CoStar Group looks too cheap for its growth, expanding margins and network-effect assets. Its unique data network and marketplaces have durable network effects, and it could lead in agentic real-estate tech, with Homes AI an early look. If earnings compound over 20% organically for three years, below consensus, a high-teens to 20x 2027 P/E is probably too cheap. Q1 EBITDA more than doubled on lower personnel costs and AI efficiencies, and residential should turn profitable in Q2. x.com  x.com  … Show more

CoStar is super cheap and left for dead even though its proprietary data makes the business bulletproof against AI. All of software is getting a bid as the sky has not fallen, and the hostile activists have been forced to move on. It ranks as a top-conviction idea. x.com 

CoStar is still too expensive because its valuation needs a bigger discount for the AI threat. Its data remains ingrained in brokers' workflows and CEO Andy Florance is still well regarded, but the moat has shrunk. Below $30 there might be a way to structure an opportunity. x.com 

CoStar Group is a favored name because of its massive hold on commercial real estate. x.com