in this name. FY2025 GAAP diluted EPS was $1.84 (down 14.8% YoY); adjusted EPS was $2.70 (up 5.9%). The $0.86/share bridge — 47% of GAAP EPS — is not the usual non-cash purchase-accounting amortization investors readily forgive. It is dominated by cash deal and integration spend:

¹ Includes purchase-accounting inventory step-up (the driver of the depressed Q1 2025 gross margin). Per-share on 209.2M (FY2025) / ~212.6M (Q1 2026) diluted shares.

Two things stand out. Positively, SGI does not add back a large standalone “amortization of acquired intangibles” line — cleaner than most PE-style roll-ups. But ~$194M of the FY2025 pre-tax add-backs are real cash outflows. Calling them “one-time” is defensible for the MF close, but the LEG acquisition (targeted to close by year-end 2026) guarantees another multi-year wave of the same “adjusted-away” cash integration spend. This is a serial acquirer whose “adjusted” numbers structurally exclude the recurring cost of acquiring. Value on adjusted EPS as the operating run-rate — but haircut it for the fact that the deal machine keeps running.

Margins — the compression is two-thirds mix, one-third one-time. Operating margin fell from ~18% (2021) to a GAAP 10.1% in FY2025, but adjusted operating margin held at 13.6% ($1,018.7M). The quarterly path shows the one-time drag clearly:

Consolidated gross margin actually rose to 44–45% in the clean 2025 quarters versus ~42% pre-MF; the FY2025 42.6% is dragged down almost entirely by the Q1 2025 inventory step-up. The honest structural read: adding MF’s retail layer lifts blended gross margin (retail markup) but is operating-margin dilutive (store rent and labor), and adds operating-deleverage risk into a soft-demand cycle.

Interest burden is now a structural headwind. Net interest expense hit $261–268M in FY2025 (up from ~$100M pre-MF) — the ~$2.8B cash portion of the MF consideration was debt-funded. 2026 guidance is ~$230M as SGI deleverages toward its 2–3x target (3.1x at Q1 2026). The LEG deal, being all-stock with LEG bringing its own EBITDA, actually lowers pro-forma net leverage — the one clearly shareholder-friendly structural feature of that transaction.

FCF quality — the headline overstates true owner cash. FY2025 OCF $800M − capex $167M = ~$633M “FCF.” Three haircuts: (1) capex was under-run — 2026 guidance is ~$225M (incl. $75M MF store refresh), normalizing to ~$200M; (2) the MF fleet is ~$2B of finance/operating leases whose principal repayment lands in financing, so it never touches the FCF line despite being a hard cash cost of running the retail estate; (3) working capital swings violently by season. Normalized owner FCF — after ~$225M capex and lease-principal amortization — is materially below the headline, closer to ~$450–500M, or a ~3% FCF yield on the ~$16B cap.

ROIC is the whole thesis in one number. ROIC fell 24.3% (2021) → 8.1% (2025) as invested capital ballooned — $4.6B goodwill + $2.6B intangibles + ~$2B capitalized leases — far faster than NOPAT. Even normalizing out the one-time deal costs, ROIC is only ~10–11%, near WACC and a fraction of the pre-MF peak. Tangible common equity is deeply negative ($7.18B goodwill+intangibles vs $3.12B total equity; TCE ratio ~−92%).

Verdict: scale is diluting returns, not improving them. The high-return, capital-light Tempur-Pedic manufacturer (24% ROIC) has been wrapped in a levered, lease-heavy, lower-margin retail roll-up that has cut ROIC to ~8% GAAP / ~10–11% normalized, doubled the interest bill, driven TCE negative, and requires a ~$0.86/share stack of cash “adjustments” to present a clean number. Economics do not improve with this flavor of scale; vertical integration is buying revenue, share, and channel control at the explicit cost of return on capital.

7. Capital Allocation

Somnigroup is, at its core, a serial acquirer of a structurally weak, cyclical category — and its capital-allocation record must be judged as a roll-up, not a compounder. The arc runs from the 2012–13 acquisition of the Sealy brand, to the ~$4–5B cash-and-stock purchase of Mattress Firm (closed Feb 5, 2025; $2.82B of cash left the 2025 cash-flow statement plus ~34M SGI shares), to the pending all-stock Leggett & Platt deal signed April 13, 2026. Each step buys further vertical control; each has been financed by some combination of leverage and equity issuance at a demand trough. That is the central tension: the strategy is coherent and the timing defensible (buying distressed assets cheap), but the funding has quietly reversed years of shareholder-friendly behavior and driven returns on capital down, not up.

The buyback history makes the pivot vivid. In its pre-deal era the company was an aggressive repurchaser — roughly $801M in FY2021 and $621–667M in FY2022 (~$1.4B over two years, at ~$30–48/share, excellent value against today’s ~$78). It then paused almost entirely (~$5M in 2023, $43M in 2024) to marshal capital for Mattress Firm, and did only ~$132M in FY2025. Worse for per-share value, the diluted share count rose from ~173.6M to ~206M on the MF stock consideration — the company spent years shrinking the count, then re-inflated it in a single deal, and the LEG transaction issues ~20M more shares (~10% dilution). Management bought back low and is now issuing equity for acquisitions; that sequencing is the opposite of what a disciplined capital allocator wants.

The rest of the toolkit is conservative-to-unremarkable. The dividend (~$0.62/share, ~$127M in FY2025, ~33% payout) is modest and well covered. SBC ($36–53M/yr, ~0.5% of sales) is not a leakage problem. R&D is trivial ($29–32M/yr, ~0.5% of sales); the real “moat spend” is advertising — $448–470M/yr on the manufacturing side, jumping to $692M in FY2025 as MF’s retail marketing consolidated in. This is a brand-and-distribution model, not an innovation model — fine, but the intangible being protected is share-of-voice, which competitors can also buy.

The leverage is where we push back hardest. FY2025 net debt is $4.55B (including ~$2B of MF leases); goodwill + intangibles of $7.18B tower over $3.12B of total equity, leaving tangible common equity deeply negative; and the LEG deal layers on LEG’s ~$0.91B net debt to push pro-forma net debt toward ~$5.5B. None of this is fatal for a cash-generative business (OCF ~$800M, FCF ~$633M), but it is a balance sheet with no margin for a deep or prolonged mattress recession — and mattresses are a discretionary, big-ticket, housing-correlated purchase near a multi-year volume trough.

The governance tell ties it together. The 2026 DEF 14A keys incentive comp to adjusted EPS, adjusted EBITDA, Relative TSR percentile, and a “strategic initiatives / ESG” component (the latter repeatedly paid at the 300% cap — soft); adjusted EPS was added in 2025. There is no ROIC or return-on-capital metric anywhere in the plan. For a company whose ROIC has fallen from 24.3% to 8.1% while it levers up to buy ever-larger businesses, paying management on growth in EPS/EBITDA and TSR but not on the return earned on the capital deployed is precisely the incentive that rewards empire-building. CEO Scott Thompson’s 2025 “compensation actually paid” of ~$45.9M (stock-inflated) is a symptom; the plan design is the disease.

Verdict: mixed, tilting skeptical. Management is not reckless — the dividend is safe, SBC and R&D are contained, and buying distressed vertical assets at a trough can be intelligent counter-cyclical allocation if the cycle turns. But the honest read is that a disciplined repurchaser has become a debt-and-equity-funded roll-up with negative tangible equity, deteriorating returns on capital, and a comp plan that doesn’t police those returns. Capital has been deployed ambitiously, not obviously allocated intelligently — the ROIC trend is the referee, and it is going the wrong way.

7a. The Leggett & Platt Deal (Special Situation)

Terms (Fact). All-stock, fixed exchange ratio 0.1455 SGI per LEG share, worth ~$11.36/LEG at the April 10, 2026 SGI close (~$78) — a 13.7% premium to LEG’s $9.99. SGI first went public with a $12.00/share proposal on Nov 28, 2025 (LEG then $10.26); the fixed ratio finally signed is worth less per share than that opening headline because SGI’s own stock fell from ~$92 to ~$78 into signing — LEG holders bear SGI’s downside. S-4 purchase accounting: total consideration $1,517M, cash acquired $511M, net consideration ~$1,006M, plus assumption of LEG’s debt (~$1.5B gross / ~$0.91B net). LEG needs a two-thirds shareholder vote. Break fees: SGI→LEG $80M; LEG→SGI $64M; plus a reverse termination fee of ~7% of LEG equity if regulatory approval fails.

Strategic rationale (Fact + skeptical Interpretation). LEG’s Bedding Products segment already supplies the “substantial majority” of Tempur Sealy’s innersprings under a long-term agreement expiring July 31, 2028. The pitch is textbook vertical integration: own the component, tighten design collaboration, capture supplier margin, secure supply. The skeptical read: SGI is buying a distressed, cyclically-depressed steel-and-spring business — LEG cut its dividend ~89% (to $0.05) in 2024, breaking a 50±year “Dividend King” streak, ran a Bedding-led restructuring, and posted FY2025 sales −7%. It is cheap (~5–6x depressed EBITDA vs SGI’s own 18.7x) and deleveraging (all-stock; the combined leverage falls), but the “E” being bought is at a trough and the acquired business is structurally lower-margin and more commodity/steel-exposed than SGI’s brand assets. SGI has not quantified accretion; the S-4 explicitly warns EPS “may be diluted” and synergies (~$50M run-rate, ~$10M year one) may not materialize. Best characterization: near-term EPS-neutral-to-modestly-accretive, contingent on a bedding-volume recovery and integration — the real bet is cyclical timing plus vertical control, not immediate earnings.

The antitrust question — the crux. SGI would own the #1 mattress manufacturer, the #1 retailer (Mattress Firm), and a key components supplier that also sells to SGI’s rival manufacturers (Serta Simmons, Purple, etc.). That is the third leg of a vertical stack and opens a clean input-foreclosure theory — the mirror of the retail-foreclosure theory the FTC ran on Mattress Firm. Two things cut the other way, and they are structural. First, the deal is built for antitrust risk: (i) a Special Dividend — if a DOJ/FTC proceeding is commenced to block/delay, LEG pays holders 6% of LEG’s 30-day VWAP × (days-of-proceeding ÷ 365), a ticking fee compensating for delay; (ii) the reverse termination fee; and (iii) — most telling — a limited regulatory-efforts covenant, explicitly NOT hell-or-high-water: “Somnigroup will not be required” to divest or accept structural remedies. SGI has pre-negotiated the right to walk rather than be forced into a divestiture. Second, the Mattress Firm precedent is directly on point and favorable: the FTC voted 5-0 to challenge that vertical deal (Jul 2024), and lost — the SDTX (Judge Eskridge) denied the preliminary injunction Jan 31/Feb 4, 2025, the FTC declined to appeal, the deal closed, and the administrative case was withdrawn. US HSR clearance on the LEG deal expired/was granted June 3, 2026, removing the US overhang; remaining approvals are foreign (EU/UK/Korea/Canada competition, Austria FDI) plus the LEG shareholder vote and S-4 effectiveness.

Deal-completion read (Interpretation). On balance, completion is more likely than not, but with a materially fatter antitrust tail than a normal strategic deal — the vertical stack is genuinely aggressive and the LEG-supplies-rivals fact is a live foreclosure hook. Timeline is the real cost: foreign reviews and any contested step could push close into 2027, and the Special Dividend / reverse-fee architecture tells you both boards already price a fight as plausible. The deal is strategically logical and cheaply priced, but it is the highest-risk of the three roll-up legs — a leveraged (via LEG’s own debt), all-stock bet that stacks the entire bedding value chain into one entity right as antitrust attention on vertical consolidation and the mattress cycle are both live.

7b. Insider Transactions

The Form 4 corpus (~200 filings, 2021–2026) reads neutral-to-mildly-negative. There have been only two open-market purchases (code P) in five years, both by director Simon Dyer on Dec 4, 2025 — ~32,000 shares at ~$93–94 (~$3.0M total, via Madad Investment Pty Ltd), immediately after the Nov 2025 LEG proposal; a genuine conviction buy, but small and now underwater versus ~$78. CEO Scott Thompson has done no open-market buying in the window — his activity is deep-in-the-money legacy-option exercises with tax withholding (retaining net shares, a mild positive) plus large charitable/estate gifts (code G, incl. 326k shares Nov 2025). Other officers show routine, discretionary (not 10b5-1) option-and-sell activity (CFO Bhaskar Rao ~$7.5M Aug 2025). The May 2026 Form 4 “cluster” is a non-event — routine annual director equity grants. Net: a single director’s ~$3M buy against a steady flow of grants, exercises, gifts and discretionary sells, with no CEO or CFO open-market accumulation.

8. Changes and Headwinds — Last Two Years

The Mattress Firm close is the dominant change — it converted a branded manufacturer into a vertically integrated manufacturer-plus-retailer, overriding a unanimous FTC block only because a federal court refused the injunction. That is a strategy pursued against the antitrust regulator’s stated judgment, and the vertical-foreclosure concerns the FTC raised are now partly visible in the disclosures (MF steering assortment to Tempur Sealy brands, guided to low-60% of MF sales; price discipline across third-party retailers). The LEG deal doubles down on the same logic at a cycle trough, but is structured shareholder-friendly (all-stock, deleveraging, EPS-accretive before synergies in year one). Guidance is quietly softening at the top line even as EPS holds: 2026 adjusted-EPS guidance is $3.00–3.40 (midpoint +18.5% over $2.70), but the sales midpoint was trimmed from ~$7.9B to ~$7.8B at Q1 on a weaker industry, with a soft Q2 flagged (EPS +5–10%, transitory commodity drag) and recovery back-half-loaded on price. Leadership is stable (Scott Thompson Chair/President/CEO; Bhaskar Rao CFO).

Verdict: net weakens the thesis, with one offset. In two years SGI has doubled its balance-sheet risk, driven ROIC to trough, doubled interest expense, and committed to a second integration-heavy roll-up — all at a demand trough and over regulatory objection, while true organic demand is flat-to-down. The offset: these are countercyclical, control-enhancing, share-taking moves that could pay off if the cycle normalizes and the pieces integrate cleanly, and the LEG deal in particular is structured to deleverage. But the changes trade quality (returns, tangible equity, clean earnings) for scale and vertical control, and raise, not lower, the execution and antitrust risk profile.

9. Risk Analysis

SGI is a levered, cyclical roll-up with deeply negative tangible equity, trading at cyclically-high multiples on trough earnings, in the middle of two simultaneous large integrations. The risk profile is dominated by financial leverage, integration execution, and category cyclicality — antitrust, notably, has de-risked on the US front (HSR cleared Jun 3, 2026) even as foreign clearances remain.

Catastrophic-loss read. A total loss is low-probability: the brands (Tempur-Pedic, Sealy, Stearns & Foster) and the #1 retail footprint have real going-concern value, the LEG deal is all-stock (no incremental cash leverage), and FCF (~$633M) comfortably services debt at current rates. But the combination of high financial leverage + negative tangible equity + trough cyclical earnings + a debt-cycle-timed roll-up makes the equity more fragile than the franchise: a deeper or longer mattress recession would compress EBITDA against a fixed ~$6.5B debt/lease load and a cyclically-high multiple simultaneously — the same one-two (earnings down + multiple down) that drove the −59% move in 2022.

10. Valuation Discussion (Embedded Expectations)

The headline “cheap” signal is an accounting artifact; the honest reads say the opposite. SGI screens at P/S 2.17x — the 7.6th percentile of its own history — but that percentile is manufactured by the Mattress Firm consolidation, which grosses up revenue by ~$3.5B of low-margin (33% gross) retail sales without a commensurate profit increase. Sales roughly doubled while operating margin fell from ~13% to ~10%, so a sales-based multiple mechanically collapses. Strip the distortion and the earnings-based multiples are where the truth lives: P/E ~32x (34.7th percentile own-history) on trough TTM EPS of $2.45, and EV/EBITDA 18.7x on an EV of ~$21.9B (correctly including ~$2B of Mattress Firm leases). Neither is cheap — both sit in the upper half of the company’s five-year range.

Own-history trend. SGI’s year-end EV/EBITDA has re-rated from 11.9x (2021) → 11.8x (2022) → 16.6x (2023) → 18.2x (2024) → ~25x (2025, at the $89 year-end high); today’s 18.7x is roughly the 2024 level and well above the 2021–22 cyclical-trough multiples of ~12x. P/E has walked from ~13–15x (2021–22) to ~24–26x (2023–24) to ~32x today. The multiple expanded as the market re-rated the consolidation story — meaning the stock has already been paid for the roll-up, at trough earnings. An investor at $78 is paying a re-rated multiple and underwriting an earnings recovery.

Comp set (approximate public figures; mattress-pure peers are distressed, so their EBITDA multiples are noisy):

SGI trades at roughly a 2x EV/EBITDA and P/E premium to every solvent furniture/bedding peer. Some premium is defensible — clear #1, the only genuine premium brand, best balance sheet in a field of distressed players. But ~19x EV/EBITDA on trough earnings, for a no-secular-growth cyclical, is a rich starting point, not a value entry.

Embedded-expectations analysis — what does $78 require? Two lenses converge: the price already discounts a substantial mattress-cycle normalization plus synergy capture.

• On EV/EBITDA: to justify 18.7x today and eventually de-rate toward a mattress-cyclical ~11–12x while holding the price, EBITDA must rise roughly 55–70% from the TTM ~$1.17B base — the market is underwriting a full demand recovery and MF/LEG synergy realization.
• On P/E: $78 against a plausible mid-cycle normalized EPS of ~$4.00–4.50 (roughly pre-MF-dilution earnings power restored, plus partial synergies) implies ~17–18x forward — a normal multiple, but only if that mid-cycle recovery arrives. On the trough $2.45, 32x is expensive; the gap between the two is precisely the recovery the buyer is paying for.

Scenario analysis (illustrative; EPS × multiple; embedded expectations only — no price target, no recommendation):

The base case lands near the current $78 — the crux: at today’s price the market pays a fair price for a successful mid-cycle outcome, leaving little margin of safety if the trough persists (bear) and requiring near-flawless execution + a demand rebound to reach the bull. Consensus (a “Strong Buy” tilt, an average target in the ~$90s) sits between base and bull — the Street is underwriting the recovery-plus-synergies path.

11. Variant Perception

Consensus view. The sell side is firmly constructive — a “Strong Buy” consensus and an average target in the mid-$90s. The bull narrative is clean and widely held: the dominant, vertically-integrated bedding leader is buying its #1 retailer and #1 components supplier at the trough of a mattress cycle, and earnings will snap back as demand normalizes while synergies compound. The re-rating from ~12x to ~19x EV/EBITDA over three years shows the market has already substantially embraced this story.

Strongest bull case. SGI is a cycle-trough consolidator with genuine assets: the only real premium brand in bedding (Tempur-Pedic), the largest US manufacturing/advertising scale ($692M ad spend), and now ownership of the #1 specialty retailer (~43% of the specialty channel). Mattress units are at multi-decade lows after three down years — a coiled spring on any housing/rate normalization — and the field has been cleared (Serta bankruptcy, Casper dead, Purple/Sleep Number loss-making), leaving SGI the last well-capitalized player standing to harvest the recovery. MF and LEG add retail margin and component profit to every mattress sold; synergies plus a demand rebound could drive mid-cycle EPS toward ~$5, and a resumed buyback (paused for deleveraging) compounds it. US antitrust just cleared.

Strongest bear case. SGI is a levered roll-up buying troubled assets in a secular-demand-challenged category near the top of a debt cycle. The tape agrees with the bear on what the stock is: FactorsToday reads SGI as a rate-sensitive housing/building-products cyclical (Home Construction industry loading +0.63; InterestRate −0.30; Growth −0.27), and its factor-similar peers are Griffon, Owens Corning, Fortune Brands, Stanley Black & Decker, UFP, Simpson, Boise Cascade — building-products cyclicals, not staple compounders. That classification clashes with the “quality compounder” framing the multiple implies. Underneath: ROIC has fallen every year, 24.3% → 8.1%, toward WACC; tangible common equity is deeply negative; the balance sheet carries ~$6.5B of debt/leases; and the “cheap” P/S is a consolidation mirage. The vertical “moat” is capped by an antitrust remedy and creates channel conflict. Buying LEG — itself a low-return, levered components maker — into a no-growth category is a Marathon-style asset-growth-anomaly red flag: growth that dilutes returns. At 32x trough earnings / 18.7x EBITDA, the price already pays for a recovery the category’s structural flatness may not deliver.

The 3–5 assumptions that matter most:

1. Does the US mattress cycle actually recover, and when? Everything — EPS, the multiple, the roll-up math — rests on units rebounding from multi-decade lows. If the trough is structural, the thesis fails.
2. Do MF + LEG synergies materialize net of integration cost and channel conflict? Two large, dissimilar integrations in parallel; the vertical benefit is remedy-capped.
3. Is 8% ROIC the new normal or a trough? If the roll-up permanently dilutes returns toward WACC, no multiple re-rating is warranted regardless of demand.
4. Is the market right to price SGI as a housing cyclical (beta ~1.16) rather than a staple compounder? The multiple says compounder; the factor loadings and peer set say cyclical. Both can’t be right.
5. Does the LEG deal close on terms, and does leverage stay serviceable through a prolonged trough?

Falsification tests. Bull is falsified if: US mattress units post a 4th consecutive down year, MF/TSNA segment margins fail to inflect, or ROIC stays sub-10% into 2027 — proving the trough is structural and the roll-up return-dilutive. Bear is falsified if: unit volumes inflect positively, consolidated ROIC recovers toward the mid-teens, synergies show up in expanding operating margin, and the buyback resumes without leverage stress — proving it a genuine cycle-trough quality compounder.

Factor-positioning read (input, not a call). Momentum is rolling over: rs_12m positive (+13.5) but rs_6m negative (−12.8), ~20% off the ATH, with a −22%/yr 6-month slump partly retraced by a +35%/yr 3-month bounce off the May-2026 $61 low. Market R² is only ~47% (specific vol ~28% annualized) — roughly half of SGI’s moves are idiosyncratic (deal/earnings-driven), not market beta. Net: the tape prices SGI as a beta-~1.16 housing-linked cyclical with a mild quality (brand) tilt and a negative growth loading — an evidence-based reason to doubt that the re-rated, compounder-like multiple is the right lens. This is where consensus may be offsides: paying a quality multiple for what the stock’s own factor signature says is a rate-sensitive cyclical.

12. Fact vs. Interpretation Table

13. Open Questions

1. What is normalized mid-cycle EPS, really? Our ~$4.00–4.50 estimate hinges on how much of the pre-MF earnings power is restored net of permanent dilution and interest — management has not framed a clean mid-cycle bridge.
2. When (if) do US mattress units inflect? Three-plus down years is already historically long; the timing of the turn is the single largest swing factor and is unknowable from the inside.
3. How much of the ~$194M FY2025 “adjustments” recurs annually while the deal machine runs? The LEG integration guarantees more; the steady-state normalized number matters for valuation.
4. Does a DOJ/FTC Second Request land on the LEG deal, and how long does any foreign review take? The Special-Dividend clock and reverse fee tell us both boards price a fight as plausible.
5. What is true owner FCF after MF lease-principal amortization? Our ~$450–500M estimate needs the finance-lease principal schedule to firm up.
6. Will management resume buybacks — and at what leverage? The pre-deal repurchase machine created value; its return (and the price paid) is the clearest capital-allocation signal to watch.

14. What Must Be True

Bull case — what must be true, and its falsification test.

• US mattress unit demand inflects from its multi-decade-low trough within ~1–2 years; SGI’s organic volumes and MF same-store sales turn positive.
• The Mattress Firm and Leggett & Platt integrations deliver their margin-mix and synergy benefits net of cost and channel conflict; consolidated ROIC climbs back toward the mid-teens.
• The LEG deal closes on terms without a forced divestiture; leverage stays serviceable and a buyback resumes.
• Falsification: a fourth consecutive down year in US mattress units, MF/Tempur Sealy North America margins failing to inflect, or ROIC stuck below 10% into 2027 — any one proves the trough is structural and the roll-up return-dilutive.

Bear case — what must be true, and its falsification test.

• The mattress category is structurally flat-to-declining, not merely cyclically depressed; 8% ROIC is the new normal as the roll-up permanently dilutes returns toward WACC.
• The re-rated ~19x EV/EBITDA / ~32x trough-earnings multiple compresses toward the cyclical ~11–12x as the market re-classifies SGI as the housing cyclical its factor loadings say it is.
• Leverage + negative tangible equity + a prolonged trough force a goodwill impairment and/or squeeze the equity.
• Falsification: unit volumes inflect positively, consolidated ROIC recovers toward the mid-teens, synergies show up in expanding operating margin, and the buyback resumes without leverage stress — proving SGI a genuine cycle-trough quality compounder rather than a debt-cycle roll-up.

15. Source Appendix

See Appendix B (Source Appendix) below for the full, dated source list. Primary sources include: Somnigroup FY2025 Form 10-K (filed 2026-02-27) and FY2021–2024 10-Ks; Q1 2026 Form 10-Q (filed 2026-05-08); FY2025 and Q1 2026 earnings press releases (PR Newswire, 2026-02-17 and 2026-05-07); the Leggett & Platt merger announcement (PR Newswire, 2026-04-13); the Form S-4 registration statement (filed 2026-06-24) and related 425 communications; the 2026 DEF 14A proxy; the SEC Form 4 corpus (CIK 1206264); FTC matter 231-0016 (Tempur Sealy/Mattress Firm) and the SDTX preliminary-injunction ruling; ISPA / BedTimes industry data; USITC / trade.gov AD-CVD determinations (2021); ROIC.ai fundamentals and valuation data; AZI valuation-index percentiles and 5-year price CSV; and the FactorsToday factor model. Facts reconcile to primary filings; third-party aggregated data (ROIC.ai, AZI, FactorsToday) are labeled as such and used as cross-checks, not as primary authority.

\pagebreak

APPENDIX A — Standard Diligence Questionnaire

Somnigroup International, Inc. (NYSE: SGI) — as of July 3, 2026. Supplemental to the main article. Fact / Interpretation / Assumption labels applied where they matter.

General

What thoughtful questions have other investors asked about this company? The central debate is whether SGI is a cycle-trough quality consolidator (bull) or a levered roll-up buying troubled assets in a no-growth category (bear). Recurring investor questions: (1) how much of FY2025’s 51.6% revenue growth is real vs. Mattress-Firm consolidation optics (Fact: almost all is consolidation); (2) whether the ~$0.86/share adjusted-vs-GAAP gap is truly one-time (Interpretation: largely recurring while the deal machine runs); (3) whether owning the #1 retailer plus the #1 components supplier is a durable vertical moat or an antitrust liability; (4) whether ROIC’s 24%→8% slide is a trough or the new normal; and (5) whether the LEG deal closes without a forced divestiture.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? A cyclical low (Fact). US mattress units have fallen three-plus consecutive years (2023 −8.0%, 2024 −5.7%) from the 2021 stimulus peak, to near multi-decade lows; TTM EPS $2.45 and GAAP $1.84 are trough-level. Adjusted EPS $2.70 (2025) with 2026 guidance $3.00–3.40.

Driven by the external environment or internal actions? Both. The external cycle (housing turnover, rates, big-ticket discretionary) drives volume; internal actions (the MF/LEG roll-up, share-of-voice advertising, price discipline) drive share gains and mix. Growth today is internal share-taking in a shrinking pie.

How stable are revenues? Low stability — bedding is discretionary, big-ticket (replaced every ~8–10 years), housing-correlated, with zero recurring revenue (franchise royalties ~$39.5M, ~0.5% of sales). Essentially 100% must be re-won each cycle.

Outlook for products/services; how big will this market be? The US mattress pool (~$10B wholesale) is a no-secular-growth category tracking household formation + replacement (low-single-digit at best). Near-term flat-to-down; the investment case rests on cyclical recovery, not secular expansion. International is the higher-growth line (mid-single-digit constant-currency).

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Less crowded at the top after consolidation (Serta Ch.11 2023; Casper dead; Purple/Sleep Number loss-making) — but the value tier is more commoditized (private label, Costco, Amazon, Walmart). Net: consolidating but structurally low-moat.

How profitable is the business (ROIC, ROE)? Deteriorating (Fact): ROIC 24.3%→8.1% (2021→2025), ~10–11% normalized — near WACC; ROE 26.8%→10.4%. Pre-MF the manufacturer earned high-teens-to-low-20s ROIC; the retail roll-up diluted it.

How profitable is the industry — competitors, barriers to entry? Thin through-cycle returns; low barriers (DTC entrants reached national scale cheaply a decade ago). Modest concentration (top-5 manufacturers ~22% share). AD/CVD duties protect domestic manufacturers but are porous.

Can the business be easily understood? Yes — make and sell mattresses; own the stores that sell them. The complexity is financial (leverage, consolidation optics, adjustments), not operational.

Can it be undermined by foreign low-cost labor? Partly mitigated by 2021 AD/CVD orders on imported mattresses (China antidumping up to ~1,732%), but the protection is porous (sourcing migrates to non-covered countries). LEG adds domestic component supply.

Do brands matter? At the premium tier, yes — Tempur-Pedic is the closest thing bedding has to a genuine franchise (real pricing power/recall). Down-ladder (Sealy) competes largely on price. Brands matter but are re-won each ~decade with no switching costs.

Nature of competition; customers’ switching costs? Competition is on price, feel, advertising share-of-voice, and shelf access. Switching costs are essentially zero — no ecosystem, subscription, or lock-in. This is the structural ceiling on any bedding moat.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The Tempur-Pedic/Sealy/Stearns & Foster brand equity is partly internally generated and under-carried; conversely, $4.6B goodwill + $2.6B intangibles from acquisitions are over-represented relative to tangible value.

Off-balance-sheet liabilities? Largely on-balance now — ~$2B of Mattress Firm operating/finance leases are capitalized (part of the $6.5B debt/lease load). Standard product-warranty and purchase obligations.

How conservative is the accounting? Mixed. GAAP is honest (the $1.84 EPS is not flattered), but the adjusted framework strips ~$194M of recurring cash deal/integration costs — aggressive in that it excludes the ongoing cost of being a serial acquirer. No large intangible-amortization add-back (a point in its favor).

How CapEx-hungry is the business? Moderate — capex ~$167M (2025), guided ~$225M (2026, incl. $75M MF store refresh), ~2–3% of sales. Retail adds store-refresh and lease obligations; manufacturing is not capital-intensive.

Capital Allocation & Management

How much FCF, and how is it used? Headline FCF ~$633M (2025); normalized owner FCF ~$450–500M after normalized capex and lease-principal. Uses: dividend (~$127M), debt paydown (deleveraging toward 2–3x), minimal buyback (~$132M in 2025, down from ~$800M in 2021).

Significant acquisitions recently? Yes — the defining feature. Mattress Firm (~$4–5B, closed Feb 2025) and the pending Leggett & Platt (all-stock, ~$1.5B equity + ~$0.9B net debt, signed Apr 2026). A serial acquirer.

Buying back shares / issuing to insiders? Reversed on buybacks — heavy repurchaser pre-2023 (~$1.4B in 2021–22 at ~$30–48), then paused; share count rose 173.6M→206M on MF stock, with ~20M more (~10% dilution) for LEG. SBC modest (~0.5% of sales).

Compensation policy / motivations of management? Incentive comp keyed to adjusted EPS, adjusted EBITDA, Relative TSR, and a soft strategic/ESG component — no ROIC/return-on-capital metric (Interpretation: rewards growth/empire-building, not returns). CEO Scott Thompson “comp actually paid” ~$45.9M in 2025 (stock-inflated). Thompson is the architect of the entire roll-up — concentrated strategic vision.

Valuation & Market Data

ADR, MLP, or K-1 issuer? No — a Delaware C-corp, common stock, NYSE: SGI. Standard 1099 treatment.

Dividend policy? ~$0.62/share (~$127M, ~33% payout), modest and well covered; ~0.8% yield. Growing but not a primary capital-return lever.

How profitable is the business? Operating margin ~10% GAAP / ~13.6% adjusted (down from ~18% in 2021); gross margin ~42.6% (blended manufacturing ~50% / retail ~33%). Profitable but with deteriorating returns on capital.

Is net income diverging from cash from operations? OCF ($800M) exceeds net income ($384M) — a positive cash-conversion sign (D&A, non-cash charges) — but the gap partly reflects the same deal/integration costs added back in adjusted EPS, and headline FCF overstates true owner FCF once lease-principal is counted.

Risks & Downside

What factors would cause the stock to decline? A persisting/deepening mattress recession; a DOJ/FTC challenge or foreign-clearance delay on LEG; integration missteps; margin compression on trade-down; a goodwill impairment; rate/housing weakness; multiple de-rating toward cyclical norms (~11–12x EV/EBITDA).

Risk of a catastrophic loss? Low-to-moderate. The equity is more fragile than the franchise: high leverage + negative tangible equity + trough earnings mean a prolonged downturn compresses EBITDA against a fixed ~$6.5B debt/lease load — the −59% 2022 move is the template. But real brand/retail going-concern value and manageable debt service limit tail risk.

Chance of a total loss? Very low. Cash-generative, brands have durable value, LEG deal is all-stock (no incremental cash leverage). A total loss would require a multi-year structural collapse of mattress demand plus a debt crisis — not the base case.

Recent News & Events

Has the business environment changed recently? Yes — softer-than-expected global bedding demand (down mid-single digits in Q1 2026 vs. flat-to-positive expected), ~$100M oil-derived commodity inflation offset by a ~4% (profit-neutral) list price increase, and the June-3-2026 US HSR clearance of the LEG deal.

Significant acquisitions? Mattress Firm (closed Feb 2025) and pending Leggett & Platt (signed Apr 2026) — see above.

Change in accounting policies? No material change; the visible shift is the consolidation of Mattress Firm (gross-up of retail revenue, elimination of intercompany wholesale) from Feb 5, 2025.

Recent changes — new markets, facilities, management? Rename from Tempur Sealy to Somnigroup (ticker TPX→SGI) in 2025; a $150M Mattress Firm store-refresh program through 2027; Stearns & Foster relaunch (H2 2026). Leadership stable (Thompson CEO/Chair, Rao CFO).

\pagebreak

APPENDIX B — Source Appendix

Somnigroup International, Inc. (NYSE: SGI) — sources as of July 3, 2026. Primary sources prioritized over secondary; third-party aggregated data (ROIC.ai, AZI, FactorsToday) labeled as cross-checks, not primary authority. Facts reconcile to primary filings.

Primary — SEC Filings (CIK 0001206264; via SEC EDGAR (CIK 0001206264))

Primary — Earnings Releases & Transcripts

Primary — Regulatory / Legal

Secondary — Industry Data

Third-Party Quantitative (cross-checks, not primary)

This report is fresh coverage built entirely from the public and third-party sources above.