CoStar Group, Inc. (NASDAQ: CSGP) — A World-Class Data Monopoly Priced Like Its Residential Bet Already Failed

⚡ Claude’s Take

This block is the author’s own independent opinion and general information only — not investment advice. The analytical sections that follow take no position and carry no price target; the single opinion in this article is confined to this block.

Verdict: BUY-the-business / accumulate-on-weakness for patient capital, sub-$32; conviction MEDIUM. CoStar is two companies stapled together: an extraordinary, ~90%-share commercial-real-estate information monopoly (CoStar Suite + LoopNet + STR) plus the #1 US apartment marketplace (Apartments.com) — together a mid-30s%-EBITDA-margin cash machine growing double digits — bolted to a multi-year, multi-billion-dollar residential land-grab (Homes.com) that has, since 2024, incinerated the consolidated margin from 29% to ~6% and, in H1 2026, taken the stock down roughly 70% from its August-2025 high (~$97) to a fresh five-year low (~$28) and its cheapest-ever valuation on both sales (~3.7× vs. a 14–21× history) and book (~1.5×). The market is now paying almost nothing for the residential option and pricing the crown-jewel commercial franchise at a data-services discount. My framing is contrarian value on a falling knife — and I stress knife: the tape is a one-way street (12-month return −63%, six-month Sharpe −1.7), the residential thesis is unproven, and a genuinely new June-2026 threat (Google moving directly into home listings) hangs over the entire portal model. But the downside is unusually well-protected for a stock this beaten: net cash, a commercial segment throwing off ~$640M+ of segment EBITDA that alone can justify most of the enterprise value, an activist-installed Capital Allocation Committee now policing the residential burn, a $1.5B buyback retiring stock at trough prices, and management guiding the residential segment to EBITDA breakeven in Q2 2026. You are buying a great business at a data-industry-trough multiple and getting the residential lottery ticket for free.

Framing & catalysts. This is not a momentum name — it is an abandoned compounder. Reasonable entry zone ~$26–$32 (roughly 3.3–4.0× sales / ~10–12× a normalized ~$1.6–1.8B “whole-company-run-at-commercial-margins” EBITDA). The single most encouraging tell already on the tape: founder-CEO Andrew Florance has been buying stock in the open market with his own cash all the way down — ~141,900 shares for ~$6.1M at descending prices of $74.67 (Oct 2024), $44.52 (Feb 2026) and ~$35 (May 2026), joined by the Marketplaces president and two directors, with no discretionary insider selling — a real, primary-source conviction signal that coincided almost exactly with activist Third Point selling out at the lows. The single fact that flips me decisively bullish: Homes.com sustaining >$106M ARR growth and the residential segment printing durable positive EBITDA through 2026–27 while the commercial core keeps compounding — proof the burn was an investment, not a subsidy. The single fact that flips me bearish: evidence that Google’s listings entry (or an AI-search shift) is structurally cannibalizing Homes.com/Apartments.com SEO traffic, because CoStar’s residential monetization is umbilically tied to organic search. Not a table-pounder given the governance and residential-ROI overhangs — but the risk/reward on a ~90%-share data monopoly at 1.5× book, with the founder averaging down, is asymmetric to the upside for anyone who can stomach the volatility.

📈 Stock Price Action — Five-Year Event Map

CoStar round-tripped a full cycle and then some: from a five-year high of ~$99.74 (Oct 2021) through four years of range-bound “dead money” ($65–$97) while investors waited on the Homes.com payoff, to a fresh five-year low of $28.32 (30 Jun 2026). It trades at ~$30 today (2 Jul 2026), roughly −70% from its 2021 high and −69% from its August-2025 52-week high of $96.83, with a 52-week range of $28.32–$96.83. Almost the entire decline is a 2026 event: the stock entered the year at $65.69 and more than halved in six months. (Prices split-adjusted; CoStar executed a 10-for-1 split in June 2024.)

The price moves are facts from the five-year price series; the attributed causes are interpretation, cross-referenced to earnings dates, 8-K events, guidance and the news feed. No price target or recommendation is implied here — see Claude’s Take above for the opportunity judgment.

1. Executive Summary

CoStar Group is the dominant information utility of the commercial real estate (CRE) industry and, increasingly, a residential-marketplace challenger. Its flagship CoStar Suite — a subscription database of ~8.5 million properties, ~22 million lease/sale comps and millions of tenant records, maintained by a proprietary research field force — commands an estimated ~90%+ share of US CRE information and renews at ~92% quarterly (95%+ for 5-year customers). Around that core sit LoopNet (the leading CRE marketplace), Apartments.com (the #1 US multifamily rental network, ~99% monthly renewal), STR (hospitality benchmarking), and a fast-growing international footprint (Domain in Australia, OnTheMarket in the UK, launches underway in France/Germany/Canada). This is a genuine wide-moat franchise: 79% gross margins, subscription revenue, data network effects, and pricing power evidenced by decades of price-plus-volume growth — CoStar has now delivered 60 consecutive quarters of double-digit revenue growth.

The controversy is entirely about capital allocation. Since launching Homes.com as a residential portal assault on Zillow in February 2024, CoStar has spent aggressively — a planned $550M net investment in 2026 alone — to buy its way into US residential search. That spend, layered on top of a $2.35B 2025 M&A campaign (Matterport, Domain, Ag-Analytics) and campus construction, has collapsed consolidated profitability: EBITDA margin fell from 29.4% (2021) to 5.9% (2025), FY2025 GAAP operating income was −$72M, and reported free cash flow shrank to ~$41M. The market lost patience: the stock fell ~70% from its 2021 high and more than halved in H1 2026, to its cheapest-ever multiples on sales (~3.7×) and book (~1.5×).

The bull case is a sum-of-the-parts: the commercial segment earns a ~34% EBITDA margin (~$161M in Q1’26 alone) and, valued as the data monopoly it is, plausibly supports much of the current enterprise value on its own — leaving the residential business, and its ~$1.7B annualized revenue, thrown in cheaply. Management guides the residential segment to EBITDA breakeven in Q2 2026 and full-company adjusted EBITDA of $780–820M for 2026, and an activist-installed Capital Allocation Committee (April 2025 board settlement) now polices the burn. The bear case is that Homes.com never earns its cost of capital against an entrenched Zillow (~230M monthly uniques vs. CoStar’s ~131M-uniques network) — the residential segment lost an estimated ~$361M in 2024 and ~$230M in 2025 (Third Point figures) on ~$100M of Homes.com revenue; that the ~$5B in goodwill/intangibles from serial M&A masks value destruction; that “adjusted EBITDA” flatters economics burdened by ~$194M of stock-based compensation; and — the newest risk — that Google’s June-2026 move into home listings structurally undermines the SEO-driven traffic on which both Homes.com and Apartments.com depend. Activist Third Point (Daniel Loeb) ran a public campaign against the residential spend and Founder-CEO Andrew Florance’s pay through 2025, then sold its entire stake and dropped the fight in April 2026. This memo takes no investment position; it lays out the mechanism, the numbers, and the falsification tests for each side.

2. Business Overview

CoStar Group operates two reportable segments — Commercial Real Estate and Residential Real Estate — but is best understood as four economic engines of very different quality.

(1) CoStar Suite (information & analytics) — the crown jewel. Subscription access to the most comprehensive CRE database in the world: property inventory (office, industrial, retail, multifamily, hospitality, land), comparable sales, lease comps, tenant data, market analytics, and increasingly debt/lender analytics. Q1’26 CoStar revenue was $331M (+9% YoY), with users up 22% YoY to 317,000, a 92% quarterly renewal rate and an NPS of 69. The data is built and continuously refreshed by a large proprietary research operation — thousands of researchers making calls, driving markets, and canvassing listings — which is the source of the moat. Adjacent modules are compounding on top: CoStar Debt Solutions crossed $100M revenue serving 500+ lenders; new products (rent benchmark, new-homes tracking, CRE loan origination workflow) extend the wallet.

(2) Marketplaces — LoopNet, Apartments.com, Land.com, Ten-X, BizBuySell. LoopNet ($85M Q1’26, +16%) is the leading global CRE listings marketplace, now rolling out asset-based advertising pricing. Apartments.com ($312M Q1’26, +10%, 15th consecutive double-digit-growth quarter, ~99% monthly renewal) is the #1 US multifamily rental network and a second cash-cow franchise — a genuine two-sided marketplace with brand dominance (its long-running national ad campaign) and best-in-class SEO/SEM. Land.com, Ten-X (CRE auctions), and BizBuySell (business-for-sale marketplace) round out the marketplace stack.

(3) Residential — Homes.com + international portals. The contested growth bet. Homes.com ($26M Q1’26, +58%) is CoStar’s residential-for-sale portal, using a differentiated “your listing, your lead” agent-membership model (the listing agent, not a buyer’s-agent advertiser, gets the lead) versus Zillow’s Premier Agent auction. It had 35,175 agent subscribers at Q1’26 (+4,300 in the quarter, +205% YoY), ~$3,400 average annual subscription, and a March annualized run-rate of $106M (+92%). Internationally, Domain (Australia, taken over in 2025) and OnTheMarket (UK, #2 by inventory) extend the residential model, with CoStar-branded launches underway in France, Germany, Australia and Canada.

(4) Matterport & data infrastructure. Acquired February 2025 for ~$1.6B, Matterport supplies 3D “digital twin” tours now embedded across Apartments.com (~250,000 tours; listings with a Matterport generate 56× more tour requests), Homes.com, LoopNet and Domain — a differentiation and proprietary-data play.

Business model. ~73% of revenue is annual-contract subscription (Domain is transactional), which produces high revenue visibility, deferred-revenue float and negative working capital at the franchise level. Revenue is overwhelmingly US (with a growing international minority). The economic character is a classic information/marketplace network: very high gross margin (~79%), high incremental margin on the mature franchises, and — critically — a management team currently choosing to reinvest nearly all of the core’s profit into the residential build.

Verdict: A rare, genuinely dominant information franchise (CoStar Suite + Apartments.com) yoked to an unproven, cash-hungry residential challenger. The quality of the business and the quality of the current capital allocation are two very different questions.

3. Industry Dynamics

CoStar straddles three industries of very different structural quality.

CRE information & analytics — an excellent industry, and CoStar owns it. The commercial-real-estate data market is a natural quasi-monopoly: the value of a CRE database rises with its completeness, completeness requires an enormous fixed-cost research operation, and once a broker/appraiser/lender/investor community standardizes on one dataset, switching means abandoning the shared language of comps the whole market quotes. CoStar has spent 35+ years and billions building that dataset; competitors (Moody’s/REIS, Altus, MSCI/Real Capital Analytics, Yardi Matrix, VTS, and venture-funded challengers like Crexi and Reonomy) hold niches but none approaches CoStar’s breadth. The result is durable pricing power (annual price increases plus seat/module expansion), ~90%+ share, and mid-90s% long-tenured renewal. In Greenwald’s taxonomy this is a textbook economies-of-scale-plus-customer-captivity advantage — the strongest kind. Marathon’s capital-cycle lens is favorable too: high returns here have not attracted enough well-capitalized supply to erode CoStar’s position, because the fixed-cost research moat deters entry.

Multifamily rental marketplaces — a good industry with a strong #1. Apartments.com is the category-defining brand with two-sided network effects (renters go where the listings are; property managers advertise where the renters are) and ~99% renewal. Competition (Zillow Rentals, Rent./Redfin, Zumper, Costar’s own Homes.com syndication) is real but Apartments.com’s brand + SEO + sales force keep it dominant. This is the second jewel and, notably, sits inside the “Residential” reporting segment — meaning the segment’s optics conflate a superb franchise (Apartments) with a speculative one (Homes).

Residential for-sale portals — a structurally worse, winner-take-most arms race. US residential search has an entrenched leader (Zillow, with its Premier Agent monetization, Showcase, ShowingTime and rental network) and a duopoly-ish structure with Realtor.com (News Corp/Move) and Redfin. Monetization depends on consumer traffic, which depends on brand and — heavily — organic search (SEO) and paid search (SEM). This is the arena CoStar chose to attack with Homes.com, spending ~$1B+ to buy brand awareness (Super Bowl ads, Oscars/Olympics) and agent memberships. Two structural risks define the industry: (a) it is a marketing-spend arms race where the incumbent’s installed base and habit are hard to dislodge; and (b) it is existentially exposed to changes in search distribution — the June-2026 news that Google is expanding directly into home listings threatens the SEO/SEM traffic funnel that every portal (Zillow and Homes.com alike) relies on, and the broader shift toward AI-answer search could disintermediate portals entirely. The NAR commission settlement (Sitzer/Burnett) has also reshaped buyer-agent economics, with ambiguous portal implications.

Verdict: Structurally excellent in CRE data (wide-moat, low capital-cycle threat), good in multifamily marketplaces, and structurally challenged and contested in residential-for-sale portals — the one arena where CoStar is spending the most and the market is most skeptical.

4. Competitive Position

The moat is real and it is in the data, not the software. CoStar Suite’s advantage is a self-reinforcing loop: the largest proprietary research operation in CRE produces the most complete dataset → the most complete dataset attracts the most subscribers → subscription revenue funds the largest research operation. A new entrant cannot replicate the field-collected comps, tenant intelligence and historical depth without spending a decade and hundreds of millions with no revenue — and even then would face a market that already speaks in CoStar’s numbers. The financial fingerprint of this moat is unambiguous: 79% gross margins, ~92% quarterly renewal (95%+ for tenured accounts), 22% user growth, and decades of price-plus-volume compounding. If the moat weren’t real, none of those would survive competition. This is a supply-advantage (unmatchable cost-per-datapoint at scale) fused with demand-side captivity (switching costs, workflow embedding). Apartments.com adds a second, different moat: two-sided marketplace network effects plus consumer brand.

Where the moat is thin or unproven — residential. Homes.com is trying to manufacture a network effect in a market where Zillow already has one. CoStar’s counter-strategy is clever: a differentiated “your listing, your lead” agent model that appeals to listing agents alienated by Zillow’s practice of selling their listings’ leads to competitors, plus enormous brand spend and SEO investment. Early evidence is genuinely encouraging — 205% YoY subscriber growth, +92% ARR, claimed 11× agent ROI, rising engagement (AI users spend ~4× longer on site), and comScore data showing Homes.com as the fastest-growing residential/rental site while Zillow’s unique visitors have declined year-over-year for 15 consecutive months. But it is early, the monetization base ($106M ARR) is a rounding error against the ~$550M annual spend, and the entire model rests on organic search traffic that Google could disrupt. This is a moat under construction, not a moat in hand.

Head-to-head. vs. Zillow: Zillow remains the traffic and profit leader (~230M monthly uniques vs. CoStar’s ~131M-unique network; ~$1.4B residential revenue vs. Homes.com’s ~$100M), and Homes.com would need on the order of ~250,000 paying agents at current ARPA to match Zillow’s residential revenue — against ~35,000 today. CoStar is out-growing Zillow on traffic momentum (Zillow uniques have fallen YoY for 15 straight months) and has a structurally cleaner agent proposition, aided by the NAR/Sitzer-Burnett settlement (effective Aug 2024), which bans seller-set buyer-agent commissions and undercuts the buyer-lead economics Zillow’s Premier Agent depends on (on settlement day CoStar rose ~8% while Zillow fell ~13%). But transactions are finite and zero-sum, so traffic gains have not yet translated into rival revenue capture. vs. Moody’s/MSCI/Altus in CRE data: CoStar is the clear leader by breadth and share (~1,500-researcher field force, ~38-year/$5B+ data investment). vs. Rightmove/REA internationally: CoStar is the challenger (OnTheMarket a distant #3 in the UK behind Rightmove/Zoopla; Domain #2 in Australia behind REA — a more rational duopoly).

Verdict: A durable, wide moat in CRE information and multifamily marketplaces — among the best franchises in information services — attached to a contested, still-unproven residential position. The consolidated financials understate the core’s quality and overstate the whole’s risk only if the residential burn is truly optional; if it is structurally necessary to defend Apartments.com’s SEO ecosystem, the two are more entangled than the sum-of-parts bulls assume.

5. Growth History and Forward Opportunities

History. CoStar’s growth record is exceptional in its consistency: 60 consecutive quarters of double-digit revenue growth, from $1.66B (2020) to $3.25B (2025) — an ~18% CAGR, a blend of ~7–10% organic (price + seats + new modules) and acquired growth (STR, Ten-X, Homesnap, OnTheMarket, Visual Lease, Matterport, Domain). The organic engine is the CoStar Suite’s price-plus-volume machine and Apartments.com’s 15-straight double-digit quarters. Q1’26: total revenue +23% (10% organic), Commercial +15% (7% organic), Residential +32% (13% organic).

Forward opportunities, ranked by quality:

• CoStar Suite module expansion (high quality): Debt Solutions (>$100M, +26% bookings), rent benchmark, new-homes data, lender workflow — extending ARPU within a captive base. This is the best growth the company has.
• International CRE (high quality): launching the CoStar/LoopNet playbook in the UK (bookings +44%), Canada (+22%), France (Q2’26, cross-selling 32,000 Business Immo subscribers), Germany and Australia (Q3–Q4’26) — replicating a proven monopoly model in greenfield markets.
• Apartments.com (good): durable double-digit growth, price transparency leadership, AI search (Smart Search, Apartments AI).
• Homes.com (high-risk, high-reward): the swing factor — if agent memberships and price increases (new-customer prices raised 1 May 2026) compound toward a multi-hundred-million-dollar profitable business, it re-rates the whole company; if it stalls, it is a value sink.
• Matterport / AI (optionality): 3D + AI (“Homes AI,” “Apartments AI”) as engagement and proprietary-data flywheels.

Verdict: High-quality, durable growth in the commercial and international data businesses; genuinely double-digit organic growth even ex-residential. The quality of the residential growth is the open question — it is real revenue growth bought at a steep and still-unrecouped cost.

6. Financial Quality

This is where the controversy lives. CoStar’s underlying franchise economics are superb; its reported consolidated economics have been deliberately buried under residential investment.

Revenue & margins. Revenue compounded from $1.66B (2020) to $3.25B (2025). But EBITDA margin collapsed as the residential build accelerated:

The gross margin barely moved (~79–82%) — the collapse is 100% operating-expense-driven: SG&A more than doubled ($879M→$2,109M) and R&D doubled ($201M→$406M), the direct cost of the Homes.com brand/sales assault plus M&A integration. The core did not deteriorate; management chose to spend its profits. Proof: Q1’26 Commercial adjusted EBITDA margin was 34%, and management guides full-company adjusted EBITDA back to $780–820M (~21% margin) in 2026 — the earnings power is intact and re-emerging as residential losses narrow.

Quality-of-earnings flags (be skeptical of “adjusted”):

• GAAP vs. adjusted gap is large. Management leads with “adjusted EBITDA doubled to $132M” in Q1’26, but FY2025 GAAP operating income was negative and net income was $7M. Adjusted EBITDA adds back ~$194M of stock-based compensation (more than 2× the $89M of 2024, swollen by Matterport retention), plus M&A/integration and the residential “investment.” SBC is a real economic cost and dilutes shareholders; treat adjusted EBITDA as an aspirational, not a cash, figure.
• Real FCF is currently minimal. Cash-flow-statement free cash flow was ~$41M in 2025 and −$245M in 2024 — because capex ran $389M–$638M (new Arlington HQ + Richmond campus) on top of the opex burn. (ROIC’s “$819M FCFF” is an unlevered add-back construct, not distributable cash.) FCF will inflect sharply upward as campus capex rolls off and residential turns, but today the cash generation is thin.
• Interest income has been a real earnings prop. With $4–5B of cash in 2023–24, interest income was ~$210M/year — a meaningful chunk of pre-tax income that is now shrinking as the cash was spent on Matterport/Domain (cash fell to $1.63B).
• Effective tax rate distortions. With pre-tax income compressed to ~$30M in 2025, the tax rate optically spiked (76%), making GAAP EPS ($0.017) meaningless — hence the 497× “P/E” is noise, not signal.

Balance sheet — a fortress, still. Even after spending ~$3.5B on M&A, CoStar ended 2025 with $1.63B cash vs. $1.18B debt = ~$640M net cash, a 2.8× current ratio, and modest leverage. Equity is $8.37B, but ~$6.7B of assets are goodwill/intangibles (Matterport $1.1B goodwill, Domain $1.2B), so tangible book is thin (~$1.6B; TBVPS ~$4). Net cash + a highly cash-generative core means liquidity/solvency risk is negligible — this is not a balance-sheet story, it is a returns-on-investment story.

Returns on capital. Ignore the vendor ratio feeds here — third-party ROE/ROIC fields are garbled for CSGP (e.g., a nonsensical 610% “ROE” for 2024). Computed manually, consolidated ROE is near zero (net income $7M on $8.4B equity) and consolidated ROIC is negative — but this reflects the residential drag, not the core. The unlevered CoStar Suite / Apartments franchises earn very high returns on their (small) tangible capital; the negative consolidated number is the cost of the residential option, capitalized through the P&L.

Verdict: Economics unambiguously improve with scale in the core (79% gross margin, 34% commercial EBITDA margin, negative working capital). The consolidated financials are temporarily — and by choice — poor. The honest question is not “are the economics good” (they are) but “will the residential capital deployed ever earn an adequate return” (unproven).

7. Capital Allocation

Capital allocation is the entire debate, and CoStar’s record is genuinely two-sided.

The bull reading: Andrew Florance built a company once worth >$40B from scratch since 1987, compounding revenue for 60 straight double-digit quarters, and has repeatedly made contrarian, ahead-of-consensus bets (LoopNet, Apartments.com, STR) that looked expensive and proved brilliant. Homes.com is the same playbook — attack an incumbent-dominated adjacency with data + brand — and early KPIs (subscriber and ARR growth, engagement, claimed 11× agent ROI) are tracking. The M&A has strategic logic: Matterport ($1.6B, Feb 2025, proprietary 3D data across all portals), Domain (~US$1.92B/A$3B, Aug 2025, the #2 Australian portal to run the residential playbook in a rational duopoly), Visual Lease (Oct 2024, lease-accounting workflow into the CoStar Suite).

The bear reading: CoStar has spent an enormous sum — a planned $550M net in 2026 alone, on top of an estimated ~$361M (2024) and ~$230M (2025) of residential losses and cumulative Homes.com spend north of $3B — with, so far, ~$106M of Homes.com ARR to show for it. The serial M&A has parked ~$5B of goodwill/intangibles on the balance sheet whose returns are unproven; Domain (~A$3B) and Matterport (~$1.6B, half in stock, chronically unprofitable) were bought at full prices near a market peak — and, notably, no goodwill impairment has been taken despite the residential losses, an accounting-quality item worth watching. Stock-based compensation doubled to ~$194M, diluting holders even as the company began buying back stock. And the residential bet was pursued over shareholder objections — which is why two activists, D.E. Shaw and Daniel Loeb’s Third Point, ran overlapping campaigns in 2025 attacking the “misallocation of billions into Homes.com” (Third Point publicly called the board “feckless” and flagged Florance’s ~$37M pay; D.E. Shaw published a detailed “Perspectives on CoStar” analysis alleging up to ~$11B of value destruction). CoStar settled in April 2025: new independent directors, board cut to eight (seven independent), the co-founder Chairman Michael Klein replaced by independent chair Louise Sams, and a Capital Allocation Committee to review Homes.com profitability — though, tellingly, that committee is chaired by Florance himself, blunting its independence. The company then cut ~200 staff and trimmed 2026 residential spend to ~$550M. Third Point ultimately sold its entire stake and dropped its proxy fight in April 2026 (“thesis no longer holds”) — a capitulation that removed the “distraction” management complained of but also removed the loudest external check. That outside investors had to force capital-allocation discipline in the first place — and that the founder chairs the very committee meant to police his own spend — is the central governance red flag.

The core roll-up has been genuinely value-accretive. The CRE/data acquisitions built the moat at sensible prices: LoopNet (2012), Apartments.com ($585M, 2014, ~7× EBITDA — a home run), STR ($450M, 2019), Ten-X ($190M, 2020). Management also showed real discipline walking away — abandoning the RentPath deal under FTC objection (2020) and dropping a ~$7.35B bid for CoreLogic (2021) rather than overpay. This is the track record that earns Florance the benefit of the doubt. The residential/recent deals are where the discipline frays: OnTheMarket (2023), Visual Lease (Oct 2024), Matterport and Domain at full prices.

Capital returns — new, and defensive. CoStar historically paid no dividend and did no buybacks — every dollar went into M&A and Homes.com. Only under activist pressure did it initiate its first-ever buyback in 2025 ($500M, 7.1M shares), authorize a $1.5B program (January 2026), and repurchase $505M (11.4M shares, largely via ASR) in Q1’26 at trough prices — an intelligent use of the net-cash balance sheet at the cheapest-ever valuation, but a concession extracted rather than an owner’s philosophy.

Incentive alignment — a live governance sore spot. The 2025 say-on-pay vote collapsed to ~53% support (from ~92% the prior year) — a severe shareholder rebuke — and CoStar filed a defensive solicitation to get it passed. Florance’s total compensation was $36.4M (2025), $37.4M (2024), $29.2M (2023) — large and rising while the stock round-tripped, against a founder stake of only ~0.75% (single share class, no super-voting; all insiders 1.18%). Short-term incentive is 100% EBITDA-based (paid at the 200% cap); long-term is now ~80% PSUs. In response the board overhauled 2026 pay — eliminating options, moving to 80% PSUs, raising and capping the relative-TSR target, and removing the CEO’s tax gross-up. Pay is therefore defensible on paper (tied to EBITDA and relative TSR), but the ~$36M quantum for a sub-1%-owner founder, plus the 53% vote, keeps compensation a genuine risk.

The insider tell cuts the other way. Against all of the above sits the most concrete alignment signal available: Florance bought ~141,900 shares (~$6.1M) in the open market with his own cash, averaging down from $74.67 (Oct 2024) to $44.52 (Feb 2026) to ~$35 (May 2026) — accelerating as the stock fell and as Third Point sold out. He was joined by the Marketplaces president (~$907K at $45.33) and two independent directors. There was no discretionary insider selling during the crash (only routine tax-withholding on vesting). Whatever one thinks of the pay and the empire-building, the founder is putting personal capital behind the stock near its lows.

Verdict: Two capital allocators in one — a disciplined, value-creating CRE/data roll-up and an aggressive, unproven, empire-building residential investor whose spending required outside investors to demand oversight. The buyback-at-trough, the activist-installed cost discipline, and the founder’s open-market buying are genuine positives; the unrecouped residential spend, full-price residential M&A, un-impaired goodwill, and ~$36M pay against a 53% say-on-pay vote are genuine negatives. The verdict hinges on whether Homes.com converts — the same open question as everywhere else in this memo.

8. Changes and Headwinds — Last Two Years

• Homes.com national launch (Feb 2024) and escalating spend — the defining strategic change; the source of both the growth narrative and the margin collapse.
• Matterport acquisition (~$1.6B, closed Feb 2025) — 3D digital-twin platform; ~$1.1B goodwill; doubled SBC via retention awards.
• Domain Australia takeover (2025, ~A$2.8B+) — a controlling/complete acquisition of the #2 Australian residential portal; ~$1.2B goodwill; transactional (non-subscription) revenue model.
• Visual Lease (Oct 2024) and Ag-Analytics (Feb 2025) — bolt-ons into CoStar Real Estate Manager and Land.com.
• Twin activist campaigns (D.E. Shaw + Third Point/Loeb, 2024–2026) and the April 2025 board settlement — new independent directors, board cut to eight, co-founder Chairman Klein replaced by independent chair Sams, and a Florance-chaired Capital Allocation Committee. Third Point sold its entire stake and dropped its proxy fight in April 2026 — the “activist distraction” management says is “behind us.”
• Founder open-market buying — Florance bought ~$6.1M of stock averaging down to ~$35 (Oct 2024–May 2026), with a director/officer buying cluster and no discretionary selling.
• First-ever buybacks — $500M (2025) then a $1.5B authorization (Jan 2026); ~$700M planned 2026 outlay.
• 2025 say-on-pay collapse to ~53% and a resulting 2026 compensation overhaul (options eliminated, tax gross-up removed).
• Google’s entry into home listings (news ~June 2026) — a new, potentially structural threat to portal SEO traffic; a proximate driver of the H1’26 selloff.
• The stock’s ~70% decline and cheapest-ever valuation — itself a “change” that shifts the risk/reward and enables value-accretive buybacks.
• International CRE launches (France Q2’26, Germany/Australia H2’26) — the next leg of high-quality growth.

Verdict: The last two years mixed genuine strategic progress (international, Debt Solutions, Apartments strength, Matterport data) with a controversial, margin-destroying residential bet, an activist intervention, and a new competitive threat. Net effect on the thesis: the business is broader and the balance sheet lighter on cash; the stock is dramatically cheaper; and the make-or-break residential question is closer to resolution (breakeven guided for Q2’26) but not yet answered.

9. Risk Analysis

Overall: The dominant risks are strategic (residential ROI, Google/AI distribution) rather than financial (balance sheet, solvency). Downside is unusually well-protected by the net-cash balance sheet and the high-margin commercial core; the debate is about upside realization, not survival.

10. Valuation Discussion (Embedded Expectations)

CoStar trades at its cheapest valuation as a public company on the two multiples that aren’t distorted by the residential P&L: ~3.7× TTM sales (vs. an 8.6× at YE2025, 10.6× in 2024, and a 14–21× range across 2019–2023) and ~1.5× book (vs. 12–30× historically). Measured against its own multi-year history, both the P/S and P/B multiples sit at all-time lows. The P/E (~497×) is meaningless because GAAP EPS is compressed to ~zero by the residential burn; ignore it (a classic distorted-EPS gotcha).

What is the market pricing? At ~$12B enterprise value (net-cash-adjusted ~$11.4B) against ~$3.25B sales and a guided $780–820M of 2026 adjusted EBITDA, the market is paying ~14–15× adjusted EBITDA — but that adjusted figure is flattered by SBC. The more telling frame is the sum-of-the-parts:

• Commercial segment: ~$1.9B revenue at a ~34% adjusted EBITDA margin ≈ $640M+ segment EBITDA, growing high-single/low-double digits, with data-monopoly economics. Peer information-services franchises (S&P Global, Moody’s, Verisk, Gartner, Equifax) trade at ~15–25× EBITDA / ~6–12× sales even after their own de-ratings. Put a conservative 15–18× on ~$640M+ and the commercial segment alone is worth ~$10–12B — i.e., roughly the entire current enterprise value.
• Residential segment: ~$1.7B revenue (Apartments.com — a #1 franchise ~99% renewal — plus Homes.com plus Domain/OnTheMarket), guided to EBITDA breakeven in Q2’26. Apartments.com alone, as a dominant marketplace, is plausibly worth several billion dollars. In the SOTP, the market is assigning this segment ~zero-to-negative value.

Embedded expectations, restated: at ~$30, the market is underwriting that (a) the commercial core is worth roughly what a de-rated data peer would fetch, and (b) the entire residential segment — including the #1 US apartment marketplace — is worth approximately nothing net of the Homes.com burn. That is a coherent bear view only if Homes.com’s losses are permanent and structurally required to defend Apartments.com’s ecosystem. If residential merely reaches breakeven (Q2’26 guide) and Apartments.com is valued as the franchise it is, the SOTP implies material upside.

Scenario sketch (illustrative, not a target):

• Bear: Homes.com burn persists/deepens, Google erodes portal traffic, organic growth slows to mid-single-digits. Whole company worth ~2.5–3× sales → high-$20s/low-$30s; the stock is roughly fairly valued for this outcome, which is why it fell here.
• Base: Residential reaches sustained breakeven-to-modest-profit, commercial compounds ~8–10%, adjusted EBITDA marches to $780–820M then higher. Re-rate toward ~5–6× sales / ~mid-teens normalized EBITDA → ~$45–60.
• Bull: Homes.com converts to a profitable multi-hundred-million-dollar franchise, international scales, margins normalize toward the mid-30s% the core demonstrates. ~7–9× sales → $70–90+ (a return toward prior highs on higher revenue).

No price target and no recommendation — this section frames embedded expectations only. The asymmetry the SOTP reveals is the reason the name is interesting; the residential ROI question is the reason it is not a certainty.

11. Variant Perception

Consensus view: “A great company that made a bad, ego-driven capital-allocation decision. Homes.com is a money pit against an unbeatable Zillow, the residential dream is dying, Google just made portals structurally worse, and the stock deserves to be down 70%.” Sell-side is split and cautious (Wells Fargo Underweight, $26; Goldman Buy but PT cut to $46) — a stock the market has largely given up on, priced for the residential bet to fail.

Strongest bull case: The market is conflating a temporary, self-inflicted, and optional margin depression with permanent value destruction. Strip out the residential investment and CoStar is a ~$1.9B-revenue, 34%-margin, wide-moat data monopoly compounding double digits — worth roughly the entire current EV on its own — with Apartments.com (a #1 marketplace) and a $1.7B residential business thrown in for free. Management is turning the residential segment to breakeven (Q2’26), buying back stock at trough prices, and now operates under activist-installed capital discipline. You are paying a data-industry-trough multiple for the best franchise in CRE information and getting a free option on the residential build.

Strongest bear case: Homes.com is a structural loser — Zillow’s consumer habit and agent network are unassailable, the ~$550M annual spend is a permanent subsidy dressed up as “investment,” and Google’s entry plus AI-search disruption make the entire portal model a melting ice cube. “Adjusted EBITDA” is a fiction that ignores $194M of real SBC dilution; GAAP earnings and FCF are near zero; the balance sheet’s cash cushion was already spent on full-price M&A parking $5B of dubious goodwill. A founder-CEO who had to be reined in by an activist will find the next expensive adjacency. The stock is cheap for a reason.

The 3–5 assumptions that decide it:

1. Does Homes.com reach durable, meaningful profitability? (Watch: residential segment EBITDA turning and staying positive from Q2’26; Homes.com ARR trajectory beyond $106M; retention after the May-2026 price increase.)
2. Does Google / AI search cannibalize portal traffic? (Watch: Homes.com and Apartments.com organic-traffic trends; Google’s rollout scope; any SEO disruption.)
3. Does the commercial core keep compounding at ~34% margins? (Watch: CoStar Suite bookings/renewals; international CRE ramp; a CRE-cycle downturn.)
4. Is capital allocation now disciplined? (Watch: Capital Allocation Committee actions; buyback pace; absence of another mega-deal.)
5. Is the SOTP real or an accounting mirage? (Watch: whether segment-level economics, once residential breaks even, validate the parts-worth-more-than-whole math.)

Factor-positioning read. The tape is unambiguously a one-way street: 12-month return −63%, six-month annualized Sharpe −1.7, relative strength −70% vs. peak, and a 5-year annualized return of −19%. Factor-model data shows CSGP now loads like a beaten-down growth/quality name (factor-similar peers: SS&C, Autodesk, GoDaddy, Genpact — de-rated software/services, not real-estate cyclicals), with market beta ~0.8–1.0. This is a falling knife with a value valuation — a stock whose momentum is horrendous precisely because its multiple has collapsed to a floor. For a contrarian, that is the setup; for a trend-follower, it is a name to avoid until the tape stabilizes. The factor evidence supports “consensus is maximally bearish,” not “consensus is wrong” — those are different claims, and only the fundamentals (assumptions 1–3 above) resolve which is right.

12. Fact vs. Interpretation

13. Open Questions

1. What is the segment-level EBITDA of Commercial vs. Residential in detail, and what does Apartments.com earn on a standalone basis (the key SOTP input)?
2. What are Homes.com unit economics — CAC, agent churn/retention after the May-2026 price increase, and the LTV that justifies the spend?
3. How exactly does Google’s home-listings product work, and what share of Homes.com/Apartments.com traffic is exposed to it (vs. direct/brand traffic)?
4. What is the path and ceiling for the residential margin — breakeven in Q2’26, but then what steady-state margin, and by when?
5. What is the compensation structure and any say-on-pay history, and what precisely is the Capital Allocation Committee empowered to do?
6. (Answered) Yes — Florance bought ~141,900 shares (~$6.1M) in the open market, averaging down from $74.67 to ~$35 (Oct 2024–May 2026), with a director/officer buying cluster and no discretionary selling — a genuine conviction signal. The open question is whether his read on Homes.com’s payoff is right.
7. How cyclical is the CoStar Suite in a genuine CRE downturn (office in particular)?
8. Will the Capital Allocation Committee remain a real check now that Third Point has exited and it is chaired by Florance himself?

14. What Must Be True

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

• The commercial core keeps compounding ~8–10% at mid-30s% margins and the residential segment converts from a subsidy into a self-funding, growing profit stream.
• Falsification: Residential segment EBITDA fails to turn/hold positive through 2026–27, or Homes.com ARR growth decelerates sharply after the price increase, or commercial organic growth drops below ~6% — any of which would prove the SOTP is a mirage and the burn is permanent.

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

• Homes.com is a structural loser and Google/AI search is dismantling the portal model, such that the residential segment is worth ≤0 and drags the whole.
• Falsification: Residential prints durable positive EBITDA while Homes.com ARR compounds >50% and Apartments.com sustains ~99% renewal and double-digit growth despite Google’s entry — which would prove the residential business is real, defensible, and mispriced at zero.

The single most important number to watch: the Residential segment’s quarterly EBITDA from Q2’26 forward. Positive and rising validates the bull SOTP; slipping back into loss validates the bear.

15. Source Appendix

See CSGP_source_appendix.md (Appendix B in the combined report) for the full source list.

APPENDIX A — Standard Diligence Questionnaire — CoStar Group, Inc. (NASDAQ: CSGP)

As-of 2026-07-03. Fact / Interpretation / Assumption labeled where it matters.

General

What thoughtful questions have other investors asked? The debate is dominated by capital allocation, and two activists (D.E. Shaw and Daniel Loeb’s Third Point) crystallized the questions in 2025–26: (1) Will Homes.com ever earn its cost of capital, or is the ~$550M/yr net spend a permanent subsidy against an entrenched Zillow? (2) What is the sum-of-the-parts worth if the commercial data monopoly is valued as the ~34%-margin franchise it is and the residential burn is stripped out? (3) Is “adjusted EBITDA” a fair proxy for economics given ~$194M of SBC and near-zero GAAP earnings/FCF? (4) Post-Third-Point-exit, is there any remaining check on a founder-CEO who chairs his own Capital Allocation Committee? (5) The newest question (June 2026): does Google’s entry into home listings structurally break the portal traffic model? Interpretation: these are the right questions; the memo’s SOTP frames the mispricing, and the residential-ROI and Google questions are the genuine unknowns.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? A self-inflicted low. Consolidated EBITDA margin is ~6% (2025) vs. a ~29% mid-cycle level (2021), and GAAP operating income is negative — but this reflects a deliberate residential investment, not a cyclical trough. Normalized/“run-at-commercial-margins” earnings power is far higher (Commercial adj EBITDA margin ~34%; FY26 guided adj EBITDA $780–820M). Interpretation: earnings are depressed by choice, which is the bull’s entire point.

Driven by the external environment or internal actions? Overwhelmingly internal (the Homes.com spend). Secondary external factors: CRE cycle (office softness, transaction volumes) and interest income (falling as cash was spent on M&A).

How stable are revenues? Very — ~73% annual-contract subscription, ~89% blended renewal (95%+ tenured), 60 consecutive quarters of double-digit revenue growth. Domain (transactional) and Ten-X (auction) are the cyclical minority.

Outlook for products/services; how big is the market? CoStar cites a >$100B global real-estate information + marketplaces TAM (aspirational). Durable growth in CRE data, international CRE, and Apartments.com; the residential opportunity is enormous but contested. Growing, international, multi-product.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? CRE data: stable, CoStar-dominated (challengers like Crexi/Reonomy nibble edges). Residential portals: more competitive and now structurally threatened by Google/AI search.

How profitable is the business (ROIC/ROE)? Aggregator ratios are garbled for CSGP — computed manually. Consolidated ROE is ~0% (net income $7M on $8.4B equity) and consolidated ROIC negative — but this is the residential drag. The core CoStar Suite / Apartments.com franchises earn very high returns on modest tangible capital (79% gross margin, ~34% commercial segment EBITDA margin, negative working capital). Assumption: unlevered core ROIC is well into the double digits.

How profitable is the industry — competitors, barriers? CRE data is a high-barrier oligopoly (data scale + field-force + switching costs). Residential portals are a lower-barrier, marketing-driven arms race.

Can the business be easily understood? Yes — subscription data + marketplaces. The complication is the two-speed segment mix.

Undermined by foreign low-cost labor? No — the moat is proprietary US/local data and brand, not labor arbitrage.

Do brands matter? Critically — Apartments.com and Homes.com are brand/SEO plays; CoStar is the industry-standard data brand.

Nature of competition / switching costs? CRE Suite: very high switching costs (workflow embedding, shared comp language). Marketplaces: two-sided network effects. Residential-for-sale: low switching costs, traffic-driven.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The CoStar Suite dataset and the Apartments.com/CoStar brands — internally built, largely unrecognized, and the core of the value. Conversely, ~$6.7B of goodwill/intangibles from M&A is recognized and arguably over-stated (no impairment despite residential losses).

Off-balance-sheet liabilities? Operating leases (new campuses); nothing unusual flagged.

How conservative is the accounting? Mixed. Revenue recognition is clean (subscription). The flags: heavy reliance on “adjusted EBITDA” that adds back ~$194M SBC; no goodwill impairment despite ~$230–361M residential losses (watch item); interest income propping pre-tax income historically.

How CapEx-hungry? Normally light (software/data), but 2024–25 capex spiked to $389–638M on new Arlington HQ + Richmond campus builds — a temporary drag on FCF that should roll off.

Capital Allocation & Management

How much FCF, and how is it used? Reported FCF collapsed to ~$41M (2025) on opex burn + campus capex; normalized FCF power is far higher. Historically 100% reinvested (M&A + Homes.com); since 2025, buybacks added under activist pressure.

Significant acquisitions? Matterport (~$1.6B, Feb 2025), Domain Australia (~A$3B, Aug 2025), Visual Lease (Oct 2024), Ag-Analytics (2025). Earlier value-accretive deals: Apartments.com ($585M, 2014, ~7× EBITDA), STR ($450M, 2019), Ten-X ($190M, 2020). Discipline shown walking from RentPath (2020) and CoreLogic (~$7.35B bid, 2021).

Buying back shares? Yes — first-ever: $500M (2025), $1.5B authorization (Jan 2026), ~$700M planned 2026. Issuing shares to insiders? SBC ~$194M (2025, doubled on Matterport retention); net dilution partly offset by buybacks.

Compensation policy / motivations? Florance total comp ~$36M (2025), largely PSU-based (STI 100% EBITDA); 2025 say-on-pay collapsed to ~53%, prompting a 2026 overhaul (options eliminated, tax gross-up removed). Founder stake ~0.75% (no super-voting). Motivation read: a founder-builder pursuing a legacy residential win; the ~$6.1M of open-market stock purchases averaging down to ~$35 signal genuine conviction and alignment.

Valuation & Market Data

ADR / MLP / K-1? No — US common stock, single class, NASDAQ.

Dividend policy? None.

How profitable / is net income diverging from cash flow? GAAP net income ($7M) and reported FCF ($41M) are both depressed and roughly aligned (both near zero); “adjusted EBITDA” ($780–820M guided) diverges sharply upward — the gap is SBC + residential investment + M&A add-backs. Watch the GAAP-to-adjusted bridge.

Risks & Downside

What would cause the stock to decline (further)? Homes.com losses persisting/deepening; Google/AI cannibalizing portal traffic; a CRE downturn hitting the Suite; another large full-price acquisition; commercial organic growth decelerating below ~6%.

Catastrophic / total-loss risk? Very low. Net cash (~$640M), a cash-generative wide-moat core, no refinancing cliff. This is a returns-on-reinvestment story, not a solvency story. A total loss is not a realistic scenario.

Recent News & Events

Has the business environment changed recently? Yes — three material 2026 developments: (1) Google’s nationwide home-listings expansion (June 2026) — a new structural threat to portal traffic; (2) Third Point’s exit / end of the activist fight (April 2026); (3) continued residential-segment improvement toward EBITDA breakeven (guided Q2 2026). Sell-side is split (Goldman Buy/$46; Wells Fargo Underweight/$26).

Significant acquisitions / accounting changes / new markets? Domain (Australia) and Matterport integrations ongoing; CoStar commercial platform launched in France (July 2026, €300B TAM), with Germany/Australia launches and European M&A flagged. No adverse accounting-policy change; the un-impaired goodwill is the item to monitor.

APPENDIX B — Source Appendix — CoStar Group, Inc. (NASDAQ: CSGP)

Prepared 2026-07-03. Primary (public) sources first; third-party aggregated data reconciled to filings.

Primary — SEC Filings (EDGAR, CIK 0001057352)

• Form 10-K FY2025 (filed 2026-02-26) — revenue, segments (Commercial vs. Residential), margins, goodwill/intangibles ($4.94B GW; Matterport $1.1B, Domain $1.2B), reportable-segment structure, net new bookings ($308M).
• Form 10-K FY2021–FY2024 — five-year financial history (revenue, EBITDA-margin trajectory, SG&A/R&D build).
• Form 10-Q Q1 2026 and prior quarterlies.
• DEF 14A (2026-04-30) — beneficial ownership (Florance 3,044,568 sh / ~0.75%; insiders 1.18%; Vanguard 9.38%, BlackRock 7.79%), compensation ($36.4M CEO 2025), Capital Allocation Committee membership.
• DEF 14A (2023, 2024) — Michael R. Klein listed as Chairman (governance history).
• DEFA14A (2025-06-13) — defensive say-on-pay solicitation (vote ~53%); DEFA14A (2026-05-18) — activist-designated directors.
• 8-K (2025-04-07) — activist Support Agreements; board refresh; independent chair (Sams); Capital Allocation Committee.
• Form 4 filings — Andrew Florance open-market purchases: 14,731 sh @ $74.67 (2024-10-25); 55,720 sh @ $44.52 (2026-02-27); 71,430 sh @ ~$35.2 (2026-05-01). Cluster buys: Frederick Saint (20,000 @ $45.33), Rachel Glaser, Louise Sams. (Accessions 0001066061-26-000012, -26-000006; 0001057352-24-000127.)

Primary — Company Disclosure

• Q1 2026 earnings call transcript (2026-04-28) — segment revenue/EBITDA, Homes.com KPIs (35,175 subscribers, $106M ARR, 11× ROI claim), guidance ($3.78–3.82B revenue; $780–820M adj EBITDA; $1.32–1.39 adj EPS; Residential breakeven Q2’26), buyback detail, renewal rates.
• CoStar Q1 2026 press release (2026-04-28), investors.costargroup.com.
• CoStar press releases — Domain completion (2025-08-27), Matterport completion (Feb 2025), France launch (2026-07-01, businesswire), Wikicasa 30% stake (Jul 2026).

Quantitative Data Providers (reconciled to filings)

• Company financial statements (10-K/10-Q via EDGAR) — income statement, balance sheet, cash flow, enterprise value, valuation multiples (own-history P/S 3.7×→8.6×→14–21×; P/B 1.5×), per-share and profitability metrics. Where third-party ratio feeds were unreliable (ROE/ROIC, GAAP P/E), figures were recomputed from the filings.
• Own-history valuation range — P/S (~3.7×) and P/B (~1.5×) at their lowest in CoStar’s public history; GAAP P/E disregarded (near-zero GAAP EPS). Five-year daily price series.
• Factor/price-action data — risk-adjusted track record (y1 −63%, m6 Sharpe −1.7, 5yr −19%), factor loadings (beta ~0.8–1.0; factor-similar peers SS&C, Autodesk, GoDaddy, Genpact), stock-info (RS vs peak −70%).

Industry, Competitive & News Sources (public)

• Zillow Group Q1 2026 results (2026-05-06), zillow.mediaroom.com — residential rev +8% to $450M; rentals +42% ($183M), multifamily +57%; ~230M uniques.
• The Real Deal (2026-04-27), therealdeal.com — CoStar residential losses (−$361M 2024, −$230M 2025), Third Point exit, staff cuts, spend trim.
• The Real Deal / HousingWire / Inman (2026-06-11) — Google/HouseCanary 50-state home-listings expansion; analyst Jake Fuller commentary; Barron’s market reaction (Zillow −4.9%, CoStar −3.6%).
• Inman (2026-04-13; 2025-07-31) — Third Point drops fight; Florance “portal wars” claims.
• HousingWire / OnlineMarketplaces — NAR/Sitzer-Burnett $418M settlement (effective 2024-08-17); portal implications; Homes.com “your listing, your lead” model.
• Reuters / Hedgeweek (2026-04-10/13) — Third Point sells entire stake, abandons proxy fight.
• CNBC (2025-04-12) — D.E. Shaw sub-1% stake, governance engagement; D.E. Shaw “Perspectives on CoStar” (Feb 2026).
• SEMrush (Mar 2026) — Homes.com traffic mix (~41.6% Google organic, ~28.6% direct).
• Competitor / industry references — latterly.org, dcfmodeling.com, mbi-deepdives.com, onlinemarketplaces.com, bisnow.com, compstak.com (CoStar Suite share, ~1,500 researchers, competitor set); virginiabusiness.com (Matterport $1.6B); eliteagent.com/businesswire (Domain ~A$3B).
• Analyst actions (via news feed) — Goldman Sachs Buy, PT $46 (2026-06-25); Wells Fargo Underweight, PT $26 (2026-06-24).

Third-party aggregated data is treated as a hypothesis and reconciled to the underlying filing; where vendor ratio fields were unreliable (ROE/ROIC, GAAP P/E), figures were recomputed from the 10-K.