Venture Capital

Sequoia operates as a true partnership where influence flows to expertise not hierarchy; co-stewards enable partners to move at founder speed; 12 years of data show contentious investments perform equally to consensus ones — only presence of conviction predicts outcomes.

Radical Ventures' framework: neo labs typically have unclear business models, product roadmaps, and technical direction at billion-dollar entry valuations, making risk/reward unfavorable; however, teams that are 'incapable of not being wildly successful' attacking infinite markets (scientific discovery) can generate Anthropic-scale returns, justifying selective conviction bets.

The Late Stage Management case reveals how complex SPV vehicles with triple/quadruple layers obscure misappropriation, as $528M scheme and SpaceX share misreporting show systemic due diligence gaps for retail-access private funds.

Airtable's clean 1x non-participating preferences allowed all investors to recover capital at $1.28B sale despite 90% valuation drop; early investors still profited, late-stage got money back — a 'good failure' model for AI-era overcapitalized startups.

The $50-100M seed fund model is broken — too large to write collaborative small checks, too small to lead $8-10M rounds — forcing funds to either stay tiny and collaborative or raise growth vehicles; Frankel's firm chooses discipline over AUM growth.

Coastal Ventures argues billion-dollar seed rounds in 'neolabs' produce aquihires where investors lose (Windsurf, Scale AI examples). Sustainable venture returns require 20%+ early ownership, concentrated bets, and proactive pro-rata decisions (3x when conviction high, 1/3x when cooling) — never default pro-rata. Rocket Lab 28% ownership at IPO validates model.

The surge of 'neo labs' raising massive seed rounds at billion-plus valuations generally offers poor risk/reward due to lack of product, business model, or technical clarity. However, funds should be selective for generational founding teams (Discovery Loop, Anthropic) where the team's track record and market scope make the entry valuation pencil out, accepting that most will fail but the next Anthropic will return the fund.

SoftBank's strategy of borrowing against its marked-up OpenAI stake to fund further investment creates a reflexivity risk where any valuation decline could trigger forced selling that further depresses the stake's value, a risky maneuver that OpenAI should not welcome.

Hodak advises founders to calculate the actual cost of running the next key experiment (not the whole company), then raise double that amount, noting some ideas are only fundable at $50M or $0 — not at intermediate levels that yield ambiguous results.

Enterprises are adopting prompt routers (Weave) and model-agnostic platforms (Replit) to cut token costs 50-80%, enforce model routing policies, and automate workflows, creating a new infrastructure layer between applications and foundation models.

US limited partners backing new European funds are importing American risk tolerance, forcing a generational shift in European venture culture toward higher conviction, earlier bets.

Data shows immigrant co-founders have twice the probability of building public companies but receive only 25% of VC funding, suggesting a structural mispricing of founder grit and conviction.

Compressed graduation timelines and massive outcome scenarios make traditional Series A ownership targets obsolete; firms now pursue extreme early ownership or late-stage concentration in proven outliers.

Fund strategy rejects neo-lab acqui-hire bets and default pro-rata. Instead: incubate hard tech (fusion, space), take 20-30% ownership early, then either 3x pro-rata or 1/3 pro-rata — never passive. Private market froth creates 'debt' requiring eventual liquidity; only early ownership in technical outliers returns funds.

Founders should calculate the actual cost to run the next critical experiment, raise 2x that amount, and push for profitability sooner to escape the 'money cancer' of perpetual fundraising cycles.

Altman observes that very few investors provide constant, proactive operational support, citing Josh Kushner/Thrive Capital as a singular exception, implying that high-touch venture partnership is a scarce and valuable differentiator.

Office of Strategic Capital and JP Morgan's $10B defense fund create a new capital stack where government de-risking makes critical mineral mining, actuator manufacturing, and PCB production venture-backable, bridging the valley of death for hard tech.

Strategic M&A combining IDE interfaces with autonomous agents (Cognition+Windsurf) and rumored large deals (Cursor) indicate category maturation where product integration and go-to-market scale drive consolidation.

The gap between Y Combinator and the next-best accelerator has expanded dramatically; early immersion in dense founder networks (especially in the Bay Area) creates unpredictable but high-value long-term connections that compound for decades.

Allocate CEO Samir Kaju and hosts highlight massive growth in private capital (3k to 30k managers, $1T to $17T) but warn of disconnect between private markups and exit valuations, predicting pullback within 2 years.

Venture capital is a micro game where success comes from investing in generational companies at reasonable markups; current billion-dollar seed rounds create intoxication and misaligned incentives, while disciplined 2x step-up financing keeps companies tethered to performance.

New co-stewards Alfred Lin and Pat Grady replaced Monday meetings with 7-day deal flow, empower partners to invest on founder timelines, and prioritize conviction over consensus — data shows contentious and consensus investments perform equally when conviction is present.

StarCloud rejected by 100+ VCs at $2M/10M post in 2023; 17 months later Benchmark leads $170M unicorn round. Catalyst: public SaaS stocks crashed on AI coding agent launches (Claude Code), convincing VCs software moats are gone. Hard tech now seen as only defensible frontier. YC demo day sentiment shifted 'fastest I've seen public markets affect demo day' — investors now demand hard tech exposure.

The seed landscape has exploded to ~2,000 funds, but returns remain power-law. Sindana wins by anchoring funds with structural sourcing advantages (Neo's scholar program, Zfellows' campus network, Nova's youth) rather than raw picking skill. Fund managers with hustle, live networks, and downstream capital access outperform; former operators excel at earliest stages, ex-bankers at late stages.

At least 75 Neolabs exist; 2/3 will return zero (acqui-hires). Current valuations based on team option value (acqui-hire floor) but next round requires $4B+ revenue trajectory to justify 10x; revenue concentration criticism is lazy (TSMC/Anduril counterexamples).

One of China's most prolific AI venture backers seeking additional capital, indicating continued investment momentum in Chinese foundation model development despite geopolitical tensions.

South Park Commons Fund 4 at $575M reflects shift from SaaS to nuclear, semiconductors, energy grid; Series B/C cap tables 25% less diluted vs 5 years ago due to smaller AI-leveraged teams and abundant capital.

Hedge fund lineage shows strategic split: Viking Global reduced AI exposure and avoided July selloff but missed prior gains, while Coatue and Tiger Global maintained concentrated AI bets accepting volatility for supercycle upside; Coatue also active in private AI company cap tables.

US term sheets tend to be simpler, longer-term, and founder-aligned (e.g., additional shares vesting over time), while European sheets are longer, more restrictive, and reflect a short-term fund-raising mindset. This cultural gap affects which founders stay in Europe and shapes the risk profile of the entire ecosystem.

Combining concentrated late-stage direct investments (Eternals) with emerging manager fund-of-funds (Tribes) creates a pipeline from seed to pre-IPO, de-risking pension allocation while capturing upside across the venture spectrum.

Top-tier VC firms aggressively pursue secondary allocations in leading AI private companies like Anthropic, with Sequoia and Green Oaks leading a $3.5B bid for Leopold's stake, though the fund ultimately retained the position.

The 17x growth in private market AUM vs public markets creates a structural allocation gap: endowments target 20-50% alternatives but $10T wealth advisor channel sits at 3%; Allocate and similar platforms will capture fees on this rebalancing, though Samir Kaju predicts a private market pullback within 2 years as 'gravity returns'.

Mega-funds must deploy billions into few legendary operators regardless of business logic (Boring Company 'crazier than Adams'); market continually enables this process but value evaporates if sentiment shifts — Twitter/X worth far less than $44B without Musk's x.ai roll-up bailout.

Billion-dollar seed rounds force premature scaling and culture of overspending; Anduril's disciplined 2x annual markups with metered capital proved more sustainable; most founders cannot responsibly manage war chests.

With infinite new companies forming yearly, top LPs like Sindana Capital prioritize fund managers with structural access to elite founders (NEO's scholar program, ZFellows campus presence, Nova's youth focus) because deal flow volume and early access compound more reliably than individual picking ability in a power-law market.

Levy identifies five stack layers (agent-first apps, horizontal agent platforms, agent infrastructure, post-training/model tuning, data/infra) where startups will capture value by translating raw model capability into domain-specific automation.

Valar Atomics ($1B/$6B) and Base Power ($1B/$13B) demonstrate investor appetite for capital-intensive energy infrastructure; strategic investors (Sequoia, Ribbit, Coatue) provide operational expertise beyond capital.

Repeat founders now create the majority of European unicorns with $2T in enterprise value, forming a self-reinforcing community that attracts global capital and mentorship networks.

Across Patrick's 70+ investments, all 7 companies reaching $100M+ revenue had immigrant co-founders. Broad data shows >50% of VC-backed IPOs have immigrant founders vs 25% of funding allocation, signaling a structural mispricing of founder grit and conviction.

US investors (e.g., Sequoia) provided n8n with simple, founder-aligned term sheets including additional equity grants, while European investors proposed complex governance-heavy terms with clawbacks. US VCs think long-term about company building; European VCs optimize for near-term fund metrics.

European pension funds hold massive dormant capital pools that are now being unlocked by regulatory changes and demand from retail savers, creating a structural tailwind for venture and growth investing in Europe similar to the US/Canada model where pensions provide 40%+ of VC LP capital.

Private-market opportunity is now 'never been better in the application layer' because frontier models advance too fast to replicate; value accrues to startups that solve deployment, partnerships, and trust in regulated verticals (healthcare, defense, sovereign) rather than building foundational models.

Park notes fundraising events now happen far faster than traditional 12-18 month cycles (Simile raised seed, Series A, and a $200M round within a year), advising founders to stay prepared as investor interest often leads company progress.

Y Combinator's brand creates massive investor inbound before demo day, allowing first-time founders with limited networks to schedule 80 investor meetings in a single week and create competitive fundraising dynamics.

Seedcamp's new Select Fund deploys at Series B+ into breakout portfolio companies across all vintage funds, complementing new lead investors and leveraging the firm's US team on both coasts to support transatlantic scaling.

Europe has the talent and education to build great tech companies but lacks the entrepreneurial culture of risk-taking, fast iteration, and comfort with failure that drives US innovation; adopting these cultural norms could accelerate European tech success.

Jack Zhang advocates hiring for first-principles thinking over industry experience, trusting founder intuition in recruitment, and firing mismatches in weeks not years — after costly mistakes hiring ex-bankers who imposed legacy thinking and took years to remove.

A cookie-cutter VC playbook — fund seat-based B2B SaaS addressing known pain points — created a 'sixth big tech' category with predictable but low-alpha outcomes; returns compressed as the model became too legible and imitable.

The founder notes European investors often panic at aggressive 5-year plans compressed to 2.5 years, while US investors view the same timeline as slow, highlighting a transatlantic gap in risk appetite for capital-intensive physical AI scaling that may drive future fundraising toward US partners.

Investors were hyped in early AI deals just like early SPACs; companies now doing the same thing as early winners will find it hard to raise. Over-capitalization before product-market fit distorts hiring, salary baselines, culture, and creates pressure to grow into unrealistic valuations.

Pre-money valuations for AI startups have jumped from ~€1.7M (2016) to €15M+ today; investors question moat durability when 'everyone adds AI to pitch decks,' and anticipate EU AI Act may invalidate entire product categories post-investment.

Sourcing tools are shifting from alpha to beta; VCs should focus capital and talent on evaluation not gathering; AI agents via MCP enable broader, deeper screening in natural language without technical expertise.

Case study: $3M raised via SAFEs at $20M cap, no shareholders' agreement, no vesting. Co-founders deadlocked; one demanded $500k to leave while keeping $2.5M cash. Delaware litigation will consume remaining cash. Lessons: SAFEs delay equity issuance → no legal shareholding → no drag-along/tag-along → investors unprotected. Always require priced round with vesting and SHA before scale.

Equity investors have become more risk-averse, requiring clear paths to profitability, which blurs the line with non-dilutive funding and expands the capital options available to founders beyond traditional venture capital.

Index's single integrated partnership across Tel Aviv, London, SF/NY enables real-time frontier access, agile deal teams, and founder trust, though it demands intense communication discipline.

European pensions allocate <0.05% AUM to venture vs US peers with materially higher returns; education is working but structural risk-aversion and subscale teams slow deployment — closing this gap is the highest-leverage lever for European GDP growth.

Median European VC experience neutral to negative; they cannot imagine trillion-dollar outcomes beyond domestic markets and prefer safe B2B SaaS over hard tech. US VCs at least engage with outlier possibilities. Since European VC LP capital largely comes from US anyway, founders should optimize for GP quality over geography.

European Investment Fund backs angels (€10k tickets) and large funds (EQT) alike; public money dominates European VC but hosts argue it's better than bailing out airlines, though fee/carry layers reduce efficiency vs direct sovereign investing.

One of venture's oldest firms raises a large early-stage fund to back the next generation of AI entrepreneurs, deploying a three-layer strategy (model, infrastructure, applications) with conviction that the defining companies of this wave have not yet started.

Vento's strategy of staying early-stage, building a global Italian founder diaspora network (160 portfolio companies, half outside Italy), and benchmarking against Kima Ventures' decade-long early-stage focus creates a structural advantage to capture power-law winners in a nascent ecosystem.

Early-stage AI investing splits into three buckets: reasonable pre-seed for new founders with traction, expensive middle for early traction, and eye-popping but risky valuations for repeat DeepMind-type teams where acquihire risk makes risk-reward unfavorable.

Traditional SaaS relied on high software production costs which are now collapsing. Best companies to start in 2026 look more applied/industrial (energy, turbines) vs. 2015. Procurement from small specialized vendors will replace internal builds at large corps, sustaining venture as an asset class.

USV's Fred Wilson/Brad Burnham partnership prioritizes intellectual engagement over founder charisma, with market sizing dismissed as backward-looking; this philosophy allowed Vanta's $10M TAM in 2018 to be funded, proving category creation expands markets rather than serving existing ones.

Early-stage European ecosystems in UK, Sweden, Nordics, Munich, and Zurich are seeing angel money return and better company building, though corporate incumbents lag in urgency compared to startups.

The 2025 Stripe cohort is the highest-performing ever: 57% non-US, 50% faster growth than 2024, double the companies hitting $10M ARR in three months, GitHub code pushes up 41% YoY, Atlas formations up 41%, and 20% of startups charging customers within 30 days vs 8% in 2020, signaling an AI-fueled inflection in company building velocity.

US multi-stage funds are winning European pre-seed deals through aggressive speed (flying to founders within hours), forcing European VCs like byFounders to differentiate through even faster decision-making and local cultural alignment that second-time founders increasingly value.

After 25 years and hundreds of DD processes, zero clean reports exist — yet investors regularly override material findings (tax compliance, GDPR liability, cap table defects) because they 'like the team/market,' leading to post-close disasters or renegotiations; pre-DD self-audits are essential.

The Rodriguez family office runs an evergreen venture fund targeting 12-15% IRR (2x every 5-6 years) via 45+ investments, 250-600k initial tickets, follow-on reserves, and strict diversification across vintage years, stages, sectors, and geography (mainly Catalonia).

After Spanish and European VCs only offered follow-on commitments contingent on a lead, the founder flew to SF, secured a lead within 24 hours via warm intros, and triggered a competitive oversubscribed round, demonstrating the importance of US lead investors for deep-tech AI startups.

In an AI-heavy funding environment, fintechs must demonstrate profitability and unit economics to raise; Brite's 2022 profitability decision enabled a successful 2023 Series A, and the company plans similar discipline ahead of its next round.

Workday's strategic VC succeeded by refusing board seats/exclusivity, focusing on product feedback loops and M&A optionality; Vilela replicated this as independent secondary investor, proving model scales beyond corporate balance sheets.

VC succession fails when founding partners retain majority carry; Credo's transition to equal partnership for the next generation (Maciek, Jakub, Max, Matěj) aligns risk, responsibility, and economics, which is rare but necessary for multi-decade firm continuity.

Right-sizing fund capital to the addressable deal flow in a constrained niche (CEE pre-seed) preserves the mathematical possibility of outlier 10-20x fund returns, whereas upsizing forces deployment into lower-quality opportunities and caps upside at 3-4x.

At pre-seed where bets are on people, the ability to reference-check founders through shared cultural context (high school, university, local reputation) creates a defensible moat that multi-stage US funds cannot replicate, even as they move earlier.

When VCs are already power users of a product, fundraising becomes more efficient because investors understand the product experience viscerally rather than through slide decks.

Earlybird's new structure ensures management company ownership stays with active partners, preventing senior partner entrenchment and enabling multi-generational legacy like McKinsey, with legal/tax mechanisms to enforce continuity.

Long-standing, sophisticated venture firms (Index, Balderton) are essential infrastructure for UK growth; government should lean into their expertise for AI policy and sovereign capability building.

Early liquidity for Spiegel and co-founder Bobby Murphy removed personal financial pressure, letting them evaluate Meta's offer on strategic/values grounds rather than survival needs — a structural advantage most founders lack.

Founders Fund's anti-thesis approach—investing purely in founder conviction and market fit across disparate industries (rockets, payments, short-term rentals, defense)—has generated the vast majority of returns from a handful of world-class operators, proving sector maps are irrelevant compared to backing the right person for the specific problem.

Raising mega rounds ($60M seed, $100M Series A) forces premature scaling, destroys negotiating leverage with candidates and customers, and creates valuation overhang that stalls momentum; normal rounds with patient markup cycles build healthier companies and better employee alignment via regular tender offers.

European deep tech check sizes and ambition levels remain far below US; while dedicated deep tech VCs have emerged in Europe, cross-Atlantic operations add constant management complexity for hardware companies requiring physical US presence.

Space launch is the 'anti-portfolio' for typical VCs: infrastructure-heavy, no product for years, state customers, decade horizons. Success requires tranche-based milestones, family offices with flexible mandates, and agencies like CDTI covering market failures. PLD's 55-line cap table reflects fragmented but persistent investor conviction.

The exit disclosure process forces founders to confront every operational mistake and warranty risk, creating a knowledge advantage that makes second-time founders better at avoiding catastrophic risks and building scalable companies.

Valuations have doubled in two years, forcing VCs to be more selective; West Coast consensus anoints category winners regardless of moat, while disciplined funds like Lerer Hippeau avoid foundation models to focus on domain-expert founders building second-gen AI applications.

University tech transfer offices created corporate venture vehicles to generate revenue, but these programs prioritize institutional interests over founder interests, resulting in biased company structures. Founder-first models like ETH Zurich and UT Austin attract talent by aligning with founders rather than capturing them.

Invest America/Trump accounts would create 401k for every child; if federal government adds sovereign wealth fund buying private equity, all shares could distribute to accounts — but Friedberg prefers free-market S&P 500 approach over government picking private winners.

The customer value fund provides non-dilutive capital to finance sales and marketing (CAC) that is repaid from customer LTV, aligning investor returns with company growth without equity dilution; General Catalyst is scaling this to $440M for Factorial.

After 2022-2024 recalibration, 2025 shows true stabilization: 34% of rounds are Series C, seed/Series A sizes growing sustainably, and 77% of deals now carry standard 1x non-participating liquidation preferences — signaling healthier founder-investor dynamics.

A rising share of M&A exits are strategic (not distressed) and involve US-based acquirers, creating a viable alternative to IPOs and reinforcing the US-UK/EU cross-border corridor as a key liquidity driver for European venture-backed companies.

Silicon Valley's norm of helping strangers without tracking reciprocity eliminates the conservation law of favors, allowing exponential growth of social capital that compounds across generations of founders.

Concentration of ambitious peers in a hub generates more and better serendipitous encounters, faster decision-making by investors, and competitive peer pressure that raises execution standards, making geographic clustering a self-reinforcing advantage.

Transitioning from informal angel investing to a professional VC fund (Oyon Ventures, €300M AUM) requires hiring investment professionals and standardizing processes; the venture builder model has yielded a SaaS company (Planet) with €1M ARR, but the founder's simultaneous management of real estate, hospitality, construction, and fund ventures risks dilution of focus.

The decision to raise venture capital irreversibly commits a company to the growth expectations of the VC model; without genuine ambition for massive scale, founders will underperform because every stakeholder (investors, employees) optimizes for growth. Ambition becomes a necessary condition, not a preference, once VC dollars are accepted.

Venture capital is seeing unprecedented markup velocity with companies doubling revenue quarterly; 8VC plans to hold its best defense and AI positions indefinitely rather than exit.

Small teams of former founders with high autonomy and talent density can execute faster than larger organizations, especially in AI where leverage amplifies individual output, justifying extreme hiring selectivity.

Both CFOs credit venture capital backgrounds (VC investing, short selling) for developing contrarian thinking, IC-style data-driven influence, and pattern recognition across portfolio companies that accelerates operating decisions.

The Wiz acquisition at 100x the founders' prior $300M exit establishes a new ceiling that raises ambition across the ecosystem, encouraging founders to target larger M&A or independent public company outcomes rather than anchoring to historical benchmarks.

Liaw observes the US-expansion threshold has dropped from ~$50M ARR to $10-15M; he advises sending a founding-team member to New York early to establish culture and hire US sales leadership, as US sales talent requires local vetting.

SeedRocket-affiliated VC vehicle run by former founders launches third fund to invest from early stage to multi-million revenue companies, differentiating via operator network rather than pure capital.

Seven VC-backed body-cam AI companies chase the same 17,000 buyers. Flock, Motorola, and Axon already own the sales channels. Langley predicts most will shut down or be acquired for pennies; the market cannot support more than a few integrated platforms.