US · the leverage tail
The 2021–22 software vintage (Citrix, Proofpoint, Anaplan, Coupa, Medallia) was bought at 8–16× revenue on PIK-heavy debt. Several are now distressed; Medallia was handed to its lenders in 2026, wiping ~$5B of equity. The sophisticated take is leverage plus AI disruption, not "recurring revenue is safe."
Europe · compounders & cyber
No European large-cap public software champions existed to take private at $20B+, so the marquee deals are decade-long buy-and-build holds (Visma) or cybersecurity take-privates. Permira and EQT both now say that in B2B software, "no AI thesis = no deal."
Asia · carve-outs win on entry price
"Asia tech PE" is Japanese/Korean hardware & semi carve-outs plus Indian IT-services control deals; no US-style software take-private exists. Kioxia is the lesson: in cyclical hardware, a fire-sale entry and the staying power to hold through a down-cycle produced a ~20× win an AI boom no one underwrote.
The thread through all three
AI is the swing factor in every region: the disruption stalking the US software vintage (seat-based SaaS), the demand shock behind the Asia hardware win (Kioxia's NAND super-cycle), and the growth story now sold across Indian IT-services (Mphasis, Hexaware). The same force is the biggest risk in one region and the biggest tailwind in another. And the deals increasingly share a source in corporate carve-outs: SAP shed Qualtrics, Toshiba shed Kioxia, HP shed Mphasis, Micro Focus shed SUSE.