Oct 21, 2026 · 1h 4m
Venture returns reward access more than startup-picking skill
E445: AngelList: Why Picking Winners Is the Wrong Way to Invest in Startups
The episode examines who gets exposure to venture’s rare winners, how investors preserve that exposure, and whether private-market access can broaden beyond the wealthy.
- 1Venture investing often hinges on access, reputation, and allocation rather than discovering mispriced startups.
- 2Power-law returns make follow-on ownership and carefully chosen diversification central to venture portfolio construction.
- 3USVC aims to widen ongoing venture access through an evergreen fund while accreditation rules remain contested.
Don't miss
Erik Syvertsen’s Anthropic investment experience sharpens the discussion of bubbles, taking profits, and the signaling risks of selling venture positions.
The brief
Erik Syvertsen and David Whiteman challenge the familiar venture story: returns may depend less on spotting hidden value than on reaching the best deals.
At seed stage, attractive startups can command expensive valuations and competitive fundraising processes, making reputation, access, and self-selection more decisive than secret information.
Because a small share of companies drives most returns, investors must balance broad exposure with follow-on capital and enough ownership in potential breakouts.
The cathedral-and-bazaar framework contrasts large firms’ durable brands and networks with diversified exposure across managers, while warning that blind indexing can miss who has a right to win.
The conversation turns personal with Erik Syvertsen’s Anthropic investment experience, then widens to the difficulty of timing bubbles, selling positions, and interpreting the signals that sales send.
USVC’s evergreen structure becomes the practical test: can AngelList’s network and manager data create broader, ongoing access to private-market wealth without erasing the risks?
What was said on this episode
34 statements · 16 positive · 11 negative · 1 mixed · 6 neutral
Agentic coding and distribution rails permanently accelerate startup growth signals
“that's a permanent trend because you have with agentic coding and the distribution rails that are in place today. The signal arrives so much faster”
Listen at 8:44
Successful startups can reach $10 billion valuations within 18 months
“valuations hit 10 billion kind of, you know, 18 months into a company's journey”
Listen at 9:06
Slow-growth startups led by misunderstood founders are now exceptions
“the slow burn companies where you have kind of a gifted, you know, misunderstood founder is really, you know, the exception of the rule today”
Listen at 9:12
Instinct raised another round at a $10 billion valuation shortly afterward
“a week later they were raising out a 10 billion dollars valuation”
Listen at 9:55
Higgsfield was discussing a valuation roughly double its prior $5.4 billion round
“their last round was you know 5.4 billion and now there's like they're in talks to do something double”
Listen at 10:42
Fundraising momentum reinforces further fundraising momentum
“momentum begets itself”
Listen at 11:44
Multi-tranche valuation rounds are a minority of financing rounds
“it's a minority of rounds”
Listen at 13:55
Bazaar investing requires understanding each manager’s competitive advantage
“you have to understand why this particular manager has a right to win”
Listen at 31:56
A fund of funds is often appropriate for LPs lacking manager-selection expertise
“for a lot of LPs, then the right answer is a fund of funds”
Listen at 33:02
A 15-company early-stage portfolio has relatively low odds of producing a 1,000x–10,000x breakout
“if they have a portfolio of 15 companies at the early stage, what are the odds that they invest in the next anthropic or open AI and they hit this thousand or ten thousand extra turn? They're still relatively low”
Listen at 34:10
Approximately 0.5% of companies become breakout investments
“0.5% of companies end up being breakouts”
Listen at 34:36
A 200-company portfolio has about a 65% chance of containing a breakout
“If you get something like 200 companies, you'll get, this is just basic probabilities math, right? You'll get something like 65% probability that one of those companies will be a breakout”
Listen at 34:40
A 500-company portfolio raises breakout probability to about 90%
“If you size it up to about 500, then your probability goes up to about 90%”
Listen at 34:52
The 2021 venture vintage was challenging for investors
“2021, right? Very challenging year”
Listen at 35:51
Investors cannot know when a market bubble will burst
“I don't think you could know when the bubble will burst”
Listen at 37:47
Precisely timing market bubbles is not humanly possible
“timing it, I don't even think it's humanly possible”
Listen at 38:28
Selling Anthropic Series C was a terrible investment decision in hindsight
“that was obviously a terrible decision”
Listen at 39:27
Systematic trimming can manage but cannot eliminate venture selling risk
“there is a way to manage that risk, but you certainly you can't eliminate it”
Listen at 40:16
Investors should consider partial sales rather than selling everything or nothing
“the rational thing is sell 10% or sell the basis or sell a 2x”
Listen at 40:43
Disciplined investing requires setting rules and following them systematically
“the difference, I think, between more kind of endowment style, disciplined investing is you're really running a program and you might, you know, you just set rules and follow those rules”
Listen at 41:11
Investors should remain invested to stay in the game
“rule number one is you have to stay in the game. Rule number two is don't break rule number one.”
Listen at 42:18
Selling venture positions carries substantial signaling risk
“there is huge signaling risk in selling”
Listen at 42:44
Venture managers may keep holding to preserve founder relationships
“they don't want to piss off founders and will just keep holding”
Listen at 43:01
Most retail investors currently lack access to private startups
“most retail investors don't have access to the private markets, to private startups”
Listen at 45:34
The Magnificent Seven generate roughly 40%–50% of S&P returns
“40 to 50% of the returns, the S&P coming from the MAG7 alone”
Listen at 46:11
OpenAI, Anthropic, and SpaceX represent about $5.2 trillion in value
“you have kind of 5.2 trillion in value coming from OpenAI Anthropic and SpaceX alone”
Listen at 46:59
Private-market access vehicles are rapidly evolving
“the market is very rapidly evolving there”
Listen at 47:59
Individual investors are unlikely to build institutional-quality startup portfolios alone
“it's very unlikely that an individual investor is going to have the time, the expertise, and the access to build a portfolio of companies”
Listen at 49:41
Available online venture access has not enabled easy institutional-grade portfolios
“It's still been very, very difficult to build a, you know, institutional grade portfolio”
Listen at 50:17
Roughly one million accredited investors participate in venture through access platforms
“let's say it's 10x that. That's still only about a million accredited investors participating in venture”
Listen at 51:00
About 10% of accredited investors have meaningful venture access
“to say 10% of accredited investors have meaningful access in this asset class today is probably about right”
Listen at 51:09
Private startup exposure will become necessary in portfolios within five years
“there will be a sea change over the next, you know, call it five years, where it becomes a necessary part of people's portfolios to have private startup exposure”
Listen at 52:02
Multiple private companies are valued above one trillion dollars
“you have multiple trillion dollar private companies”
Listen at 53:09
Average investors need responsible venture access or will miss substantial returns
“you're going to need kind of the venture capital asset class and and then have a responsible way for the average investor to include that in their portfolio or they're just going to be missing a huge part of the of the kind of returns”
Listen at 53:36
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.
Books & mentions
Some links are affiliate links — PodLume may earn a commission if you buy.
Weekly email
State of the Pod, every Monday
The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.