CSGP — CoStar Group, Inc.

The tape and the narrative agree: as of 2026-09-13 CSGP was down 30.2% over six months and more than 50% year-to-date, sitting near $30.46 against a $90.72 52-week high. The proximate cause is the July 28, 2026 Q2 call, on which management cut FY revenue guidance to $3.715-3.755B (implying ~15% growth, down from a ~17% prior midpoint) after net new bookings fell 26% YoY, and disclosed a deliberate 21% reduction of the Homes.com sales force. Sell-side sentiment has cooled from a prior bull consensus to roughly 16 Buys, 9 Holds and 2 Sells, with Baird downgrading to Neutral in July and RBC holding a cautious Sector Perform with a $34 target in September. The offsetting bullish signals are thin but real: management raised FY adjusted EBITDA guidance to $780-820M, CEO Andrew Florance has been buying stock in the open market, and the Apartments.com/ChatGPT integration is showing ~40% referral-traffic growth since August 2026. That has not been enough to change the tape.

Revenue tells a decelerating-but-still-growing story: $624.7M in Q3 2023 to $925.0M in Q2 2026, with ten straight sequential increases into Q4 2025, a trivial $3M dip in Q1 2026, then a new high. But YoY growth has clearly decelerated: +26.9% in Q4 2025, +22.5% in Q1 2026, +18.4% in Q2 2026, and guidance implies ~15% for FY2026. Profitability tells the opposite, and more interesting, story: operating income went -$42.8M (Q1 2025) to -$27.2M, -$51.1M (-6.1% margin, Q3 2025), then inflected to +$49.5M (Q4 2025), +$3.0M (Q1 2026) and +$76.0M / 8.2% margin in Q2 2026 - the best operating margin since Q4 2023's 10.8%, and $76M versus a -$27.2M loss in the year-ago quarter. Net income has been positive for three straight quarters ($47M, $3M, $55M). The genuine weak spot is cash: trailing-four-quarter free cash flow is roughly -$257M (OCF $497M versus capex $754M), and cash on the balance sheet has fallen from $4.68B at 2024-12-31 to $1.27B at 2026-06-30. Q2 2026 FCF of +$12M was the first positive print in five quarters, so the cash trend is stabilizing, but this is not yet a cash-generative business at current spending levels.

Judged on trailing numbers CSGP looks absurd (165.8x TTM P/E, negative PEG, 2.2% operating margin, -0.2% ROIC) and judged against its own five-year history it looks free (3.47x TTM sales versus a 12.73x five-year average; 1.55x book). Both framings are misleading. The five-year averages are polluted by the 2020-22 bubble (average five-year P/E of 899x), so the deterministic 'Cheap vs. its own history' label is weaker evidence than it appears. On forward numbers - consensus FY EPS of $1.366 and FY revenue of $3.75B - the stock trades at ~22.3x earnings and ~3.3x sales, against roughly 15% revenue growth, a 76.5% gross margin, a net-cash balance sheet (cash $1.266B versus LT debt $994M) and EV/EBITDA of ~15x on management's $800M FY adjusted EBITDA midpoint. That is modestly cheap for the franchise quality, but it is a valuation that requires the earnings recovery to show up. It is not a deep-value cushion.

This is a case of the market pricing in a partial earnings recovery while discounting the odds it gets delivered - not the classic 'pricing in growth that hasn't shown up.' At $30.46 and $1.366 of consensus FY EPS, you are paying 22.3x for earnings that are themselves up roughly 7x from the $0.183 TTM base, so the multiple is only 'cheap' if the estimate holds. The single line item that decides it is net new bookings, which fell 26% YoY in Q2 2026 and is a leading indicator of subscription revenue 3-6 quarters out. If bookings stabilize, the FY2027 EPS path toward $1.70-1.80 makes the stock look like ~17x forward-2027 and the current price is a genuine turnaround discount. If bookings stay negative, the $1.366 estimate gets cut, the 'cheap' look evaporates, and the stock can trade in the $20s again. Watch the Q3 2026 bookings line, not the trailing EPS.

As of 2026-09-13, the technical read was neutral-to-weak, not oversold: RSI(14) at 46.77, price $30.46 sitting 3.68% below its 20-day SMA, only 0.38% above its 50-day SMA, and a painful 29.16% below its 200-day SMA. Returns were -1.46% over one week, -0.13% over one month and -7.25% over three months, with a +3.29% bounce on the session off a $29.49 prior close. The interpretation: the waterfall is over - the one-month return is essentially flat - but there is no uptrend, and the 200-day is far overhead. The stock is attempting to base in the high-$20s / low-$30s, and RSI in the mid-40s gives no oversold entry signal on its own. The timing implication is that there is no urgency to chase; a weekly close back above the 50-day that holds, or a higher low above the $25.89 52-week low, would be the first real confirmation. Buying here is a fundamental bet, not a technical one.

• High Treasury yields and sticky inflation keep both commercial real estate transaction volumes and residential turnover depressed - directly reducing the transactional and marketing revenue that supplements CoStar's subscription base.
• Sector rotation has punished heavy-capex growth spenders for two years; CSGP's Homes.com burn made it a poster child and forced the pivot to margin preservation, but the market has not yet re-rated the stock for that pivot.
• Frozen residential turnover (affordability, mortgage-rate lock-in) caps how fast Homes.com can monetize traffic, regardless of how good the product gets.
• Competitive intensity in residential portals is intense - Zillow's dominant traffic and agent-ad products plus realtor.com/News Corp keep Homes.com's customer acquisition cost structurally high.
• AI-driven search is a double-edged structural force: the Apartments.com/ChatGPT integration is lifting referral traffic (~40% growth since August 2026), but LLM intermediation could eventually disintermediate portal models.
• A Fed easing cycle or a thaw in CRE lending would be the single biggest macro tailwind, since CoStar's revenue durability is tied to transaction velocity.

• Q3 2026 earnings (late October / early November 2026) - the critical test of whether net new bookings stabilize from the -26% YoY Q2 decline and whether the 21% Homes.com sales-force cut preserved margin without killing top-of-funnel.
• FY2026 adjusted EBITDA landing at or above the raised $780-820M guidance, which would validate that the cost/AI-efficiency pivot is real rather than a one-quarter accounting effect.
• Conversion and engagement metrics from the Apartments.com-OpenAI/ChatGPT integration, which management says drove ~40% referral-traffic growth since August 2026.
• Further open-market insider buying by CEO Andrew Florance, and any acceleration of buybacks with $1.27B of cash against a $12.3B market cap.
• A turn in interest rates or a CRE transaction-volume recovery, which would lift both the fundamental base and the multiple.
• Any strategic action on Homes.com - a partnership, monetization, or further restructuring - that resolves the biggest capital-allocation question.

• Net new bookings down 26% YoY is a leading indicator for subscription revenue, and it is the one number that can invalidate the consensus $1.366 FY EPS estimate and the 15% growth guidance.
• Cash generation is still negative on a trailing basis (~-$257M TTM FCF) and the cash pile fell from $4.68B to $1.27B over six quarters - the balance-sheet fortress is being spent, and it is finite.
• Homes.com may be a structural capital sink if it cannot win share from Zillow; the bear case is that the last three years of spend bought traffic rather than durable economics.
• Current profitability is close to nil (2.2% TTM operating margin, 0.1% ROE, -0.2% ROIC) - the earnings recovery is a forecast, not an established fact.
• The revenue base is rate-sensitive: renewed yield spikes would hit both CRE transaction volumes and residential velocity simultaneously.
• Medium-term AI disintermediation risk to the portal/search model, which the OpenAI partnership mitigates but does not eliminate.

This is a good franchise trading at a modestly discounted price with a genuine, unproven turnaround attached - not a comfortable value stock. The positives are concrete: 18.4% YoY revenue growth in Q2 2026, a 76.5% gross margin, three straight quarters of positive net income, an operating margin that has inflected from -6.1% to +8.2% in three quarters, net cash of ~$272M, 1.55x book, ~22x forward EPS, and a CEO buying stock with his own money after a 66% drawdown. The negatives are equally concrete: bookings down 26%, a self-inflicted guidance cut, negative trailing free cash flow, and a 200-day moving average 29% overhead. My read is that the risk/reward skews favorable on a 12-24 month horizon because the margin inflection is already visible in the reported numbers while the market is still pricing terminal decline - but the position only works if Q3 bookings show even a modest sequential improvement. This is an early, needs-proof entry, not a safe one, and it is analysis rather than personalized advice.

Blended two forward multiples applied to consensus/guidance inputs as of 2026-09-13: (1) an 17.5x EV/EBITDA multiple on management's $800M FY2026 adjusted EBITDA midpoint, giving a $14.0B enterprise value, plus $272M of net cash (cash $1.266B less LT debt $994M), divided by ~405M diluted shares = $35.25; and (2) a 25x forward P/E on consensus FY EPS of $1.366 = $34.15. A 50/50 blend gives $34.70. The 25x P/E and 17.5x EV/EBITDA multiples are set below the quality-software peer set (Verisk/MSCI-type 25x+ EBITDA) to reflect the 26% bookings decline and unproven Homes.com economics, but above FactSet-type low-teens multiples given 15%+ revenue growth and a 76.5% gross margin.