The Twenty Minute VCJeff Wang: Sequoia Capital's $9BN Global Equities Fund on The Future for NVIDIA, Google & Meta|E1212
CHAPTERS
- 0:00 – 1:10
SCGE’s investing philosophy: themes first, companies second
Jeff opens with the core mental model that drives Sequoia Capital Global Equities (SCGE): picking the right “island” (theme) matters more than picking the perfect “ship” (stock). He frames how this perspective shows up in their views on Google’s threats and why NVIDIA can still be reasonably priced if AI momentum persists.
- •70% of research is on the theme; 30% on the company selection
- •“Pirate and treasure map” analogy: right theme determines outcomes
- •Early signals (“wiggles”) matter more than perfect forecasting
- •Google faces more threats than ever; NVIDIA valuation depends on continued growth
- 1:10 – 2:59
Joining SCGE: building a Sequoia-backed public equities startup
Jeff recounts the formation of SCGE and his decision to join early, before external fundraising. He explains the initial risks, the economic sacrifice, and why Sequoia’s ecosystem advantages in tech investing were compelling.
- •SCGE incubated in 2009; Jeff joined in 2010 pre-external launch
- •Started with ~$50M internal capital; scaled to ~$9B AUM
- •Portfolio roughly two-thirds public, one-third private (mostly Sequoia co-invests)
- •Joined despite ~70% pay cut and high failure risk of hedge fund launches
- 2:59 – 5:28
From shorts to longs: expressing disruption without leverage
Jeff explains SCGE’s distinctive approach to shorting: not fraud-hunting or pure valuation calls, but a way to strengthen a thematic stance. He also outlines why SCGE uses low leverage, preferring concentrated, high-conviction thematic exposure over balance-sheet amplification.
- •Shorts used to express the inverse of long thematic conviction
- •Uses private-market ecosystem insight to see disruption early (e.g., Starlink impacts)
- •Low leverage by design; tech already provides volatility and power-law upside
- •Goal is a specific viewpoint, not a low-volatility “Citadel-like” product
- 5:28 – 6:51
When long/short breaks: factor rotations and surviving drawdowns
The conversation turns to periods when the portfolio structure can suffer—especially when markets rotate from growth to value. Jeff emphasizes designing the business for endurance and aligning with long-term LPs who understand tech volatility.
- •SCGE tends to be long growth, short incumbents/value being disrupted
- •Factor rotations driven by rates/macro can be painful but often short-lived
- •Track record framing: performance judged over long periods, not quarters
- •LP alignment matters; many LPs overlap with Sequoia’s long-term base
- 6:51 – 8:42
Knowing when you’re wrong: dispassion, data science, and re-underwrites
Jeff outlines how SCGE decides whether adverse signals mean a thesis is broken or just noisy data. He highlights integrating a data science team into investment routines and institutionalizing quarterly re-underwrites, including devil’s-advocate reviews on contentious positions.
- •Dispassionate interpretation of evidence is essential
- •Public markets: same info for everyone (Reg FD) → need outside-in verification
- •Data science team embedded in weekly pipeline meetings
- •Quarterly portfolio re-underwrite; fresh partner underwrite for controversial names
- 8:42 – 11:10
Mistakes in practice: Shopify post-COVID modeling and the Twilio warning signs
Jeff shares two learning cases: Shopify, where post-COVID demand normalization was underestimated, and Twilio, where multiple small issues collectively signaled deeper competitive pressure. The lesson is to respect step-downs in trendlines and to treat clustered negatives as meaningful.
- •Shopify: modeled COVID acceleration as persistent; reality reverted toward pre-COVID trend
- •Holding out hope too long despite changing data is costly
- •Twilio: margin slippage, questionable M&A, exec departures—each explainable alone
- •Quarterly re-underwrites and devil’s advocacy help surface thesis decay sooner
- 11:10 – 14:31
2016 ‘re-founding’ crisis: earning SCGE’s right to exist
Jeff details the pivotal 2016 leadership transition when Sequoia considered shutting SCGE down after parting ways with the original PM. He describes crafting a Version 2.0 plan focused on Sequoia’s edge—growth tech and ecosystem synergy—and presenting it in Sequoia’s “hot seat.”
- •Original PM didn’t unlock Sequoia ecosystem advantages; drifted into non-tech bets
- •Sequoia considered shutting the business; Jeff had to propose a new plan
- •Team relocated temporarily for clarity; narrowed focus to growth tech + overlap + co-invests
- •Mike Moritz framing: less leap of faith than 2009 due to established team, systems, proof points
- 14:31 – 18:49
Why a dedicated public equities team (not a continuation fund)
Harry challenges whether SCGE is simply a continuation vehicle for Sequoia holdings. Jeff argues public-market investing is a separate discipline: it includes net-new names, active capital recycling, and the responsibility to redeploy proceeds into better opportunities.
- •Public investing requires daily buy/sell discipline and opportunity-cost decisions
- •Only ~1/3 of portfolio overlaps with Sequoia portfolio; many winners are net-new (e.g., Shopify)
- •Bias to hold great businesses; avoid over-trading small valuation differences
- •Trim decisions matter when pricing gets meaningfully ahead of fundamentals (e.g., 2022)
- 18:49 – 21:39
Crossover into privates: why it helps—and how much illiquidity is too much
Jeff explains why SCGE participates in late-stage private rounds: better company understanding, competitive context, and allocation flexibility across public and private opportunity sets. He also describes preferred steady-state private exposure and how SCGE sizes concentrated positions.
- •Private participation can deepen knowledge (example: Nubank across secondary, rounds, IPO, post-IPO)
- •Flex allocation: currently fewer attractive private opportunities → more focus on publics
- •Crossover is improving as ‘tourists’ exit and IPO markets reopen
- •Target private allocation ~20–25% (vs. ~1/3 currently); concentrated book of ~15–20 longs with top five ~35–40%
- 21:39 – 42:23
Will privates overtake publics? Continuation funds, IPO reopening, and ‘why go public’
They debate whether private markets can supplant public markets and the role of continuation funds amid limited liquidity. Jeff expects IPOs to reopen as companies repair financial profiles and regain predictability, and argues only exceptional companies can indefinitely avoid public markets via secondaries.
- •Public markets still dominate scale, liquidity, and branding value
- •Continuation funds may be cyclical, driven by IPO shutdown and liquidity needs
- •IPO readiness now requires cleaner growth/profit tradeoffs (e.g., ‘30+10’ vs. ‘60/-20’)
- •Only a small set of elite firms can stay private indefinitely; most still need public markets
- 42:23 – 46:58
AI and the Mag 7: data, distribution, and Google’s growing threat surface
Jeff explains why big platforms may sustain leadership in early AI: they own distribution and data, enabling rapid productization and monetization. He contrasts Meta’s ability to reprice via ad auctions with Google’s search intentionality—and flags that Google faces unprecedented competitive threats, including from Meta.
- •AI productization favors companies that own customer + data; no new distribution channel yet
- •Meta: immediate AI rollout and instant monetization via auction pricing and improved ROAS
- •Google: search is already ‘high-signal’; AI uplift may be harder and business model riskier
- •SCGE is neither long nor short Google, but sees more threats than ever; Meta AI as a major challenger
- 46:58 – 53:14
AI monetization: ARPU vs. experience, and the $600B capex-to-revenue gap
The conversation moves from AI excitement to measurable economics: whether AI boosts ARPU or mainly improves UX. Jeff argues monetization is already happening in ad-driven models (especially Meta) and will come to software more slowly due to contracting cycles, while acknowledging real risk that capex outruns application ROI.
- •AI can drive immediate incremental ARPU where pricing is dynamic (ads)
- •Estimate: AI already contributing substantial incremental EBIT at Meta via time spent + CPM lift
- •Software monetization slower: requires proving value and renegotiating contracts (e.g., price hikes like Canva)
- •Agrees there’s a ‘$600B problem’: application ROI must justify massive capex; model-training spend is the biggest sink
- 53:14 – 55:19
Why NVIDIA can still be ‘reasonable’: capex momentum and hedged semiconductor exposure
Jeff outlines how to think about NVIDIA’s valuation in the context of hyperscaler capex acceleration and uncertainty about 2026. He explains SCGE’s approach to AI hardware exposure: long the best-positioned winner (NVIDIA) while shorting structurally weaker adjacent businesses to hedge overinvestment risk.
- •NVIDIA valuation framed by forward earnings and the bigger question of 2026 demand
- •Hyperscaler capex growth is surprising; sustainability is uncertain but momentum is strong
- •SCGE approach: long NVIDIA for moat/position; short weaker hardware/server names with margin risk
- •Overinvestment risk exists, so pairing longs/shorts helps manage scenario uncertainty
- 55:19 – 56:30
Infrastructure & data center deployment: why ‘shovels’ may lack moats
Harry asks about investing in the ‘steel and servers’ layer—data center construction, power electronics, and deployment. Jeff explains why SCGE largely avoids it: high near-term demand attracts competition, and many of these businesses may not sustain strong moats over a multi-year holding period.
- •Data center deployment is booming, but durability of margins/moats is questionable
- •High growth draws competitors; many picks are ‘okay businesses’ not compounding franchises
- •SCGE prioritizes long-term ownership over tactical 1–2 year trades
- •Prefers application-layer businesses with a core franchise plus an AI call option
- 56:30 – 1:01:02
Geography and macro: Europe’s prize, deglobalization, China opacity, and India optimism
They discuss global allocation amid Europe’s challenges, the reality of deglobalization, and increased difficulty investing in China due to transparency and trust. Jeff remains optimistic about China’s entrepreneurial talent (especially companies that go global) and is cautiously bullish on India’s evolution toward more profitable, disciplined companies.
- •Europe matters, but the best European firms must win in the US and globally (Klarna example)
- •Deglobalization is already underway and affects investability, especially in China
- •China risk: reduced transparency; difficulty understanding internal decisions despite strong fundamentals (e.g., PDD)
- •India: strong consumer/financials; fewer standout tech firms yet, but improving discipline and profitability (e.g., Zomato)
- 1:01:02 – 1:15:07
Lessons from Sequoia leaders and rapid-fire: optionality, resilience, and life views
Jeff shares personal working lessons from Doug Leone (trust and commercial instinct), Mike Moritz (talent nose and ‘dream gene’), and Roelof Botha (culture and long-term excellence). In quick-fire, he covers decision-making, staying long-term through bad years, views on AI capex uniqueness, and broader personal concerns like family structure.
- •Doug: trust-building, ‘sniffer’ for what works, and ‘listen with Dumbo ears’
- •Mike: talent recognition, communication, and pushing beyond pure numbers/EBITDA to imagination
- •Roelof: culture, thoughtful team design, patience with excellence over time
- •Quick-fire themes: optionality is expensive; 2022 as the tough down year; concern about family structure; alternate career paths and storytelling
