Orlando Bravo says some software names hit by AI deserve a valuation cut

Thoma Bravo, which manages over $183 billion across 77 companies, warned that AI will disrupt many public software firms and has contributed to sector pain (the IGV software ETF is ~28% below its September peak). Orlando Bravo said some valuation declines are "very warranted" while others are unjustified, signaling differentiated risk across the software universe. He also conceded Thoma Bravo overestimated growth when paying $6.4 billion for Medallia in 2021, calling it a mistake that led to overpayment.

Analysis

AI is accelerating reallocation of value inside software incumbents rather than across the entire tech sector: companies that sell differentiated data, orchestration layers, or embed AI into high-switching-cost workflows will see margin expansion, while point-solution vendors that rely on routine automation face rapid revenue churn and multiple compression. Expect this re-allocation to play out unevenly over 3–24 months as enterprise procurement cycles and product roadmaps determine who can convert pilot wins into recurring ARR.

Second-order winners include cloud infrastructure and observability providers that capture increased telemetry and compute spend; second-order losers include professional services and reseller channels that historically extracted implementation rents but will be squeezed by standardized agentic deployments. The private markets arbitrage window widens when public multiples contract: disciplined acquirers with operational playbooks can buy market share at lower entry EV/ARR and reprice via bundling AI-enabled services over 24–36 months.

Key tail risks are (1) a rapid drop in compute costs that commoditizes model differentiation, (2) open-source model adoption that lowers switching costs for buyers, and (3) regulatory pushback on agentic systems that delays enterprise rollouts — any of which could compress conversion rates over 6–18 months. Conversely, clear monetization metrics (ARR uplift, net retention expansion, AI-driven price per seat) disclosed over 2–4 quarters would reverse the current haircut quickly and create outsized upside for names already trading at depressed multiples.

The current market offers asymmetric opportunities to pair concentrated winners against broad software beta: pick operators with engineering-led moats and repeatable enterprise monetization, pressure-test them for durable SMAs (sustainable margin advantages), and size positions to survive a 12–24 month product cycle before expecting full re-rating.

Key Decisions for Investors

• Pair trade (12–18 months): Long MSFT (platform + enterprise AI governance) vs short IGV ETF — equal-dollar exposure sized 3–6% net. Rationale: capture platform re-rating if enterprises consolidate AI stacks; expected payoff 1.5–3x if MSFT outperforms IGV by 20–30%. Stop-loss: 12% on either leg.
• Long Snowflake (SNOW) 12–24 month exposure (cash or LEAPs) 2–4% portfolio: SNOW sits on differentiated data infrastructure that monetizes increased telemetry. Target double if ARR growth accelerates and gross margins expand; downside is -40% in a prolonged SaaS multiple contraction.
• Tactical short set (6–12 months): identify 3–5 niche, high-licensing SaaS micro-cap names with >30% revenue from routine automation and poor gross margins — establish small, concentrated short basket (total 1–2% portfolio) with 20–30% stop losses. Rationale: fastest path to multiple compression as features get commoditized.
• Opportunistic private-equity exposure (24–36 months): increase allocation to buyout vehicles or rotates to APOS-like vehicles that can acquire public software assets at distressed multiples — target IRR 15–25% if execution reduces churn and bundles AI services. Limit allocation to 2–5% due to execution and vintage timing risk.