Menlo Ventures' Ganesan on Anthropic Bet, Backing AI Startups

Menlo Ventures raised $3 billion for new AI investments, the largest fundraise in the firm's 50-year history. The deal underscores continued investor appetite for artificial intelligence and the shift in venture capital toward backing next-generation AI winners. The news is positive for the AI and venture capital ecosystem, though likely not a broad market-moving event.

Analysis

This is less a single-fund event than a signal that private AI capital formation is entering a second phase: the bottleneck is shifting from model training to distribution, infra efficiency, and application monetization. A large, dedicated pool of dry powder tends to compress diligence cycles and inflate pre-round pricing for anything with an AI label, but the incremental winner is not necessarily frontier model builders; it is the picks-and-shovels layer that can absorb capital at scale without needing perfect consumer adoption. The likely second-order effect is an even stronger capital advantage for the most brand-name private platforms, while smaller incumbents in software and services face a harsher “prove ROI now” standard.

The tradeable implication is that public-market AI exposure should be split between beneficiaries of persistent capex and names vulnerable to valuation crowding. The winners are compute, networking, datacenter power, and cooling ecosystems; those revenues can monetize faster than private AI software can. By contrast, overfunding at the venture layer can be a negative for adjacent public SaaS and IT services, because it subsidizes new entrants that attack existing workflows with lower pricing and faster product iteration.

The contrarian risk is that massive AI fundraises can become a late-cycle tell: abundant capital often leads to duplicate bets, inflated exit expectations, and slower markups over the next 12-24 months. If AI revenue conversion disappoints, the unwind will first show up in secondary/private valuations, then in public comps with the most stretched forward multiples. A smaller but real catalyst for reversal is a sustained slowdown in enterprise AI spend or evidence that inference economics remain too expensive to support broad deployment outside a handful of use cases.

Key Decisions for Investors

• Long AI infra beneficiaries over AI application hype: buy NVDA, ANET, and VRT on 2-6 week pullbacks; target 15-25% upside if capex re-accelerates, with 10-12% downside if enterprise spend pauses.
• Pair trade: long SMH / short IGV for the next 1-3 months; the fundraise reinforces near-term demand for semis and networking more than for high-multiple software, with a favorable relative-value skew if AI adoption stays capital-intensive.
• Buy medium-dated call spreads in VRT or ETN as a way to express datacenter power/cooling demand; risk is lower than outright AI software longs, with upside tied to continued buildout over 6-12 months.
• Avoid adding to the most crowded public AI software names with weak current monetization; use rallies to trim exposure in high-multiple SaaS baskets over the next quarter as private competition increases pricing pressure.
• For a contrarian hedge, consider shorting a basket of recently re-rated enterprise software names against a long basket of AI infrastructure equities; if private capital keeps flooding the space, the first earnings revisions are more likely to hit software margins than hardware demand.