A Value Investor Lost in the Valley
What this covers
The conversation pairs host Patrick with allocator Chris Douvos, who looks at venture as a value investor would, from the seat of someone choosing fund managers rather than companies. He starts with institutional context: why private marks are stale and so calm an institution's apparent volatility, and why the lesson of Yale's venture success is "don't try this at home." The 1990s rush that followed, with about $100 billion raised in 2000, led to years of poor returns.
The middle of the discussion covers how startup economics changed. Cloud servers, open source, and the lean startup cut the cost of reaching revenue, illustrated by Josh Kopelman's three companies needing $7 million, $700,000, and $70,000. That opened a gap for micro VC funds, now about 550 firms, and Douvos argues the competition is pushing up prices. He applies Buffett's Equation, opportunity as value minus perception, to the TechCrunch effect and to IPOs pricing below their last private round. He also suggests consumers may capture much of the value startups create. Several positions are contested: that track record and fund size are poor guides, that 12 to 20 company portfolios outperform, and Bill Davidow's finding that generalists captured about 80 percent of returns in emerging spaces.
The last stretch is about people and incentives. Douvos discusses managers who prefer fees to carry, the principal-agent problem in institutions, and why standard deviation misses the technology, market, and team risks that matter. He also recounts David Salem's advice to invest without fear of being wrong and alone, and notes that a venture fund outlasts the average marriage by about two to one.
Douvos argues venture returns come from concentrated, process-driven, early-stage bets made by allocators with long horizons and high pain tolerance, while perception-driven capital inflows, fund-size growth, and hype erode the value that venture is supposed to create.
- Track record and size are lagging or negative indicators; process and early entry are leading ones
- Cheap company formation and abundant capital inflate price, so opportunity shrinks as perception grows
- Allocators should judge managers on people, strategy fit, and portfolio construction, and bet on long horizons
Crowded capital and perceived value inflate prices, compressing returns until fundamentals reassert themselves.
- Because startups stopped needing large rounds, a capital gap opened for traditional $400 million funds that could not write $500,000 checks, and this gap drove the rise of micro venture funds such as First Round, IA, True, Floodgate, Felicis, and SoftBank's early vehicles.
“if you're a $400 million fund, you can't write a $500K check. That led to the rise of these funds.”
- The real lesson of Yale's venture capital case is 'don't try this at home': after Yale's success, the 1990s rush into venture, with about $100 billion raised in 2000, produced poor returns and a six-to-eight-year depressed period.
“the lesson of the case is actually, don't try this at home. So many people then piled into venture capital in the late '90s, we saw 2000 was a $100 billion fundraising year, and returns really suffered.”
- Competitive intensity drives up the price paid for venture assets, and since the return on any asset is a function of the price paid to get in plus the capital it consumes, rising prices mechanically compress returns.
“It's really driving up price, and the return on any asset is a function of the price you pay to get in plus the capital it consumes. It's almost simple math.”
- There are now about 550 micro venture capital firms, and the resulting competitive intensity in every space is pushing up prices.
“Now, there are 550 micro VC firms.”
- Venture has barbelled into roughly 550 micro funds at the early stage and very large mega funds at the top, which creates funky pricing dynamics; mega funds keep raising because some large pools of capital need innovation exposure and cannot invest early.
“I think we had a barbelling of the ventures, 550 micro funds... And then on the high end, you've got these mega funds now.”
- Investors who fly in from far away for a few days each quarter, bringing outside money that is sophisticated but not attuned to local opacity, fund many marginal early-stage investors, creating what Chris calls chaos capital and inflated seed valuations.
“we see people fly in here from Dubuque and Dubai and spend three days each quarter... And then as a result, you've got chaos capital, especially in the early stage.”
- Perception can temporarily raise a company's value through a recursive 'TechCrunch effect', but because there must be a last buyer, either an acquirer or public markets who care about value, fundamental revenue and value ultimately decide outcomes.
“In the end, there has to be a last buyer for your shares, or next buyer for your shares.”
- Opportunity equals value minus perception, so when perception of a company rises, its opportunity shrinks; Chris calls this Buffett's equation while admitting he is not sure Buffett said it.
“But opportunity equals value minus perception. It's simple.”
- Several recent IPOs have priced below the last private round, a public down round, which shows that perception has grown so large that value is what it is perceived to be and opportunity has shrunk.
“we've seen a bunch of these IPOs go out, where the opening... is below the last round private market price.”
- Technologies tend to be over-hyped in the short term and under-hyped in the long term, so Chris prefers to invest after the hype cycle, when the trough of disappointment has cleared the field.
“I like investing post-hype, because I feel like there are a lot of technologies out there that are over-hyped in the short term and under-hyped in the long term”
Thoughtful process and concentrated selection, not fund size or track record, drive venture outperformance.
- On the margin, concentrated venture portfolios of roughly 12 to 20 companies should outperform, because each win has much more impact; diversification into 50 to 70 companies only helps if selection is random, as with darts at a dartboard, so thoughtful selection is what makes concentration produce structural alpha.
“if you can apply thoughtfulness to the problem and have a portfolio of 12 to 20, you could really make a structural alpha, I believe.”
- Venture fund size is a contra-indicator of success, because funds that show early perceived success have money thrown at them, and by the time a fund is proven good it is usually much larger, which dilutes the process that created the results.
“By the time you really know if a fund is good, it's usually a lot larger. I do think that size is a contra-indicator to success.”
- Track record is a lagging indicator rather than a leading one, so the better predictor of venture success is repeatable process: thoughtful methods for building, testing, and executing investment hypotheses.
“I actually think track record is a lagging indicator, not a leading indicator.”
- Nassim Nicholas Taleb said the American economy is the most open to optionality of any economy in the world, and Chris believes that in venture, a few surviving companies out of many failures can be transformative and make money for well-positioned investors while indexing venture wastes time.
“America's economy is the most open to optionality of any economy in the world, and I really believe that the essence of that is distilled here.”
Venture risk lies in deal concentration and team risk, so each position must be able to return the fund.
- Startups often require many things to go right in sequence, and the joint probability of all of those things happening is actually quite small, which is a source of stress for investors.
“the joint probability of all these things happening is actually pretty small. That's something that stresses me out.”
- Standard deviation is a poor proxy for venture capital risk; the relevant risks are technology, market, and team risk, such as a VP of sales whose behavior could wreck a company.
“Don't give me the standard deviation BS. How do you even begin to integrate this?”
- Venture portfolio construction should ensure that any single position, through its sizing and entry valuation, can return the whole fund, which Chris calls RTFE (return the fund equivalent), because venture returns are a power law where the top 10 deals out of about 3,000 matter most.
“any one of those companies can return the whole fund.”
- Strategy is not just what you do but also what you do not do and what you ignore, and most public managers pay little attention to constructing portfolios to create an advantage.
“because strategy is not just what you do. It's what you don't do and what you ignore.”
- Large venture funds can post weak results despite strong earlier funds, so an allocator must distinguish assets under management from alpha; a manager's $75 million and $150 million funds performed well, while the later multi-billion funds deteriorated sharply.
“they had a $1.2 billion and a $2.4 billion fund, and I later read in II... They just shat the bed.”
Early entry offers a better cost basis, but long horizons and principal-minded patience are needed for slow results.
- Early-stage venture investors are structurally advantaged over late-stage investors because they buy in at a better cost basis.
“Early-stage investors, I think, are structurally advantaged versus late-stage investors, just because they have a better cost basis.”
- Venture companies take roughly eight to 11 years from first investment to go public and about six years on average to reach an M&A exit.
“It takes eight to 11 years for companies to get public from first investment on average. It takes, I think, six years on average to get to an M&A exit.”
- The evaluation horizon in venture is often a multiple of the fundraising horizon: a manager may raise four funds before fund one shows results, which is longer than most attention spans and career tenures.
“You'll raise four funds, conceivably, before fund one starts really showing results.”
- Lean startup methods made it far cheaper to reach first revenue: Josh Kopelman's first startup needed $7 million, his second $700,000, and his third $70,000, because servers can be rented and open-source software integrated instead of built.
“Josh Kopelman at First Round always says his first startup took $7 million to get to first revenue, his second startup took $700K to get to first revenue, and then his third startup took $70K”
- High-net-worth investors with an entrepreneurial outlook, who think like principals, are well suited to venture, while investors who think like agents and chase short-term results are a problem for institutional investing.
“Think more principals than agents. That's where you get some of the problems in institutional investments, is if people think too much like agents and they want short-term results.”
- The average venture fund lasts about twice as long as the average American marriage, so allocators are effectively betting on people for a 20-year ride and typically need to back at least two funds from the same manager.
“the average venture fund lasts twice as long as the average American marriage. You're really making a bet on people, and you're probably betting on at least two funds”
- The biggest driver of venture success is a willingness to have a long horizon, because venture requires taking the most on faith for the longest time, and having high tolerance for risk, illiquidity, and horizon together is a rare trio.
“I would say the biggest driver of venture success is a willingness to have a long horizon.”
- A Harvard endowment official was penalized against a benchmark when management fees were drawn during a quiet venture year, creating a J-curve drag; the anecdote illustrates how short-term benchmark pressure pushes against the long horizon venture requires.
“they had a big management fee draw... it just created this huge J curve.”
- David Salem taught Chris to invest courageously without fear of being wrong and alone, because if you are afraid of being wrong and alone you will never be right and alone, and career risk should not enter the investor's mind.
“I want you to invest without the fear of being wrong and alone, because if you're afraid of being wrong and alone, you'll never be right and alone.”
Local proximity to founders, mentors, and campuses gives hard-to-replicate deal access and an edge.
- Berkeley is structurally underfunded by venture capitalists relative to its research base, which includes about $850 million in Department of Energy funding at Lawrence Berkeley Lab and $1.1 billion in science funding at Cal, while venture capitalists concentrate at Stanford; Chris would go long Berkeley and short Stanford as a pairs trade.
“There's $850 million in Department of Energy funding there, and another $1.1 billion in sciences funding at Cal, and yet, all the venture capitalists spend their time at Stanford.”
- Beyond money and momentum, the Bay Area has an edge in mentorship from people who know how to build companies and in entrepreneurial management, because taking a company from $10 million to $100 million in revenue, or blitzscaling, is a distinct and very hard skill from zero-to-10.
“Taking a company from $10 million to $100 million in revenue is an amazing challenge.”
- Silicon Valley is unique in history as a magnetic place, so other regions should build their own dynamism rather than try to become the next Silicon Valley; Chris hopes someone becomes the new Biocoast.
“I don't want anybody else to be the next Silicon Valley; I want somebody to be the new Biocoast, or whatever it'll end up being.”
- Money is a crucial factor in the Bay Area's magnetism, since capital moves quickly through the Electronic Herd, and the Bay Area combines this with mentorship and an ecosystem with momentum that New York and Los Angeles also have.
“Money is a really important thing... Money is crazy important.”
- In Palo Alto, almost no one within several blocks earns W2 income; they earn equity or consulting 1099 income, which creates a mentorship loop that keeps founders in the Bay Area, while a London exit typically sends founders to the South of France.
“I don't think anybody gets W2 income. They're all equity or consulting, 1099-based.”
- Students and PhD researchers at college campuses are producing sophisticated entrepreneurship, such as computational physics and computational biology software, and investing close to these sources, as in Berkeley's House Fund and MIT, offers differentiated access to deals.
“What we're seeing now is real authentic entrepreneurship among college students, and in very sophisticated ways.”
- Historically, venture investors have tried to pull intellectual property out of colleges and research institutions, but licensing offices and institutional complexity make this a thorny road with many broken plows.
“that's a tricky, thorny road, because you've got thick institution universes and licensing offices and this and that.”
Stale marks and misleading labels distort institutional accountability, while public markets ultimately price value.
- ICO fundraising changes accountability dynamics because companies raising large sums from coin sales face less pressure to answer to investors, which Chris finds unsettling; he hopes the market will find equilibrium.
“I think the whole ICO thing changes some of the accountability dynamics, and that's really unsettling to me.”
- Venture and private equity have a pacifying effect on institutional volatility because their values are stale, and Lerner and Gompers found that rolling forward quarterly marks produced a 0.65 correlation between the S&P 500 and itself, the same correlation commonly used for private equity.
“they got a .65 correlation of the S&P versus itself. By the way, .65 is what most people use for private equity correlation.”
- Endowments that churn through private equity leaders, such as one school that has gone through seven generations of PE managers, cause managers to become anxious with each new sheriff, which erodes their franchise, whereas persistent teams like Yale's keep institutional memory.
“every time there's a new sheriff in town, all the managers are anxious, and it starts to, I think, erode franchise.”
- The label hedge fund is misleading because many do not hedge, and the 'skill-based investor' label is absurd since bond managers are also skilled, so Princeton settled on the term idiosyncratic return for its hedge fund portfolio.
“Ultimately, we settle on idiosyncratic return, which I thought was actually pretty clever.”
- Every private company needs a last buyer, either an acquirer or public-market investors, and these buyers care about value rather than media mentions, which is why several recent IPOs priced below their last private round.
“there has to be a last buyer for your shares, or next buyer for your shares. And that's either an acquirer or the public markets”
Manager selection rests on judging people's edge, humility, and strategy fit, not on performance.
- Manager selection depends on whether the strategy resonates with the people running it, and Chris finds that there is often slippage between what managers say they do and what they actually do.
“What's important with respect to strategy is, is there a resonance between the strategy and the people? It's amazing to me how often there's a slippage between cup and lip there.”
- Each asset class speaks its own language: hedge fund managers cite Sortino and Treynor ratios, while venture managers justify investments by the team, so allocators must understand that venture judgments are about people and that this disconnect is poorly recognized.
“Speak Like the Locals... Why'd you invest in that fund? Oh, the team.”
- Chris's four-part manager selection process evaluates people first (whether they have an edge), then strategy (whether it resonates with the people), then the portfolio (the proof of the pudding), and performance last because it is a lagging, size-dependent indicator.
“Yeah. I'll start with my own manager selection process, which has four parts... The first is the people.”
- Allocators can assess venture managers' humility and learning by asking for their strongest held opinion and biggest professional disappointment, since a disappointment shows where their hypotheses were misproven.
“What's your strongest held opinion? What's the biggest hypothesis that you're testing?”
Venture firms must recast from passive capital sources into service providers whose claims need specific evidence.
- Crowdfunding and crypto unbundle capital from influence, since VCs historically provided both capital and value-add such as introductions to buyers and key hires, and crypto takes this further by letting founders raise large sums directly.
“I think crypto is that trend perhaps taken to the next level, because we see folks going... They can go out and they can raise, with coins... considerable capital.”
- Venture capital firms will have to recast themselves from capital sources into service providers that offer a suite of services to entrepreneurs, under pressure from crypto and unbundling, while ICOs weaken accountability.
“ultimately, venture capital firms will have to recast themselves as service providers.”
- A concerning trend is that experienced operators such as CFOs, VP sales, and heads of product, who used to recycle into new companies, are increasingly becoming venture capitalists, which could weaken the entrepreneurial management that makes the Bay Area strong, especially when their outcomes make them rich enough to stop working.
“A lot of those folks are becoming venture capitalists now.”
- Most venture capital is passive capital, and the idea that venture investors reliably provide catalytic value-add is somewhat a myth that David Swensen sold to endowments, though a few firms do it very well.
“I do think it's somewhat of a myth. There are firms that do it really well, but a lot that... The vast, vast majority are just passive capital.”
- First Round's signature investment in Uber seems to reflect being in the right place at the right time rather than its venture concierge services, since Travis Kalanick likely did not seek help from the concierge, so service-provider claims need hyper-specific evidence.
“Their signature investment, in some ways, is Uber... I don't think Travis ever asked a question of FRC's venture concierge.”
Emerging themes reward specialists and early footprints, while augmentation benefits will be unevenly distributed.
- Human augmentation through exoskeletons and robotics will deliver a net societal benefit, but since the future is not evenly distributed, some people will gain more augmentation opportunity than others, creating a challenging societal question.
“There're certain people that are going to have more opportunity than others to be augmented humans”
- Venture is becoming more esoteric, and understanding AI is much harder than understanding early computer hardware, so the barrier to entry is higher and specialist domain expertise matters more than Chris historically gave it credit for.
“I think the barrier to entry is much higher... that is what makes this specialist theme much more important than I historically have given it credit for.”
- Bill Davidow's analysis found that specialist first movers captured only about 20 percent of total returns in emerging spaces, while generalist firms that could get up to speed and bring business-building resources captured roughly 80 percent.
“those people captured all the early returns, but only 20% of the total returns, and 80% of the returns accrued to generalist firms”
- Because beta is cheap through passive investing, active management will evolve toward catalytic management, where managers bring resources and actively change companies, and public markets will look more like deep value plays with catalytic investment.
“I think that active management will look more like catalytic management. How can those active managers bring some resources to bear?”
- Data Collective's early focus on big data in 2011 put it in a catbird seat now that the market is flocking to AI, which is a cousin of big data, showing the value of building a deep footprint early in a theme.
“the footprint that they built in those early years really has put them in the catbird seat”
Student-driven and augmentation-focused startups look better than licensing university IP.
- Licensing IP from colleges and research institutions has historically been a thorny road with many broken plows, but Chris sees real, sophisticated student-driven entrepreneurship, such as PhD-led computational physics and biology software, as a better opportunity.
“What we're seeing now is real authentic entrepreneurship among college students, and in very sophisticated ways.”
- Robotic drywall finishing is augmenting rather than replacing workers in a trade where drywallers are on average seven years older than the average tradesperson and face silica dust regulations, and the robots still need people to supervise.
“Once you learn about demographics of drywallers... they're, on average, seven years older than the average tradesperson.”