CoStar Group, Inc. (CSGP)

Priced as of 2026-09-11 — market data goes stale within days; treat this as a starting point, not a current quote.

Adding up the fair value of CoStar's three Strategic Business Units (SBUs) gives a Total Fair Enterprise Value (EV) of about $14.78 billion, against a market EV of $12.24 billion today. That works out to a central fair value of **$36.75 per share** versus a price of $30.47 — a gap of about **+21%**. The stock sits 7% of the way up its own 52-week range and trades at less than half its five-year median earnings multiple, so the gap has not been priced in yet. The single biggest change from the prior run is not a more generous multiple — it is a fresher set of numbers. CoStar's first-half 2026 results and its reaffirmed full-year 2026 guidance show a business materially larger and more profitable than the fiscal-year 2025 figures the previous valuation used, including the Residential segment turning profitable for the first time. The honest caveat: today's price sits **inside** the fair-value band ($27.17 to $46.47), not below it, so conviction is capped at Moderate. The band is wide because CoStar does not disclose how profitable Apartments.com is on its own.

Fair-value band

Rather than a single point estimate, this valuation reports a range: every discretionary judgment call set to its defensible low end, its midpoint, and its high end.

Price sits inside the band — the method does not resolve this name. The decision below comes from the central estimate, but the honest read is that it depends on assumptions this analysis could not settle.

Portfolio at a glance

Key assumptions & swing factors

• Rule C (freshest base) is the single largest change from the prior 2026-09-01 valuation: Commercial adjusted EBITDA moves from $672M (full-year 2025 actual) to a $717M full-year 2026 estimate built from first-half 2026 actuals ($333M) plus a seasonally-adjusted second half; the Residential segment turned adjusted-EBITDA positive (+$12M) in Q2 2026 for the first time; and CoStar reaffirmed full-year 2026 guidance of $3.715-3.755B revenue and $780-820M adjusted EBITDA on 2026-07-28.
• Rule A cross-check narrowed the Commercial comp range rather than keeping its midpoint. The raw set (MSCI 23.73x, Verisk 17.67x, S&P Global 16.20x, FactSet 11.25x EV/EBITDA, all stockanalysis.com 2026-09-11) has a midpoint of 17.49x, which implies 6.10x EV/Revenue at Commercial's 34.9% margin. On that second metric the comps span 4.30x-13.97x - endpoints differing by 3.25x, past the 3x incoherence trigger. The spread is entirely margin-driven, so MSCI (58.9% margin, index-licensing on third-party assets under management) was dropped as not the same kind of business as a 34.9%-margin field-researched data operation. Narrowed range 11.25x-17.67x, midpoint 14.5x used. Re-run cross-check: 14.5x x 34.9% = 5.06x implied EV/Revenue, inside the narrowed 4.30x-8.54x range (endpoints 1.99x apart) - cross-check passes. Note this narrowing lowered the multiple from 17.49x, it did not raise it.
• Rule B (discount once, not twice) for Apartments.com: the Stage 7 Job Audit Safe Cow override was used ONLY to select the comp set - a going-concern marketplace peer (Zillow Group) rather than a sub-scale harvest multiple. No haircut was then applied to that multiple. The prior valuation applied both mechanisms (the override AND a ~15% discount to Zillow, 2.70x down to 2.3x); that second step was not applied here. Central multiple is Zillow's live 2.54x EV/Revenue (stockanalysis.com, 2026-09-11).
• Revised Cash Cow growth-matching rule was tested and NOT triggered. Commercial's revenue grew 11.5% year-over-year in the first half of 2026, faster than all four comps' three-year revenue growth forecasts (2.93%-9.68%). But the multiple applied is EV/EBITDA, and Commercial's adjusted EBITDA grew only 6.7% - mid-range versus the comps, with margin compressing from 36.5% to 34.9%. The midpoint therefore stands as a genuine central estimate, not a floor.
• The prior run's second swing factor remains unresolved and was NOT fabricated: no sourced split exists for how the Residential segment's adjusted EBITDA divides between Apartments.com and Homes.com. Per the Segment Profitability Fallback Rule, EV/Revenue is used for both and no margin is estimated by allocating costs. This is expressed as band width instead - roughly $6.15 of the $19.30 per-share band.
• Scope flag on the Apartments.com line: CoStar no longer discloses brand-level Residential revenue, so the $1,662M base (Residential Q2 2026 $444M annualised, less Homes.com's $114M run-rate) also contains Land.com, Domain (Australia) and OnTheMarket. These are included rather than dropped, because silently omitting an operating business would understate the Sum-of-the-Parts total.
• Price sits INSIDE the Rule D band ($27.17 conservative / $36.75 central / $46.47 generous), 12% above the conservative end and 34% below the generous end. Per the Decision Framework this means the method does not fully resolve this name, and conviction is capped at Moderate regardless of the +21% central gap.
• Conviction is independently capped at Moderate by the largest-component rule: Commercial is 70% of Total Fair Enterprise Value and inherits Stage 4's flagged borderline growth call (6-8% CAGR versus an outlying 11.0% estimate, straddling the 10% High/Low Growth threshold that separates Cash Cow from Star).
• Sum-of-the-Parts sanity check, named rather than closed: Total Fair EV of $14.78B is 17% above the market EV of $12.24B. Holding the Residential SBUs at central value, the market implies just 11.0x adjusted EBITDA for Commercial - below FactSet's 11.25x, the cheapest comp in the set. Likely cause is that CoStar still trades as one blended equity story after a 65% twelve-month drawdown driven by whole-company events (a July 2026 guidance revision and a CFO departure), plus a Residential profitability inflection that is only one quarter old. No multiple was adjusted to close this gap.
• Fact change since the prior run: Third Point sold its entire CoStar stake in April 2026, ending its activist campaign. The prior valuation cited active activist pressure to divest Homes.com as support for a harvest multiple; that pressure has lapsed. Homes.com's Dog classification still holds on Stage 4 share data (#4 of four US portals, ~6% of portal visits) and Stage 7 job scores (12/25 paying job, 6/25 demand-side job).
• Market prices go stale in days, not months. This decision reflects prices as of 2026-09-11 and must be re-run, not assumed current, before acting on it.

Portfolio-level tensions

• Matterport is a correlated, shared resource across Commercial (nearly 300,000 digital twins on CoStar/LoopNet) and Homes.com (feeding Homes AI) - a licensing or performance shock would degrade both SBUs' differentiation together, even though it is not a Rule 6 shared-job case
• CoStar holds no Rule 8 landlord position anywhere in the portfolio - Yardi, RealPage, and Entrata are the landlords over Apartments.com's PMS integrations, and mass-media advertising is the effective landlord for both demand-side search jobs; the entire moat is built on owning jobs directly, never on renting out ground to others
• The capital-allocation case against continued Homes.com funding is no longer ambiguous - the raw BCG share/rank data and the independent Job Audit (Contested paying job, No hold demand-side job) now agree it is a real Dog, converging with CoStar's own January 2026 spending cut and Third Point's activist pressure
• Apartments.com's renter-search job and Homes.com's homebuyer-search job are both Rule 5 rented positions exposed to the same diagonal AI-assistant threat (Chain Rule 5) that could absorb search behavior as a side effect of a bigger job