CoStar Group Deep Value Investment Analysis

Revenue repeatability, stability, and dependence. Repeatability is very strong on the commercial real estate side because most contracts renew automatically, are billed through fixed monthly fees, and have high renewal rates. But the company also clearly warns that if real estate markets weaken, customers consolidate, or customers cut spending, cancellation rates will rise. Apartments.com on the residential side is also closer to a "subscription advertising platform," while Homes.com, Domain, and Matterport bring transaction revenue and single-listing promotion revenue that make overall company repeatability slightly weaker than in the past. Customer concentration is not disclosed as a core risk, suggesting that the business is not obviously dependent on any single large customer. The real dependence is on real estate activity levels, traffic acquisition capability, and salesforce efficiency.

Where the industry stands. The commercial real estate information and analytics industry is more like a mature but still steadily growing industry. Demand comes from persistent needs in leasing, investment, financing, asset management, valuation, and market research. Long-term demand in residential portals is also stable because renting and buying homes will not disappear. But competition is more intense, traffic battles are fiercer, and brand marketing consumption is heavier, so it is more like an industry with "long-term demand, but inelegant competition." The U.S. homeownership rate was 65.3% in Q1 2026, the rental vacancy rate was 7.3%, and home transaction activity was weak but had not disappeared. NAR disclosed existing home sales at an annualized pace of about 4.02 million units in April 2026 and about 3.63 million units in March. This supports the judgment that demand is long-lasting while short-term cycles are significant.

Competitive landscape. In U.S. residential portals, Zillow remains one of the largest platforms: its 2024 average monthly unique users were about 221 million. Move, which owns Realtor.com and is part of News Corp, disclosed about 62 million monthly active users and about 241 million average monthly visits in News Corp's fiscal Q2 2026. CoStar itself disclosed that its full website network attracted 131 million average monthly unique visitors in Q1 2026, while Homes.com had about 35,000 members at that time, up more than 200% year over year. This shows that Homes.com is clearly growing, but it has not yet built traffic dominance at the same level as Zillow. The commercial real estate side is different: since its founding in 1987, CoStar has looked more like the strongest platform in U.S. commercial real estate data and online marketplaces, with competitive advantages far stronger than those of its residential portal business.

Would I be willing to hold it if the stock market closed for five years. View: If the stock market closed for five years, I would be willing to hold CoStar's core commercial real estate assets. But whether I would be willing to hold the entire CSGP at today's price for five years is a more cautious question, because what you are buying is not only a strong-moat commercial real estate database, but also two variables still being tested: residential portal expansion and cross-border acquisition integration. Business understandability score: 4/5. Industry attractiveness score: 3/5. The core logic behind these scores is: demand is stable and the core is strong, but the company's overall complexity has risen, while the competitive intensity of the residential portal track reduces the purity of the whole company as a "simple good business."

Moat and Management

Moat. Data advantage: strong. Since its founding, the company has continuously built standardized real estate databases. In early public filings, it explicitly said it had the "most comprehensive commercial real estate database" and the "largest commercial real estate research department" in the industry. This kind of database cannot be replicated simply by buying traffic. It depends on years of data collection, cleaning, standardization, customer feedback loops, and a national sales and research organization.

Network effects: medium-strong, but very different by business. Platforms such as LoopNet, Apartments.com, Homes.com, Domain, and OnTheMarket all have typical two-sided marketplace features: more traffic attracts more listings and advertising budgets, and more listings then attract more traffic. The issue is that the network effects of commercial real estate and rental portals are already more mature, while Homes.com remains a challenger in U.S. resale home portals. Zillow's monthly unique user scale is still higher than CoStar's full network, and Realtor.com still has meaningful traffic. So CSGP's residential network effects are not yet as solid as its commercial real estate network effects.

Switching costs: medium-strong. Fixed monthly fees, terms of at least one year, automatic renewal, and a renewal rate around 89% show that customers are not using these products as one-off purchases. They are embedded in daily workflows. Especially once professional real estate data and research tools enter investment committee processes and the daily judgment systems of brokers and investors, replacement is not easy. Still, this switching cost is not "deeply locked in" like ERP. It is more an efficiency dependence + data dependence + team habit dependence.

Brand, scale, and channels: medium-strong. The company owns multiple brands, including CoStar, LoopNet, Apartments.com, Homes.com, STR, Ten-X, Matterport, Domain, and OnTheMarket. Scale advantages show up in two ways: first, the company can maintain a large research and sales organization; second, it can use cash flow from commercial real estate and rental businesses to subsidize brand building and product development for Homes.com. The board even created a capital allocation committee to review Homes.com's long-term financial model, breakeven point, and net present value analysis before recommending that the board launch a new $1.5 billion repurchase program and the 2026 budget. For long-term owners, this at least indicates that the company formally recognizes the seriousness of capital allocation.

Cost advantage, regulatory barriers, patents and licenses: limited. CoStar's advantage is not winning through lower prices. It is "more complete and deeper data with higher commercial value." That means it is not a typical cost leader. Regulatory licenses and patents are not its main barriers. Matterport's 3D spatial technology and historical accumulation create some technical threshold, but this business is not large enough within the group to form a decisive moat.

Is the moat widening, stable, or narrowing. View: The commercial real estate core moat remains stable and may be widening slightly. At the group level, the moat is "stable but increasingly differentiated." Commercial real estate and rental platforms continue to benefit from data accumulation, customer inertia, and brand. In residential portals, CoStar is still spending money to compete for user mindshare, which shows that this part of the moat has not fully settled. In other words, the core assets have a strong moat, but the consolidated company's moat is not as pure as the commercial real estate core. Moat strength score: 4/5.

Is management trustworthy. Andrew Florance remains founder and CEO, which is a positive. But in terms of ownership, as of April 2026 he held about 3.045 million shares, still less than 1%. All directors and executives together held about 1.18%. This means the company is founder-led, but it is not a typical "high equity alignment" owner-operator model. For conservative investors, this is still some distance from Buffett's preferred setup of managers owning a lot of stock and being deeply aligned with minority shareholders.

Is capital allocation rational. The judgment here must be mixed. On the positive side: the company has long paid no dividend and has used more capital for reinvestment, acquisitions, and repurchases. Before 2026, it also introduced a capital allocation committee to review the Homes.com budget and return profile. In Q1 2026, it continued executing the new $1.5 billion repurchase program. By the end of March, under that program, it had repurchased 11.4 million shares, spending $505 million, at an average price of about $44.44/share, with about $995 million remaining.

On the negative side: under the old 2025 repurchase program, the company repurchased a total of 7.1 million shares, spending $500 million, implying an average cost of about $70/share, while the current share price is about $34. At minimum, this shows the prior repurchase round did not occur at an obviously undervalued level. In other words, CoStar does not deserve a high score on whether it can deploy capital at the best per-share value points.

Do acquisitions create value. From 2023 to 2025, the company acquired OnTheMarket, Visual Lease, Matterport, and Domain in succession. Matterport alone had total consideration of about $1.927 billion, including $902 million of cash and 11.7 million CoStar shares. Domain's total purchase consideration value was about $1.912 billion, with the large step-up ultimately landing mainly in $1.244 billion of goodwill. Matterport created about $1.105 billion of goodwill. Acquisitions of this scale are successful only if they deliver excess returns over many years. Based on current information, it is not possible to conclude that they have already created value. The right description is: still in the validation period.

Incentives and candor. The CEO's total 2025 compensation was about $36.43 million, with a large portion from stock and option awards. The company's risk disclosure is generally adequate. The 10-K clearly states that weakness in real estate markets, customer consolidation, rising cancellation rates, and insufficient sales team efficiency can all hurt growth. The proxy statement also disclosed the review process for Homes.com's long-term budget and NPV analysis, which is better than many growth companies that only tell stories. But from a conservative shareholder's perspective, high compensation + low management ownership + acquisition-led expansion still means capital allocation should be discounted. Management and capital allocation score: 3/5.

Financial Quality

The following table is mainly compiled from CoStar's 2021-2025 10-K filings. ROE/ROA are approximate calculations in this report based on average beginning and ending assets and equity. "Free cash flow" is split into two measures: one including all capital expenditures, and one excluding "campus/campus construction capital expenditures," to help understand the company's true operating cash generation capacity. Because the company holds significant cash, has acquisition amortization, and has gone through segment reorganization, a precise five-year ROIC series is highly dependent on methodology. This report does not invent a single precise number, and instead discusses directional changes after the table.

How to read the financial quality. The revenue side is strong: from 2021 to 2025, revenue increased from $1.944 billion to $3.247 billion, a four-year CAGR of about 13.7%. But the profit side has deteriorated sharply: operating profit fell from $432 million in 2021 to an operating loss of $72 million in 2025, while net income fell from $293 million to $7 million. This is not a traditional case of "the business has collapsed." It is more like the company reallocating the cash flow of its formerly high-profit commercial real estate/rental businesses into residential portal brand wars, acquisition integration, and related amortization. Q1 2026 segment data confirms this: commercial real estate EBITDA was $127 million, while residential real estate EBITDA was -$46 million.

Gross margin and cost structure. Gross margin declined slowly from 81.6% to 78.9%, showing that the underlying business remains highly software-like and data-like, but transaction businesses, hardware/services, and acquisition-driven cost structure changes are diluting margin. The real cause of the collapse in operating margin is not gross profit, but sales and marketing expense. In 2025, sales and marketing expense excluding amortization of customer base was as high as $1.560 billion, or 48% of revenue. In 2021, that ratio was only 32%. This is a very clear fact: recent profit pressure at CoStar is essentially an active decision to spend heavily to gain residential share.

Cash flow quality. Operating cash flow explains the situation better than net income. In 2025, net income was only $7 million, but operating cash flow was still $430 million. In Q1 2026, operating cash flow was $152 million, clearly higher than $53 million in the same period of 2025. This means accounting earnings currently understate the company's operating cash generation capacity, for reasons including depreciation and amortization, stock-based compensation, and accounting treatment of working capital items. In other words, CoStar's current GAAP earnings look poor, but cash flow has not dried up.

But cash flow is not "fully free" either. The company has long had significant cash absorption from deferred commissions. This item increased from about $102 million in 2021 to $184 million in 2025. In the 2025 cash flow statement, changes in deferred commissions used $152 million of cash. This shows that growth is not costless. To acquire and retain subscription contracts, the company needs to pay sales commissions upfront, capitalize them, and then amortize them gradually. For owner earnings, this is a real cash requirement and cannot simply be ignored.

Capital expenditure intensity. Total capital expenditures were large in 2024 and 2025, especially $579 million of spending on a new campus in 2024 and still $307 million in 2025. If these are all mechanically treated as maintenance capital expenditures, free cash flow looks very poor. But the company also separately presents this portion and has stated in historical filings that these investments include new office buildings, land, and corporate aircraft, which are more expansionary or one-time in nature. For long-term owners, the more meaningful approach is to observe both measures: "including all capital expenditures" and "excluding campus-type capital expenditures." The latter shows that the core operating cash capacity is still not weak.

Balance sheet and leverage. The company has about $1.0 billion of long-term debt on its books, mainly 2.8% Senior Notes due in 2030. As of Q1 2026, the company had $1.215 billion of cash and cash equivalents, $101 million of restricted cash, and net long-term debt roughly around zero to a modest net cash position. Because the company still has investment income, the traditional meaning of "interest coverage" is less important. More importantly, survival risk is low. This is not a company that would be crushed by debt as soon as the real estate cycle weakens.

Share count, dividends, and repurchases. The company pays no dividend. Cash is mainly used for reinvestment, acquisitions, and repurchases. Year-end shares outstanding increased from 394.9 million shares in 2021 to 417.9 million shares in 2025, so there has not been meaningful per-share shrinkage over this period. In 2025, the company also issued an additional 11.7 million shares for the Matterport transaction. Therefore, even though the company has begun large-scale repurchases, from the perspective of long-term shareholders over the past few years, dilution has not been fully reversed.

Accounting risk and signs of manipulation. I do not see obvious recent red flags for financial fraud or aggressive accounting, such as major restatements, control failures, or sudden abnormal surges in receivables. The company's 10-K still has 404(b) auditor coverage over internal controls. In 2024 and 2025, the main internal control references were about integration work for recently acquired assets. The real issue is not "fraud smell," but whether business investment and acquisition amortization will keep GAAP and true shareholder returns far apart for a long time, and whether acquired asset impairment could emerge in the future. Here, special attention should be paid to Matterport litigation reserves and subsequent goodwill quality. Judgment: profits are not fake, but they should not be casually glossed over with the phrase "cash flow is good." This looks more like a company currently reinvesting away its profits.

Owner Earnings and Valuation

Current valuation snapshot. Fact: Near the date of this report, CSGP's share price was about $34, equity market value was about $13.9 billion to $14.1 billion, and enterprise value was about $13.8 billion. Revenue for the latest 12 months, roughly calculated as full-year 2025 plus Q1 2026 minus Q1 2025, was about $3.412 billion. Latest 12-month EBITDA, roughly calculated using the company's disclosed quarterly data, was about $252 million. Latest 12-month operating cash flow was about $529 million. Inference: This corresponds to EV/Revenue of about 4.1x and EV/EBITDA of about 55x. If measured by latest 12-month free cash flow including all capital expenditures, the share price also corresponds to a P/FCF above 80x. This is clearly not a "cheap stock" valuation.

How to estimate owner earnings. The difficulty in Buffett-style owner earnings is distinguishing which capital expenditures are maintenance and which are expansionary. It also requires a view on whether stock compensation is a real cost to shareholders. For CoStar, I do not recommend giving only one point estimate, but rather a range from looser to more conservative.

Looser measure: Start with 2025 operating cash flow of $430 million and deduct only "non-campus/recurring capital expenditures" of about $82 million, giving about $348 million. This measure is closer to "how much after-tax cash flow the operating assets themselves have generated." More conservative measure: On that basis, treat part of stock-based compensation as a real economic cost, or understand it as the future repurchase cost required to offset dilution. 2025 stock-based compensation of $194 million was high and included acquisition-related factors, so it should not be mechanically deducted in full. But if $100 million to $120 million is conservatively treated as a "replacement cost for long-term shareholders," then 2025 owner earnings are closer to $230 million to $250 million. View: Therefore, I prefer to view CoStar's current "conservative owner earnings capacity" as about $230 million to $350 million, rather than simply treating all $430 million of operating cash flow as distributable cash.

Normalized capacity in 2026. The company's full-year 2026 guidance includes net income of about $213 million to $241 million, EBITDA of about $606 million to $646 million, and residential real estate EBITDA expected to turn positive for the full year at $50 million to $70 million. Inference: If the residential business really turns from a continuous profit drain into a modest profit contributor, and operating cash flow continues improving, CoStar's normalized owner earnings after 2026 may rise into the $350 million to $500 million range. That is precisely why the current share price is not really betting on 2025, but on "recovery earnings in 2026-2028."

What multiple of owner earnings does the current price imply. Based on the above ranges: Using $350 million of looser owner earnings, current equity value is about 40x. Using $230 million to $250 million of more conservative owner earnings, current equity value is about 56x to 61x. Neither number has typical value-stock appeal. Conservative Owner Earnings estimate: $230 million to $250 million. Looser but defensible estimate: around $350 million. All of the above are this report's inferences, not company-disclosed figures.

Discounted owner earnings valuation. To avoid giving only the conclusion, the assumptions are laid out below. Here, owner earnings are treated as after-tax cash flow to shareholders, so net debt is not added or subtracted separately. A small net cash position is treated only as a marginal safety cushion.

• Conservative scenario assumptions: 2026 owner earnings of $300 million; 8% annual growth for the first 5 years; 4% annual growth for the next 5 years; 10% discount rate; terminal value at 18x owner earnings.

• Base scenario assumptions: 2026 owner earnings of $380 million; 12% annual growth for the first 5 years; 5% annual growth for the next 5 years; 9% discount rate; terminal value at 22x owner earnings.

• Optimistic scenario assumptions: 2026 owner earnings of $500 million; 15% annual growth for the first 5 years; 6% annual growth for the next 5 years; 8.5% discount rate; terminal value at 25x owner earnings.

Under these assumptions, this report estimates per-share intrinsic value at roughly: $14-15 conservative, $26-27 base, and $44-45 optimistic. This immediately shows that once you apply even a modest discount to growth, margins, and capital allocation returns, the current share price is not cheap.

Relative valuation. Comparing CoStar with several types of peers clarifies the issue. It should be noted that there are not many truly perfect listed peers. The following group includes two categories: "real estate portal/data platforms" and "high-quality data and information platforms." Stock Analysis statistics show Zillow at about 71.5x EV/EBITDA and 24.3x EV/FCF; S&P Global at about 17.2x EV/EBITDA and 24.4x EV/FCF; Moody's at about 21.9x EV/EBITDA and 30.4x EV/FCF; and FICO at about 27.5x EV/EBITDA and 35.5x EV/FCF. CoStar's latest standardized FCF yield is only about 1.21%, and ROIC is about -0.16%. Inference: CoStar has not yet delivered high and stable capital returns like MCO, SPGI, and FICO, nor is it a purer residential portal recovery trade like Zillow. Its current valuation looks more like "the market assigning a combined premium for high-quality core assets and the option value of residential turning profitable." This is not cheap. It is high expectations.

Asset value or liquidation value. This company is not suitable for a liquidation-value purchase approach. Total equity at the end of 2025 was about $8.371 billion, but $4.944 billion of that was goodwill and $1.771 billion was intangible assets. In other words, tangible net asset support is far thinner than book equity suggests. By Q1 2026, the company still had close to $10.0 billion of total assets, but the main assets that can be quickly realized and truly protect shareholders are cash, not database goodwill itself. View: For CSGP, liquidation value gives you almost no margin of safety. Buying it must rest on the premise that it will continue as an operating business and generate cash flow above its cost of capital.

Valuation conclusion.

• Conservative intrinsic value range: $20-26/share

• Fair intrinsic value range: $27-35/share

• Optimistic intrinsic value range: $40-50/share

On this basis, the current price is roughly near the upper end of the fair range. Ideal buy price range: $22-28. Acceptable hold price range: $28-38. Clearly overvalued range: above $45. These ranges are essentially assumption-driven views, but their key constraints are clear: the residential business must deliver profitability, the commercial real estate core cannot stall, and capital allocation cannot keep making large-scale mistaken repurchases at high valuations.

Margin of Safety and the Bear Case

Is the current price cheap enough. My judgment is: no. This does not mean it is necessarily outrageously expensive. It means it is not cheap enough for a "balanced but conservative" long-term investor to easily ignore residential execution volatility and acquisition integration risk. The market is already trading the improvement expected in 2026-2028, not giving you a bargain at the 2025 trough.

The most fragile valuation assumption. The most fragile assumption is not "whether revenue can grow," but post-improvement residential margins and cash conversion. If growth at Homes.com/Domain requires sustained high marketing spend to maintain, then revenue growth may not translate into owner earnings. The company itself disclosed that residential was a clear drag in 2025, and only guided full-year EBITDA to turn positive in 2026.

Does the investment still work if growth falls short. If the commercial real estate core remains steady and the residential business improves gradually rather than explosively, long-term survival should not be impaired. But buying at the current price would likely leave returns only in the low to mid single digits, making it hard to outperform the index meaningfully. The current valuation does not give you a low starting point. It gives you a "fair, but not cheap" starting point.

If margins decline or the valuation multiple contracts. From 2021 to 2025, the company's operating margin fell from 22.2% to -2.2%, which already provided a preview: even if the core business remains healthy, when group-level capital allocation shifts toward high-investment expansion, margins and market patience can deteriorate quickly. If future residential investment produces weak results and the market reprices the company as a "low cash return platform stock," valuation multiple contraction would create very real capital losses.

The strongest bear case. The bear would say: CoStar's best assets are actually its commercial real estate database and Apartments.com, yet the company is putting large amounts of capital into the most brutally competitive residential portal battlefield, where traffic is heavily dependent on brand spending. Acquisitions such as Matterport and Domain increase complexity and goodwill, and may not deliver commensurate returns in the future. Management repurchases also have not obviously occurred during undervalued periods. I think this is a very strong bear case that must be taken seriously. It captures CSGP's central tension today: a company that once looked like a high-quality subscription data business is actively turning itself into a more complex "subscription + traffic + acquisition integration" portfolio.

What facts would overturn the investment judgment. If the following facts emerge, I would acknowledge that the original judgment needs to be rebuilt: First, commercial real estate subscription renewal rates fall materially below about 89% and drop under 85% for multiple quarters. Second, the residential business fails to turn EBITDA positive in 2026-2027, or turns positive but free cash flow remains poor. Third, acquisitions such as Matterport, Domain, and OnTheMarket show clear impairments or returns remain far below cost for a long time. Fourth, Home.com's traffic/member growth stagnates, while Zillow and Realtor widen their share gaps again. Fifth, the company keeps using large amounts of cash for repurchases during overvalued periods without a corresponding increase in intrinsic value per share.

Largest permanent capital loss scenario. I do not think the main risk is "the company goes bankrupt." It is good assets diluted by bad capital allocation and an excessive purchase price. In this scenario, the commercial real estate core still has value, but residential portal expansion earns average returns, acquisitions bring impairments, and the market values the company at a more reasonable 20-25x owner earnings. The share price then fails to return to the purchase cost for a long time. In the conservative case, a decline into the low $20s is not unimaginable.

Checklist and Final Judgment

Comparison with other opportunities. If CSGP is compared with the strongest peers or adjacent companies, I would rather compare it with S&P Global, Moody's, and FICO, which are "information/scoring/data platforms," rather than only with Zillow. The reason is simple: if you pay a high valuation, you at least want to buy companies that have already delivered high returns, high cash conversion, and strong pricing power. Compared with these mature high-quality platforms, CoStar's core commercial real estate assets are not weak, but the group's ROIC, margins, and owner earnings are currently much weaker.

If compared with the index, my conclusion is: buying it now does not provide a clear case for outperforming the S&P 500. On one hand, the historical performance chart in the company's 2025 10-K shows that from the end of 2020 to the end of 2025, CoStar's share price materially lagged the S&P 500. On the other hand, the current U.S. 10-year Treasury yield is about 4.67%, meaning any equity investment with execution risk should offer a clearly sufficient risk premium. At CSGP's current valuation, the conservative scenario struggles to offer enough excess return.

Investment checklist.

Final rating: Watch. One-sentence investment thesis: CoStar's commercial real estate core business is excellent, but the listed company currently looks like a good car pulling a heavy trailer: it can run, but it is not light. Around the current $34 price, it is more "worth tracking" than "worth acquiring heavily."

Core bull case. The underlying business models of CoStar/LoopNet/Apartments.com remain high quality, with a high subscription revenue mix and stable renewal rates. The commercial real estate segment still generated $127 million of EBITDA in Q1 2026, showing that core asset cash flow is healthy. The balance sheet is solid, net leverage is low, and the company can withstand cyclical weakness while continuing to invest. Management's 2026 guidance calls for the residential segment to turn EBITDA-positive for the full year. If delivered, group profitability would improve meaningfully.

Core bear case. Residential portal expansion is still being validated and faces strong competitors such as Zillow and Realtor. Group margins and returns on capital have deteriorated sharply in recent years. Acquisitions have brought large goodwill and intangible assets, and returns have not yet been fully proven. The current price remains expensive relative to owner earnings, with insufficient margin of safety. Past repurchase timing also cannot be called cheap.

Key assumptions. For the investment to work, the commercial real estate core must maintain high renewal rates and steady growth. Homes.com/Domain must convert traffic and member growth into sustainable profitability. Deferred commissions, marketing expense, and SBC must not consume operating cash for too long. Acquisition goodwill must avoid large impairments. Management must focus more on per-share value than scale when making repurchase and new investment decisions.

Fair buy price. I think the more comfortable long-term buy range is $22-28. If you require at least a 20%-30% discount to base intrinsic value, it is better to seriously consider it around $25 or below. The current price around $34 is closer to a level where investors can keep monitoring fundamental delivery.

Target holding period. If ultimately purchased, this type of company should be viewed over at least 5-10 years, or even longer, because real value realization depends on the residential business moving from investment phase to harvest phase, not on EPS fluctuations over the next one or two quarters.

Expected annualized return. At the current price, a rough estimate is: Conservative scenario around 0%-3%; Base scenario around 6%-9%; Optimistic scenario around 12%-15%. The meaning of this set of numbers is important: the base-case return is not bad, but for a "balanced but conservative" investor, it is not high enough to require action now.

Maximum downside risk. If residential returns fall short, acquired assets are impaired, and the market simultaneously compresses the valuation multiple, a 30%-40% permanent capital loss over the next 2-5 years would not be exaggerated. In an extreme case, if the market is only willing to price the company at conservative intrinsic value, a share price in the low $20s is imaginable.

Tracking indicators. The most important indicators to track are: commercial real estate segment EBITDA and net new bookings; whether subscription renewal rates stay around 89%; Homes.com membership and residential segment EBITDA; operating margin/operating cash flow/deferred commission trends; SBC intensity and repurchase pace; integration results for Domain, Matterport, and OnTheMarket; whether goodwill and intangible assets show impairment signals; and changes in shares outstanding on a per-share basis.

Signals that should trigger reassessment. If commercial real estate renewal rates decline for consecutive periods; residential real estate EBITDA deteriorates again; the company keeps relying on heavy marketing to sustain growth while per-share cash flow does not improve; acquisition litigation or impairments expand; or management continues aggressive repurchases and acquisitions at high valuations, the investment logic should be reassessed immediately.

Final recommendation. Calmly stated, CoStar is not a bad company. On the contrary, it owns very valuable core assets. But a good company is not the same as a good price. From the perspective of a long-term business owner, I would rather own it at a lower purchase price than bet at the current price on perfect residential execution. For conservative long-term capital, the most rational action is not to chase an "excellent business" at a high price, but to keep tracking it, wait for a thicker margin of safety, or wait until fundamental improvement truly translates into more solid owner earnings before deciding whether to pay a higher price.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.