The EBITDA multiple (EV/EBITDA) is a valuation ratio comparing a company's enterprise value to its earnings before interest, taxes, depreciation, and amortization. Enterprise value equals market capitalization plus net debt. An EV/EBITDA of 15x means investors are paying 15 times annual cash earnings for the business. Lower multiples suggest cheaper valuations; higher multiples imply growth expectations or premium quality.
EV/EBITDA is preferred over P/E ratios for cross-border comparisons because it neutralizes differences in capital structure (debt vs. equity), tax regimes, and depreciation policies. In the AI infrastructure space, multiples vary dramatically: semiconductor equipment companies trade at 20–30x EBITDA, mature chipmakers at 8–15x, and hyperscalers at 15–25x. The key insight is that headline multiples can be misleading — a company trading at 25x current EBITDA but growing EBITDA at 30% annually is cheaper on forward earnings than a company at 12x growing at 5%.
EBITDA multiples are the universal language of business valuation. When a SaaS company trades at 20x EBITDA and a semiconductor company trades at 12x, the gap reflects both growth expectations and risk perception. Multiple compression — when the market decides a company deserves a lower multiple — is one of the most powerful forces in equity markets, capable of cutting a stock price in half even as earnings grow.
Closelook's Software-Credit Nexus thesis argues that agentic AI will force systematic EBITDA multiple compression across the SaaS sector. Companies currently valued at 15-20x EBITDA on the assumption of durable pricing power face structural margin pressure as AI agents can replicate their core functionality. Understanding where multiples should be — not where they are — is the core analytical challenge.
EBITDA multiples connect to the SaaSpocalypse thesis, the Software-Credit Nexus, and the Economic Moat framework that determines which multiples are sustainable.
