Odds on Open

Ex-$6B Fund Manager Discusses Risk Management, Venture Capital, and AI Infrastructure

1h 10m Listen to the original Episode page
Risk ManagementVenture CapitalArtificial IntelligenceHealthcareMarket StructurePortfolio Construction

Summary

Speaker_0, former manager of a $6 billion platform at DB Advisors, argues that the greatest investment opportunities arise in less legible, unpopular areas rather than crowded AI infrastructure trades. He emphasizes the importance of temporal diversification and dynamic risk management to survive market shocks, contrasting modern multi-manager platforms with past failures like LTCM. The discussion extends to venture capital, where he advocates for 'temporal maximization' over 'volatility laundering' and prioritizes 'slugging percentage' over hit rate. Finally, he critiques current healthcare and education models, proposing corporate self-insurance and market-driven reforms as solutions to rising costs.

Key takeaways

  1. 1
    negative correlation can very quickly go away and in discontinuous markets. where essentially correlations all go to one. And we've seen these in any number of shocks that have occurred since I first ran DB Advisors. So it gave me an appreciation for having there be an aggregate amount of risk across all of your strategies that you're comfortable with, and not to rely too much on offsetting correlations. in times of market stress.

    Unidentified speaker (speaker_0) · 5:32

Episode sections

  1. 0:00 – 0:32

    AI Infrastructure and Market Legibility

    Speaker_0 notes that the investment community focuses heavily on fundamental AI infrastructure, creating opportunities in less legible, unpopular areas. He suggests his greatest opportunities arise when markets are not in favor or ignored.

  2. 0:32 – 2:48

    DB Advisors Platform Structure and Risk Management

    Speaker_0 describes running DB Advisors, a $6 billion platform with ~20 teams across stat arb, HFT, CTA, and special situations. Risk allocation was dynamic, based on holistic platform risk and macro views, allowing offsets between strategies like special situations and stat arb.

  3. 2:48 – 5:14

    Evolution of Risk Management Technology

    Speaker_0 compares the 2002-2005 era to modern multi-manager platforms, noting that while structure resembles Millennium (central data/tech with autonomous teams), modern AI and data availability allow for dynamic, real-time risk micro-adjustments that were impossible then.

  4. 5:14 – 8:21

    Correlations and Time Horizon Diversification

    Speaker_0 explains that while offsetting correlations work in normal times, they disappear during stress, causing correlations to go to one. He advocates for temporal diversification (mixing short-term liquid strategies with long-term lockups) to survive shocks, citing LTCM's failure due to lack of this mix.

  5. 8:21 – 10:18

    Position Sizing in High Volatility Environments

    Addressing high semiconductor volatility, Speaker_0 advises smaller position sizes to avoid margin calls and longer lockups to allow thesis realization despite short-term tactical wrongness. He compares this to poker, requiring a bankroll to survive variance.

  6. 10:18 – 15:06

    Small vs. Large Managers in AI Era

    Speaker_0 argues small, nimble managers can still generate alpha, particularly in volatility trading and sentiment-driven strategies. He suggests successful small managers will resemble high-Sharpe volatility market makers that cannot scale, while larger players dominate through structural advantages.

  7. 15:06 – 17:16

    Venture Capital as Illiquid Complement

    Speaker_0 transitions to VC, noting his appreciation for illiquidity used to advantage in special situations. He views venture as a fit for his background, leveraging long-dated lockup money to match the payout function of early startups, similar to holding illiquid assets for extended periods.

  8. 17:16 – 34:06

    Cliff Asness's 'Volatility Laundering' Critique

    Speaker_0 addresses Cliff Asness's claim that VCs launder volatility. He reframes this as 'temporal maximization,' where VCs hold companies for 7-10 years, allowing volatility to wash out and compounding returns to materialize before exiting to public markets, passing risk to new investors.

  9. 34:06 – 39:32

    Decomposing Venture Returns and Process Errors

    Speaker_1 asks how venture investors decompose returns given smaller sample sizes than hedge funds. Speaker_0 explains that process errors fall into security selection and pricing, noting that while quant methods exist for founder selection, founder-market fit remains subjective. He identifies the pain of throwing good money after bad as a key error to avoid.

  10. 39:32 – 43:18

    Traits of Winning Venture Investments

    Speaker_0 attributes his venture wins to three factors: contrarian ideas, deeply domain-knowledgeable founders, and founders who are 'mad' or have a chip on their shoulder from past failures. He contrasts this with hedge fund investing, where short time horizons favor positioning over long-term vision.

  11. 43:18 – 53:01

    Hedge Fund vs. Venture Mindsets and Wealth Creation

    Speaker_0 argues that hedge fund traits like impatience and lack of vision are detrimental in early-stage venture. He emphasizes that venture requires letting winners run for 'slugging percentage' rather than optimizing for batting average. He states his objective function is impact optimization, not wealth maximization.

  12. 53:01 – 57:33

    AI Infrastructure Interconnectedness and Volatility

    Speaker_1 describes the pervasive prosperity in New York driven by AI. Speaker_0 agrees AI will disrupt industries but warns that the interconnectedness of the AI infrastructure trade (chips, data centers, credit) creates volatility risks akin to the 2007-08 financial crisis rather than the dot-com bubble.

  13. 57:33 – 1:05:05

    Capital Starvation and Career Advice for Youth

    Speaker_0 believes the AI trade will dominate capital allocation for years, leading to 'capital starvation' in other sectors, which he views as an opportunity. He advises young people to build competitive moats in human interaction and branding, as technical skills are at risk of commoditization by AI.

  14. 1:05:05 – 1:09:59

    Role of Government and Social Contract in AI Abundance

    Speaker_0 argues that policymakers' current approach to managing the wealth created by AI is incorrect. He suggests the social contract should involve incentivizing companies to reinvest in education, retooling, and local economies rather than just taxing wealth.

Claims worth verifying

DB Advisors was a $6 billion multi-manager platform with ~20 trading teams across statistical arbitrage, HFT, CTA, and special situations.

Unidentified speaker (speaker_0) ·0:37

How to verify: SEC Form PF filings for Deutsche Bank or historical press releases from DB Advisors regarding AUM and team structure. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Primary source

DB Advisors ran $4.5-4.5 billion of proprietary money and $2.5 billion of third-party money.

Unidentified speaker (speaker_0) ·0:40

How to verify: Deutsche Bank annual reports or DB Advisors investor presentations from the 2026-2026 period. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Correlations go to one in discontinuous markets/stress.

Unidentified speaker (speaker_0) ·5:32

How to verify: Historical financial data analysis of asset correlations during 2026, 2026, or other market shocks. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

LTCM lost $6 billion.

Unidentified speaker (speaker_0) ·7:59

How to verify: SEC filings or historical records of LTCM's liquidation and losses. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Primary source

Intel stock was down 30% in a day.

Unidentified speaker (speaker_1) ·6:16

How to verify: Historical stock price data for Intel (INTC) around the time of the podcast (Nov 2026) or specific news reports of a 30% single-day drop. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Cliff Asness claims VCs are 'laundering volatility'.

Unidentified speaker (UNKNOWN) ·17:16

How to verify: Cliff Asness's public writings, interviews, or AQR Capital Management publications discussing venture capital and volatility. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Game Changers Ventures has 16 core investments and targets 25 shots on goal per fund.

Unidentified speaker (speaker_0) ·27:18

How to verify: Game Changers Ventures fund one-pagers, LP reports, or public disclosures of portfolio size. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

David Tisch of Box Group backed Cursor with a $750k check that grew to a billion.

Unidentified speaker (UNKNOWN) ·26:40

How to verify: Crunchbase or PitchBook data on Cursor's funding rounds and Box Group's investment history. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Onyx Spot Return Crude Oil ETC tracks dated Brent crude.

Unidentified speaker (UNKNOWN) ·4:45

How to verify: Onyx ETC prospectus or fact sheet confirming the underlying index and commodity exposure. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

78% of world's oil deals are based on dated Brent.

Unidentified speaker (UNKNOWN) ·4:45

How to verify: Industry reports from oil trading houses or market analysis on Brent crude pricing benchmarks. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Early firms scraped LinkedIn to identify the best talent when they left places like SignalFire.

Unidentified speaker (speaker_0) ·36:33

How to verify: Search for SignalFire's historical investment thesis and public statements regarding their use of LinkedIn data for talent identification. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Most venture money is lost on the second, third, and fourth checks, not the first.

Unidentified speaker (speaker_0) ·38:06

How to verify: Review data from Preqin or NVCA reports on venture capital follow-on investment performance and write-off rates. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Cliff Asness coined the term 'volatility laundering'.

Unidentified speaker (UNKNOWN) ·54:23

How to verify: Search for Cliff Asness's publications or interviews where he discusses 'volatility laundering' in the context of hedge fund returns. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

The current AI infrastructure trade is more akin to 2007-08 than the dot-com bubble due to interconnectedness.

Unidentified speaker (UNKNOWN) ·56:05

How to verify: Compare the leverage structures and credit exposures in the current AI capex cycle to historical financial crisis reports (e.g., Financial Crisis Inquiry Commission) and dot-com era valuations. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Speaker_0 ran DB Advisors, Deutsche Bank’s $6 billion multi-manager hedge fund platform.

Unidentified speaker (speaker_1) ·34:06

How to verify: Check Deutsche Bank's press releases or SEC filings from the period Roger Ehrenberg was managing DB Advisors. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Primary source

SignalFire used quantitative methods for security selection by scraping LinkedIn.

Unidentified speaker (speaker_0) ·36:33

How to verify: Review SignalFire's website or case studies detailing their data-driven founder identification process. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Warren Buffett would have been a great early-stage venture capitalist if not for his focus on Ben Graham's margin of safety.

Unidentified speaker (speaker_0) ·44:23

How to verify: This is a subjective opinion/analogy; verify by checking Buffett's public statements on venture capital or early-stage investing to see if he agrees or disagrees with this characterization. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

The AI trade will cause capital starvation in sectors outside of AI for many years.

Unidentified speaker (speaker_0) ·58:29

How to verify: Monitor capital allocation trends in venture capital and private equity, tracking the percentage of new funds raised for AI versus other sectors over the next 3-5 years. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Speaker_0's objective function is impact optimization, not wealth maximization.

Unidentified speaker (speaker_0) ·51:01

How to verify: Review Game Changers Ventures' mission statement, LP communications, or public interviews where Speaker_0 discusses his career motivations. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Cliff Asness's 'volatility laundering' critique applies to the podcast business's beta to AI.

Unidentified speaker (UNKNOWN) ·54:23

How to verify: This is a metaphorical use of the term by the speaker; verify by listening to the specific context of how he applied the term to podcast advertising revenue sensitivity. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

The speaker believes policymakers' current thinking on adapting to AI-created abundance is incorrect.

Unidentified speaker (speaker_0) ·1:05:13

How to verify: Review public statements, op-eds, or policy proposals by Roger Ehrenberg between 2026-08-12 and 2026-11-12 regarding AI and government policy. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Education and healthcare are notable examples of sectors where large government subsidies, combined with constrained supply, lead to inexorable price increases.

Unidentified speaker (speaker_0) ·1:06:17

How to verify: Check economic literature or government reports on education and healthcare cost drivers during the verification period (2026-08-12 to 2026-11-12) for discussions on subsidy impacts and supply constraints. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

The healthcare industry spends more on acute and end-of-life care than on prevention, which the speaker describes as the inverse of how it should be.

Unidentified speaker (speaker_0) ·1:07:24

How to verify: Consult healthcare expenditure data from CMS, OECD, or industry reports published between 2026-08-12 and 2026-11-12 comparing acute/end-of-life care spending to prevention spending. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Walmart, Anthropic, and Costco are examples of companies that rationally self-insure.

Unidentified speaker (speaker_0) ·1:07:39

How to verify: Verify the self-insurance status of Walmart and Costco through their annual SEC filings (10-K) or corporate investor pages. Verify Anthropic's self-insurance status through press releases or corporate announcements in the verification period. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Primary source

Large companies that self-insure invest most money into prevention and provide acute services at a fraction of societal pricing.

Unidentified speaker (speaker_0) ·1:07:48

How to verify: Analyze internal corporate communications or case studies from Walmart, Costco, or Anthropic published between 2026-08-12 and 2026-11-12 detailing their healthcare spending allocation between prevention and acute care. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Do your own research

  • Verify the structure and team composition of DB Advisors by reviewing public filings or company press releases.
  • Check historical correlation data for AI infrastructure stocks (e.g., NVDA, AMD) during stress periods to assess the claim of correlation convergence.
  • Research LTCM's portfolio composition and risk management failures to validate the lesson on temporal diversification.
  • Analyze the performance metrics of venture firms that prioritize 'slugging percentage' versus those focused on hit rates.
  • Review healthcare cost data for companies like Walmart, Costco, and Anthropic to verify claims about self-insurance savings.
  • Investigate government subsidies in education and healthcare to assess the claim of constrained supply and inelastic demand.
  • Look for counter-arguments regarding the scalability of AI infrastructure investments and potential regulatory risks.
  • Check for any conflicts of interest or affiliations between Speaker_0 and the companies mentioned (e.g., IA Ventures, DB Advisors).
Full transcriptexpand
0:00
Unidentified speaker (speaker_0): Because so much focus is on the fundamental AI infrastructure trade, much of the investment community doesn't care about anything else. Those are the times in my life as a market participant, be it on Wall Street or in the venture space, that I have captured the greatest opportunity is when things are
0:20
Unidentified speaker (speaker_0): not legible, when things are not in favor, and kind of when no one cares.
0:27
Unidentified speaker (speaker_1): Roger, thank you so much for coming on the pod.
0:30
Unidentified speaker (UNKNOWN): Thanks
0:30
Unidentified speaker (speaker_0): so much, Ethan. It's great to be here.
0:32
Unidentified speaker (speaker_1): What does running a $6 billion hedge fund look like from the CEO seat?
0:37
Unidentified speaker (speaker_0): Stressful.
0:40
Unidentified speaker (speaker_0): The way that DB Advisors was organized, I was basically sitting on top of this business that was running four $4.5 and a half billion billion of of proprietary proprietary money money and and two, two and a half billion of third-party money run in these peri pursu managed accounts. And I had about 20
1:00
Unidentified speaker (speaker_0): different trading teams, ranging from special situations to statistical arbitrage,
1:09
Unidentified speaker (speaker_0): l literally and CTA. I mean, everything with a quantitative bend except special sits. was to kind of provide some added convexity to the portfolio. So it was really this interesting platform for looking at single manager risk, the amount of risk we wanted to take per strategy, but then looking at it on a blended basis since
1:29
Unidentified speaker (speaker_0): there were a lot of offsetting correlations in their books.
1:31
Unidentified speaker (UNKNOWN): How
1:32
Unidentified speaker (speaker_1): would you decide when to size someone up and size someone down?
1:35
Unidentified speaker (speaker_0): So when somebody came on the platform, we would give them a deal with a set amount of basically back then kind of value risk. And we were constantly measuring exposure. And as somebody performed, we would increase their thresholds. Then it was much more of
1:55
Unidentified speaker (speaker_0): an asset allocation decision. How much a manager would cap out would be a function of how we holistically looked at the risk across the platform. So it might be that because of our
2:08
Unidentified speaker (speaker_0): scale of our special sits, that would enable us to crank up our high-frequency stat ARB program more. So it was dynamic based upon our both macro view of the markets as well as the aggregate risk of the platform itself.
2:25
Unidentified speaker (UNKNOWN): How
2:25
Unidentified speaker (speaker_1): would you compare the sophistication of risk management back then to the way the pods, the multi managers are structured today? Obviously, a lot of these guys have had to learn from periods like two thousand and eight. Um
2:40
Unidentified speaker (UNKNOWN): but
2:40
Unidentified speaker (speaker_1): I'm just curious to hear about how the business worked back then versus how you think it works today.
2:48
Unidentified speaker (speaker_0): So
2:49
Unidentified speaker (UNKNOWN): the
2:49
Unidentified speaker (speaker_0): business is definitely has
2:51
Unidentified speaker (speaker_0): evolved a lot because of technology. So the structure of DB Advisors,
2:56
Unidentified speaker (UNKNOWN): in
2:57
Unidentified speaker (speaker_0): many ways, looks like a millennium, where you've got this central core that provides data services and technology to the teams. The teams do have a lot of autonomy subject to these risk limits. Within those risk limits and these concentration limits, they can invest as they see fit.
3:16
Unidentified speaker (speaker_0): Again, back in the 2002 to 2005 period when I ran DB Advisors,
3:23
Unidentified speaker (speaker_0): you simply didn't have the
3:29
Unidentified speaker (speaker_0): ability to microadjust as quickly simply didn't exist. It was like more like using a butter knife as opposed to, you know, uh a razor's edge sword. And over time, now with forget about like m we were using machine learning algorithms in the early 2000s and spending enormous amounts
3:49
Unidentified speaker (speaker_0): on on data. Right now, with AI, so much more data being available at so much lower prices, and a lot of these platforms internalizing their their data. So they have their own golden records of data.
4:03
Unidentified speaker (speaker_0): You're able to dynamically manage risk and develop new strategies much more quickly than you could back in the era kind of the on the early multi-manager platforms.
4:14
Unidentified speaker (speaker_2): Most oil ETFs don't actually track the price of oil. They hold futures contracts. And many of them track oil for delivery up to four months from now.
4:22
Unidentified speaker (UNKNOWN): So
4:23
Unidentified speaker (speaker_2): when you buy an oil ETF, you're not betting on today's oil prices. You're betting on where the market thinks oil will be next season. That's not the kind of trade you meant to make. Which brings me to today's sponsor, Onyx. They're the world's biggest market maker in oil derivatives, and they built the solution to this problem: the Onyx Spot
4:42
Unidentified speaker (speaker_2): Return Crude Oil ETC.
4:45
Unidentified speaker (UNKNOWN): It
4:45
Unidentified speaker (speaker_2): tracks dated Brent, the price 78% of the world's oil deals are based on. And it reflects barrels shipping in the next two to three weeks. That exposure used to be reserved for oil majors and trading houses. Now you can buy it on interactive brokers. The ticker is OIL.
5:02
Unidentified speaker (speaker_1): How did running that multi-manager platform inform or teach you about the way you view risk?
5:11
Unidentified speaker (speaker_1): Allocating capital, etc. et cetera.
5:14
Unidentified speaker (speaker_0): Managing a scale multi-manager platform
5:18
Unidentified speaker (speaker_0): gave me the insight
5:22
Unidentified speaker (speaker_0): about
5:23
Unidentified speaker (speaker_0): A, in normal times, offsetting correlations.
5:28
Unidentified speaker (speaker_0): But what it also taught me is that
5:32
Unidentified speaker (speaker_0): negative correlation can very quickly go away and in discontinuous markets.
5:39
Unidentified speaker (speaker_0): where essentially correlations all go to one. And we've seen these in any number of shocks that have occurred since I first ran DB Advisors. So it gave me an appreciation for having there be an aggregate amount of risk across all of your strategies that you're comfortable with, and not to
5:59
Unidentified speaker (speaker_0): rely too much on offsetting correlations. in times of market stress. So being extremely robust in how you model the tails, because everything looks good in the middle of the curve. Once you get out to the edges, that's where things get scary.
6:16
Unidentified speaker (speaker_1): Once you get out to the edges, that's where things get scary. Fully agree. We saw that last week, obviously. Intel down 30% thirty percent in a day. Um
6:27
Unidentified speaker (UNKNOWN): if
6:28
Unidentified speaker (speaker_1): you were running a multi-manager business today, how can one actually protect themselves? Because it seems like all the hottest trades are unbelievably crowded um
6:40
Unidentified speaker (speaker_1): managers appear uncorrelated on the surface but end up being correlated. How would you think to sort of
6:47
Unidentified speaker (UNKNOWN): not
6:47
Unidentified speaker (speaker_1): even solve that problem, but how would you think about that problem?
6:51
Unidentified speaker (speaker_0): So
6:53
Unidentified speaker (UNKNOWN): when
6:53
Unidentified speaker (speaker_0): I think about constructing kind of the optimal asset management platform,
6:59
Unidentified speaker (speaker_0): it's not just strategic diversification. Because again, if you just look at that single dimension, correlations skyrocket in times of stress. The other dimension is temporal. And to me, having a spectrum of time horizons to go along with strategic
7:19
Unidentified speaker (speaker_0): diversification, it was what enables you to both survive in the short term, but to also profit in the long term.
7:29
Unidentified speaker (speaker_0): The ability to hold for extended periods and having lockup regimes that match those durations of those strategies.
7:40
Unidentified speaker (speaker_0): is the right kind of offset to the highly liquid, but at times highly correlated strategies under stress that during normal time seem uncorrelated. Essentially, what I'm talking about is if you look at LTCM, LTCM was extremely short Short-term term time horizons, seemingly a
7:59
Unidentified speaker (speaker_0): globally diversified and asset type diversity. diversified portfolio, then in stress, correlation skyrocketed and they were they were down six billion in the blink of an eye. If they had a series of strategies with different time horizons, there wouldn't have been everybody rushing for the excess at the same time and having the entire
8:19
Unidentified speaker (speaker_0): franchise blow up.
8:21
Unidentified speaker (speaker_1): How would you size a book today, given that names today are much, much, much, much more volatile? than they were previously? And I'll give an example. Um I haven't run the numbers myself, but if you look at the semis names, the vol with which they move today, even though their market caps are much bigger, are i i i it's it's
8:41
Unidentified speaker (speaker_1): just unreal. And so how do you think that should how do I you think that should inform the way a PM or trader should construct their portfolio?
8:50
Unidentified speaker (UNKNOWN): So when
8:50
Unidentified speaker (speaker_0): thinking about optimal portfolio construction in segments where volatility is extremely high and way higher than it's been historically, really there are the two approaches. One is simply position sizing. And
9:06
Unidentified speaker (speaker_0): in your models, just having kind of trying to capture the benefit of the vol and still being kind of asymmetrically negative. the VOL if you're overweight, and then all of a sudden you've a downdraft and you get a margin call. Obviously position sizing helps with that. I think the other thing that it'll be
9:26
Unidentified speaker (speaker_0): interesting to see what young managers and spin-outs from legacy managers do with this. I think having longer lockups, given the fact that you might be thesis right. and tactically wrong in the short term, you need to let some of this value play out over
9:46
Unidentified speaker (speaker_0): time, even though you've got this tremendous level of volatility. It's very similar to playing poker, where you need a big enough bankroll to live for another day and you want to put in your chips when you feel like it you have asymmetric information.
10:03
Unidentified speaker (speaker_0): But in the markets, when you're posting margin, you can be right, but if you can't hold it, you're toast, which is why I think through smaller position size and the use of longer lockups can help protect against that scenario.
10:18
Unidentified speaker (speaker_1): On the topic of background size, I like ac asking this question to many of my guests. Do you think that smaller managers are just going are
10:31
Unidentified speaker (speaker_1): riding tailwinds associated with AI because it's inherently deflationary, because costs will go down, because they can um, you know, th they have an access essentially to a team of call it, you know, an infinite number of analysts, or do you think that
10:49
Unidentified speaker (UNKNOWN): the
10:49
Unidentified speaker (speaker_1): larger players will it will be able to use their structural motes to um call it squeeze all the juice out of all this new technology and attract all the best talent and gobble up all the capital? I'd love your take.
11:04
Unidentified speaker (speaker_0): I think
11:06
Unidentified speaker (speaker_0): that there will always be room for small, nimble managers to make good money.
11:14
Unidentified speaker (UNKNOWN): In
11:15
Unidentified speaker (speaker_0): my current career as a venture capitalist, I feel similarly. I think you could make the same argument in the illic in the illiquid markets, right? If you look at the Andreasons and the Light speeds and the Menlo's and people like that that have raised massive funds. And then you can look at the the Millennium and the Citadels and too many firms to mention that have both the
11:34
Unidentified speaker (speaker_0): technology, the people and the capital to seemingly dominate, there will always be opportunities for alpha with small players. And I would tend to think that right now, where small players can make a lot of money is on momentum and sentiment. Because right now, you pointed out how
11:54
Unidentified speaker (speaker_0): volatile these stocks are.
11:57
Unidentified speaker (UNKNOWN): If
11:57
Unidentified speaker (speaker_0): you're really good at trading vol right now and you have a relatively small amount of capital, you can still make a ton. of money. So I think it's just going to be incumbent upon these small managers to be extremely crisp on what their edge is, be the absolute best at it. And I think they'll be able to develop
12:15
Unidentified speaker (UNKNOWN): and
12:15
Unidentified speaker (speaker_0): deliver way higher sharps than the larger players.
12:18
Unidentified speaker (speaker_1): You mentioned Vol and how these if you're a small manager and you have a very crisp understanding of what your edge is, and you trade things. like that, you can make a ton of money. If you were a small manager today, how would
12:31
Unidentified speaker (UNKNOWN): you
12:32
Unidentified speaker (speaker_1): garner the right type of capital to actually capitalize on that? Because a lot of these trades are volatile and um, you know, big upswings, but there's also potential for for big downswings um because you don't know what you don't know. And so how would you get a long-term LP base? when you
12:52
Unidentified speaker (speaker_1): call it you don't want you want to get any capital you can? How would you think about that?
12:58
Unidentified speaker (speaker_0): I think your point about matching the investor to the manager, to the strategy, is critical a hundred percent. In the
13:11
Unidentified speaker (speaker_0): picture that I painted around a small manager being able to effectively trade vol, I think by its nature, those because of the small AUM, they're going to need to be very high frequency and they're not holding positions for an extended period of time. That's really a much better opportunity for a
13:31
Unidentified speaker (speaker_0): larger manager that can do that kind of um temporal diversification that I was speaking to. I think the successful small manager here. is going to look like
13:41
Unidentified speaker (speaker_0): the early stat arms that were trading ultra high frequency when nobody else was. And they were effectively, they kind of looked like market makers. I could see somebody being almost like a vol market maker today with a relatively small AUM developing, generating extremely high sharps on small
14:01
Unidentified speaker (speaker_0): money. I don't think those strategies scale, which is one of the reasons why we see either these very under-the-radar small shops that mint money just staying small, or they decide they're going to really become more multi-manager, diversified, multiple strategies, extending timelines
14:21
Unidentified speaker (speaker_0): lines and raising way more AUM.
14:23
Unidentified speaker (UNKNOWN): No,
14:23
Unidentified speaker (speaker_1): I mean one thing that's shocked me just from starting this podcast and meeting all sorts of different people is the number of people who are able to run strategies that are
14:36
Unidentified speaker (speaker_1): illiquid enough so that the big players don't capture that edge, but liquid enough to where they can make a very meaningful amount of money. Um I'd love to now ask a little bit more about how
14:51
Unidentified speaker (UNKNOWN): your
14:51
Unidentified speaker (speaker_1): time in the hedge fund business, how you think about risk, um informed the way rather informed the way you bet to take on the personal risk to start a venture fund. Um
15:06
Unidentified speaker (UNKNOWN): and
15:06
Unidentified speaker (speaker_1): then how you think about making bets in that space. Love to hear your some thoughts.
15:14
Unidentified speaker (speaker_0): The transition from hedge fund manager generally trading the most liquid markets to moving to become a venture capitalist at the most illiquid end of the continuum was definitely an adjustment. I think
15:31
Unidentified speaker (UNKNOWN): what
15:32
Unidentified speaker (speaker_0): I really appreciated from our special sits book at DB was how. the manager
15:41
Unidentified speaker (speaker_0): used illiquidity to their advantage in
15:46
Unidentified speaker (speaker_0): buying right and being able to hold for an extended period. It was a fantastic complement to all of the liquid markets Stat or high-frequency high frequency trading strategies that we had.
16:01
Unidentified speaker (UNKNOWN): And
16:01
Unidentified speaker (speaker_0): there's something about that that resonated. with me. I mean, look, at the end of the day, managers,
16:06
Unidentified speaker (UNKNOWN): I
16:07
Unidentified speaker (speaker_0): think, figure out who they are and what feels right.
16:11
Unidentified speaker (UNKNOWN): And
16:11
Unidentified speaker (speaker_0): one of the benefits I had from having this top-down view is I actually did figure out what's right for me, and that is at the illiquid end of the spectrum, where I then took that mindset to the venture business where very few.
16:27
Unidentified speaker (speaker_0): people have my background. I think probably only Rob Stavis at Bessemer, who ran Risk Garb at Solomon, has a similar background to me. And
16:37
Unidentified speaker (speaker_0): I do think about bet sizing a lot. So much of the work that I did while at DB, both on the derivatives desk and running DB advisors, this notion of sizing the positions correctly. for the strategy and the time horizon was something that I brought with me to the venture world.
16:57
Unidentified speaker (speaker_0): And that when I while bringing back the, you know, the poker analogy, one of the beauties of venture, if you are good, is having long date effectively long dated lockup money to run these programs is a perfect fit for the payout function of these early startups. Aaron
17:16
Unidentified speaker (UNKNOWN): Powell
17:16
Unidentified speaker (speaker_1): What would you say to
17:19
Unidentified speaker (UNKNOWN): call
17:19
Unidentified speaker (speaker_1): it Cliff Asnes, who says that a lot of venture guys are just laundering volatility?
17:24
Unidentified speaker (UNKNOWN): Aaron Powell
17:24
Unidentified speaker (speaker_0): Cliff has always had some very provocative views of things.
17:30
Unidentified speaker (speaker_0): Rather than say
17:33
Unidentified speaker (speaker_0): laundering volatility, I would say
17:37
Unidentified speaker (speaker_0): they are
17:39
Unidentified speaker (speaker_0): temporal maximizers. They best leverage time horizon in order for the vol to largely wash out until they're able to monetize, and then they pass it off to
17:55
Unidentified speaker (UNKNOWN): the
17:56
Unidentified speaker (speaker_0): public, essentially. Because the the best exit for most companies is to go public as opposed to selling if you're looking to actually maximize long-term value. But if you can think about the fact that
18:08
Unidentified speaker (UNKNOWN): I
18:09
Unidentified speaker (speaker_0): see the company today, it is going to gestate for seven to ten years.
18:15
Unidentified speaker (UNKNOWN): I
18:16
Unidentified speaker (speaker_0): then want to take it public.
18:18
Unidentified speaker (UNKNOWN): Over
18:18
Unidentified speaker (speaker_0): that seven to ten year horizon that I've held it, the market's gone up, the market's gone down, the company's died three times, it's come back to life. But if you then plot its trend,
18:29
Unidentified speaker (speaker_0): it's compounded at 40 to 50% per year over that time horizon, then you effectively monetize, and then you're handing the risk to another set of investors. That isn't vol
18:44
Unidentified speaker (speaker_0): laundering. That is being vol smart.
18:48
Unidentified speaker (speaker_1): Where does
18:50
Unidentified speaker (speaker_1): skill come into the venture investing playbook? And I'll give some context to how I'm thinking about that question. When I look at the top end, um, you know, your Sequoyas, um, Thrive, right? I feel like all of those bets are easy to make. As
19:10
Unidentified speaker (speaker_1): in um, if you gave me a list of the hottest companies call it in you know series A, B, um, that are that are growing massively, stack teams, um, it's a very easy bet to make. And I feel like if you gave me a good chunk of
19:30
Unidentified speaker (speaker_1): money, I could make those same bets, but I there's no way I'd have access to those deals. Um In your case,
19:37
Unidentified speaker (UNKNOWN): you know,
19:38
Unidentified speaker (speaker_1): small smaller fund, obviously a lot more skill involved embedding on certain companies. What do you see as the nodes or the the different variables that you can play with um that
19:54
Unidentified speaker (speaker_1): look when you're analyzing why you would have a right to win for a particular deal?
20:01
Unidentified speaker (speaker_0): The notion of right to win is is is the right question. And
20:06
Unidentified speaker (UNKNOWN): I
20:06
Unidentified speaker (speaker_0): think you're also
20:09
Unidentified speaker (speaker_0): correct in this very
20:12
Unidentified speaker (speaker_0): unusual time in the venture industry. And I've been doing this for almost 25 years now. And I'm I can honestly say I've never seen a market quite like this, where there is so much momentum in so few companies.
20:27
Unidentified speaker (speaker_0): VCs marking up other VCs deals, that it seems that you don't need to be a genius to make money in the venture game at scale. We're playing a completely different game. And I I could we could talk about that first point for a long time. I will talk about my game, and if you want to go back,
20:47
Unidentified speaker (speaker_0): that's fine.
20:49
Unidentified speaker (UNKNOWN): Uh
20:52
Unidentified speaker (speaker_0): So much money is leaning
20:56
Unidentified speaker (speaker_0): in one direction, namely the frontier models, the enabling technologies, the infrastructure to support the AI trade. And I think that's what you're referring to with the Sequoies and the Thrives and folks like that.
21:10
Unidentified speaker (UNKNOWN): What
21:10
Unidentified speaker (speaker_0): we're doing at Game Changers Ventures, and I would say our
21:15
Unidentified speaker (speaker_0): smaller peers,
21:18
Unidentified speaker (speaker_0): right to win is even more important because
21:22
Unidentified speaker (UNKNOWN): we
21:22
Unidentified speaker (speaker_0): are also looking for these generational companies,
21:28
Unidentified speaker (speaker_0): but the very best founders oftentimes will be
21:32
Unidentified speaker (speaker_0): legible to multiple venture fund GPs.
21:38
Unidentified speaker (UNKNOWN): What we
21:38
Unidentified speaker (speaker_0): try and do is,
21:43
Unidentified speaker (speaker_0): in the first instance, have theses that we believe will lead to generational companies that are not
21:52
Unidentified speaker (speaker_0): obvious and hard to do, even with the best founders. And then in terms of the GCV product itself.
22:03
Unidentified speaker (speaker_0): having founders perceive us as the best partners to help them get to product market fit and then early scaling. Because that's really when we get involved with companies.
22:21
Unidentified speaker (speaker_1): In decision making, in access. Um,
22:24
Unidentified speaker (UNKNOWN): And
22:24
Unidentified speaker (speaker_1): I'm asking this because this podcast is all about edge, it's all about alpha, it's about all about outperformance and figuring out reasons why you are going to win um across many different bats,
22:40
Unidentified speaker (speaker_1): having a higher hit rate, etc.
22:42
Unidentified speaker (speaker_1): What are some of the things you do to make sure that hit rate is always going up, to sharpen, sharpening that hit rate, making your bats more on the ball, etc. et cetera.
22:53
Unidentified speaker (speaker_0): When thinking about
22:56
Unidentified speaker (speaker_0): not just the right to win, but how to win, which I think is the nature of your question.
23:05
Unidentified speaker (speaker_0): There are really
23:06
Unidentified speaker (speaker_0): call it three, three parts to it. One is brand and recognition of what the process. product is and having that product be attractive to founders as they're in the early stages of their journey. Another is
23:24
Unidentified speaker (speaker_0): focusing on opportunities that are extremely hard and seemingly crazy that turn off a lot of our peers.
23:36
Unidentified speaker (UNKNOWN): In
23:37
Unidentified speaker (speaker_0): my own history, some of the best companies I backed seemed nuts. And the founders seemed
23:47
Unidentified speaker (speaker_0): aspirational to the point of being deluded that they could be successful.
23:55
Unidentified speaker (speaker_0): And then the third is
23:59
Unidentified speaker (speaker_0): I think the the respect of downstream peers who view us as attractive and successful stock pickers who can help them meet their next attractive Series A or Series B investment after we've nurtured these
24:19
Unidentified speaker (speaker_0): businesses.
24:20
Unidentified speaker (UNKNOWN): And
24:20
Unidentified speaker (speaker_0): it's really not about hit rate. I really don't care about hit rate. I care about slugging percentage.
24:27
Unidentified speaker (speaker_0): And so the discipline that we have, and this is discipline that I brought to IA Ventures when I started it. And again, I think my trading background has been super helpful in my work as a venture capitalist, is being hyper disciplined about time being the
24:47
Unidentified speaker (speaker_0): scarce resource, not money, and only investing in company. companies if you think they can be Grand Slams. If you really think that that company and that team, the founder Market Fit, gives them the chance to build a generational company
25:03
Unidentified speaker (speaker_0): because the failure rate is just too high
25:07
Unidentified speaker (speaker_0): that your successes have to be outliers. They just have to be. If you want to have the 5X. the net and above type returns.
25:16
Unidentified speaker (UNKNOWN): Aaron Ross Powell You
25:17
Unidentified speaker (speaker_1): don't think about hit rate, you think about slugging rate. How does that inform the way you view concentration in the portfolio?
25:25
Unidentified speaker (speaker_0): So portfolio construction is definitely something that I've been very loud about, but not loud in the sense that I'm right. Loud in the sense that
25:37
Unidentified speaker (UNKNOWN): this
25:37
Unidentified speaker (speaker_0): is how I have done it, do it, and how I've been successful, which is I am I do believe in concentration,
25:48
Unidentified speaker (speaker_0): partly because
25:51
Unidentified speaker (speaker_0): the amount of time it takes to be a good partner to founders who are especially our pre-product Product market fit is significant. So if time is truly your scarcest resource, then the N of companies that you and your team can deal with
26:06
Unidentified speaker (UNKNOWN): is
26:06
Unidentified speaker (speaker_0): naturally going to be smaller than if you're writing
26:11
Unidentified speaker (speaker_0): lower ownership participation checks and having a much lighter touch with your companies. By the way, uh there are managers that have done very well with that kind of strategy. There aren't many.
26:27
Unidentified speaker (speaker_0): but they exist.
26:29
Unidentified speaker (UNKNOWN): Right?
26:29
Unidentified speaker (speaker_0): I mean, I'll say, yeah, my friend David Tisha Box Group, you know, with Cursor, great, right? The 750 into a 750,000 into a billion,
26:40
Unidentified speaker (UNKNOWN): you
26:40
Unidentified speaker (speaker_0): know, I think David runs that playbook pretty much the best way you can.
26:46
Unidentified speaker (UNKNOWN): Uh,
26:46
Unidentified speaker (speaker_0): but that's also a combination of brand. Founders know what they're getting with with him and his partners. and they do a good job. We're playing a different game.
26:56
Unidentified speaker (UNKNOWN): We
26:56
Unidentified speaker (speaker_0): play our game well,
26:58
Unidentified speaker (UNKNOWN): and
26:58
Unidentified speaker (speaker_0): we've done exceptionally well. So I think it's less about there being the right game. It is being great at the game you're playing and having founders know that.
27:10
Unidentified speaker (speaker_1): How many companies
27:13
Unidentified speaker (UNKNOWN): are
27:14
Unidentified speaker (speaker_1): you
27:15
Unidentified speaker (UNKNOWN): how
27:15
Unidentified speaker (speaker_1): many portfolio companies do you currently have?
27:18
Unidentified speaker (speaker_0): So we have sixteen core investments. And when we think about
27:24
Unidentified speaker (speaker_0): how many shots on goal, if you will, we want to have in a given fund,
27:30
Unidentified speaker (speaker_0): it's about twenty five.
27:33
Unidentified speaker (speaker_0): And that twenty 25 five number has been consistent from when I first started IA back in two thousand. 2009 through the four funds when I was involved, and now through to game changers. That is
27:47
Unidentified speaker (speaker_0): a really good size in an early, early stage portfolio because you have enough
27:57
Unidentified speaker (speaker_0): shots on goal such that two, 234 three, four
28:02
Unidentified speaker (UNKNOWN): are.
28:02
Unidentified speaker (speaker_0): likely to be successful, very successful, if you're a good stock picker, which is more than enough to generate the 5 to 10x net on smaller fund sizes.
28:16
Unidentified speaker (speaker_0): And so
28:17
Unidentified speaker (UNKNOWN): at
28:18
Unidentified speaker (speaker_0): the same time, we can spend enough time with the companies given that relatively small number, and you know, we have four. investors on our team to more than handle that delivering the game changer's product in the way that we want to. And then, of course, what happens is, well, what about fund two? When you tack
28:38
Unidentified speaker (speaker_0): on another 25 companies, doesn't this start to uh overwhelm you unless you constantly are increasing team size? The answer is no, because when you start to raise subsequent funds,
28:51
Unidentified speaker (speaker_0): companies in the first fund or the fund prior have matured. And they don't need as much
28:59
Unidentified speaker (speaker_0): time simply because they've either A, made it to that next level, or B, if it's not working, then we with the founders find a way to engineer a soft landing. So there tends to be that selection. that happens organically in the couple of years as you're deploying a fund
29:19
Unidentified speaker (speaker_0): before you raise the next
29:21
Unidentified speaker (UNKNOWN): one.
29:22
Unidentified speaker (speaker_1): One of the things I see that's different about the venture business compared to the hedge fund business is in the hedge fund business in general, as your winners go up in value because you're mark-to-market. you really need to cut those if
29:41
Unidentified speaker (speaker_1): you want to generate high sharp
29:44
Unidentified speaker (UNKNOWN): in
29:44
Unidentified speaker (speaker_1): say equities.
29:45
Unidentified speaker (UNKNOWN): In
29:46
Unidentified speaker (speaker_1): venture, what I tend to see is,
29:48
Unidentified speaker (UNKNOWN): you know,
29:48
Unidentified speaker (speaker_1): call it, you bet,
29:50
Unidentified speaker (UNKNOWN): you
29:50
Unidentified speaker (speaker_1): give you know, give a little bit of money to this company, all of a sudden they're they're growing really fast. And you say, hold on, this is the winner. Let's add to the position, right? How do you reconcile those two?
30:03
Unidentified speaker (speaker_1): ways of thinking? Um is it less risky as it make as as the company gets bigger? Give me your thoughts around that.
30:12
Unidentified speaker (speaker_0): So ongoing investment in companies and liquidity realization,
30:17
Unidentified speaker (UNKNOWN): I
30:17
Unidentified speaker (speaker_0): mean these are the hot button issues in early stage venture.
30:22
Unidentified speaker (UNKNOWN): Uh
30:23
Unidentified speaker (speaker_0): most money in venture isn't lost on the first check. It's lost on the second, third, and fourth checks. It's when you essentially have been undisciplined in terms of your follow-on strategy and done blind prorata participations and things like
30:43
Unidentified speaker (UNKNOWN): that.
30:45
Unidentified speaker (speaker_0): I uh have felt and feel very strongly that every investment in a portfolio company needs to stand on its own. Notwithstanding the relationship implications of a lead not doing full prorrata in a subsequent financing round,
31:05
Unidentified speaker (speaker_0): we are extremely honest with our founders from the get-go get go that we are not going to do that. There are times. we will make a small participation to show support, but won't do full prerata because we're not happy with the company's performance and where they're going. There are times
31:25
Unidentified speaker (speaker_0): we will try and preempt rounds and and or do super prorata if we can in the companies that we think are performing best. So one side of your question kind of doesn't deals
31:38
Unidentified speaker (UNKNOWN): have to be a good thing
31:39
Unidentified speaker (speaker_0): with that and how you build concentration in companies. Then the other side is actually realizations. You were talking about hedge fund managers in liquid markets trimming winners, they're also getting paid on that, whereas venture managers are not,
31:57
Unidentified speaker (speaker_0): especially because with a venture manager, if you want to generate liquidity early, meaning company isn't sold in total or isn't IPO'd,
32:07
Unidentified speaker (UNKNOWN): you're
32:08
Unidentified speaker (speaker_0): selling in a secondary. You are, on the one hand, generating liquidity. On the other hand, you're truncating your gains. So you need to be very conscious that
32:18
Unidentified speaker (UNKNOWN): if
32:18
Unidentified speaker (speaker_0): you're selling down in one of your best companies, is that the right thing to be doing? And then you still need to clear your fund size in terms of returning. capital back to LPs before you can start participating in the profits. Both them and you, which is why I
32:36
Unidentified speaker (speaker_0): excuse me, almost never
32:40
Unidentified speaker (speaker_0): I historically have used secondary very sparingly,
32:45
Unidentified speaker (speaker_0): because I have felt like our best investments are likely to compound. at really high rates, where my LPs would rather keep the money in the companies if they're compounding at 40-50% per year than me giving it back to them right now. So
33:06
Unidentified speaker (speaker_0): it's it is an art. And I would say now secondary has become an even more critical component of
33:14
Unidentified speaker (UNKNOWN): early
33:14
Unidentified speaker (speaker_0): stage. managers' portfolio strategy because of some of the explosive amount of appreciation. Now, I will say I have never backed a company that in three years has gone from
33:28
Unidentified speaker (speaker_0): twenty million to sixty billion, right? And how would I behave if I had the opportunity, if I own 10% of a company at a sixty. 60, even 5% of a company at a 60, even the even if I thought it had tons of room to run, even if I thought it could be a trillion dollar company, would I not sell some
33:48
Unidentified speaker (speaker_0): secondary at 60 billion in a company that had exploded in value over a relatively short span of time? I absolutely would.
33:57
Unidentified speaker (speaker_1): You talking about secondaries and how you think about um
34:02
Unidentified speaker (UNKNOWN): call it
34:02
Unidentified speaker (speaker_1): taking profit um
34:06
Unidentified speaker (speaker_1): has got you know it has me reflecting on the way
34:13
Unidentified speaker (speaker_1): hedge fund guys deconstruct the returns um and figure out points in time where they have won and points at of time where they have lost and them having a lot more data to to to actually look. at. You know, if you're
34:28
Unidentified speaker (UNKNOWN): you
34:28
Unidentified speaker (speaker_1): have a call it 50 names, right? You can look
34:32
Unidentified speaker (UNKNOWN): at when
34:33
Unidentified speaker (speaker_1): things do well, when things don't. Um to give an example, I
34:38
Unidentified speaker (UNKNOWN): know
34:39
Unidentified speaker (speaker_1): a fund manager who hired a quant
34:42
Unidentified speaker (UNKNOWN): and
34:42
Unidentified speaker (speaker_1): and was, you know, he was decomposing Composing his returns, and he said, every time it
34:48
Unidentified speaker (UNKNOWN): it
34:49
Unidentified speaker (speaker_1): it seemed like every time. this fund manager would have the idea to buy a stock, it would run up a little bit before he would actually pile in. Um, and so it's there's all these different behavioral things that you can
35:04
Unidentified speaker (UNKNOWN): that
35:05
Unidentified speaker (speaker_1): that you can see in one's returns because you have a much larger sample size. I'm wondering how does that translate to venture where you have fewer. sample, you know, your your your portfolio is not marked on a daily basis. You can't see the volume, you can't see if your entry was optimal or suboptimal until the future. And even then,
35:25
Unidentified speaker (speaker_1): you're still looking back with hindsight bias. Broadly, I guess my general question is how do you decompose your returns and figure out um errors in your process?
35:38
Unidentified speaker (UNKNOWN): So
35:38
Unidentified speaker (speaker_0): thinking about about process errors
35:42
Unidentified speaker (speaker_0): fall into two buckets in venture. One is security selection. Should I have invested in this company or not? And
35:53
Unidentified speaker (speaker_0): I'll call it kind of pricing, which is when I got in, what price that I did I pay? And then as I've made subsequent investment decisions, were those rational. decisions price and volume when I made
36:06
Unidentified speaker (UNKNOWN): them? In
36:07
Unidentified speaker (speaker_0): the first case,
36:10
Unidentified speaker (speaker_0): this is where
36:13
Unidentified speaker (speaker_0): venture and quant strategies diverge. There is much less science in the work that we're doing at the very earliest stages of venture investing. There have been and are funds that do use quantitative methods for security
36:33
Unidentified speaker (speaker_0): selection. Early firms that were scraping LinkedIn and trying to identify the best talent when they left places like Signalfire. They've done very well. There's a lot of firms now, obviously with using AI tools to try and build similar frameworks for. identifying which founders are
36:53
Unidentified speaker (speaker_0): likely to be the most successful.
36:56
Unidentified speaker (speaker_0): Notwithstanding my quant background, I have not
37:00
Unidentified speaker (speaker_0): spent a ton of time around that because I personally don't believe that that is either the only way or even the best way for founder selection, because it's not just great
37:15
Unidentified speaker (speaker_0): people, it's great people building the right things into the right markets.
37:20
Unidentified speaker (UNKNOWN): So
37:21
Unidentified speaker (speaker_0): this notion of founder market fit is something that we talk about a lot at Chi C V.
37:26
Unidentified speaker (speaker_0): So
37:28
Unidentified speaker (speaker_0): that that so we're constantly scrutinizing our process for identifying best founder,
37:38
Unidentified speaker (speaker_0): best product vision, best market opportunity.
37:43
Unidentified speaker (speaker_0): Then there's the subsequent investment decision. And that's something that, of course, without hindsight,
37:52
Unidentified speaker (UNKNOWN): you
37:52
Unidentified speaker (speaker_0): don't know.
37:54
Unidentified speaker (speaker_0): GCB is really too young for us to have material data on that yet. We do have material data from my work at IA.
38:04
Unidentified speaker (speaker_0): And um
38:06
Unidentified speaker (speaker_0): Ethan when I had spoken about most of venture money being lost, not on the first check, but on the second, third, and fourth checks. We spend
38:18
Unidentified speaker (speaker_0): so learnings from IA,
38:21
Unidentified speaker (speaker_0): I a few times I had invested two, three, four times. in companies that ultimately did poorly upon exit. That has led to a tremendous amount of introspection. What information did I have when I made those decisions? What did I get wrong? Definitely it
38:41
Unidentified speaker (speaker_0): brought that to our work at GCV. That those to me are the most painful process errors when you actually are thrown throwing good money after what turned out to be bad,
38:57
Unidentified speaker (speaker_0): where somehow either because you became so wedded to the idea, so wedded to the founders that you know I was blind to the fact that things weren't as good as they seemed, and as such, I continued to invest more. money and then lost that money when the company
39:17
Unidentified speaker (speaker_0): became unsuccessful.
39:18
Unidentified speaker (speaker_1): Yeah, I mean, I guess I want to ask you a personal question. What are the things you specifically learned from those moments? Um
39:26
Unidentified speaker (speaker_1): when do you personally win with your venture investments?
39:32
Unidentified speaker (speaker_0): So the times that I have lost,
39:35
Unidentified speaker (speaker_0): I have lost because
39:39
Unidentified speaker (speaker_0): I let emotions get in the way of critical decision making.
39:45
Unidentified speaker (speaker_0): Part of being a venture capitalist by nature is having an optimistic abundance mindset. So that that is the that is the frame of pretty much every successful venture investor.
40:01
Unidentified speaker (UNKNOWN): And
40:01
Unidentified speaker (speaker_0): that's great. And that's part of what makes the job so exciting. The
40:07
Unidentified speaker (speaker_0): potential downside of that is not being able to acknowledge when things aren't working because you so want them to work. And there is also the sunk On cost of having invested enormous amounts of time, forget about money, in building relationships. with the
40:27
Unidentified speaker (speaker_0): team. And it becomes really, really hard to have the conversation,
40:34
Unidentified speaker (UNKNOWN): I'm
40:34
Unidentified speaker (speaker_0): not going to write another check because I've lost faith.
40:39
Unidentified speaker (speaker_0): So
40:42
Unidentified speaker (speaker_0): I try very hard not to do that anymore. I think I've gotten way better at it, because at the end of the day, Ethan. it actually helps the founders. These are really super bright, high potential people that could do lots of things with their time. I don't want them wasting time
40:59
Unidentified speaker (UNKNOWN): if
40:59
Unidentified speaker (speaker_0): they're working on a company that should just not exist. And I don't want to be an enabler of that.
41:07
Unidentified speaker (speaker_0): When I win
41:10
Unidentified speaker (speaker_0): in venture,
41:12
Unidentified speaker (speaker_0): it is because I have
41:15
Unidentified speaker (speaker_0): generally identified a
41:19
Unidentified speaker (speaker_0): contrarian idea
41:23
Unidentified speaker (speaker_0): with
41:25
Unidentified speaker (speaker_0): deeply domain knowledgeable founders who are mad.
41:31
Unidentified speaker (speaker_0): Those are the three things, those are the three common threads. They're mad because
41:36
Unidentified speaker (speaker_0): they tried to do this somewhere else and it didn't work out. There was co-founder conflict. Their company was subscale, got bought by a big company and that ruined
41:47
Unidentified speaker (UNKNOWN): it.
41:49
Unidentified speaker (speaker_0): They
41:50
Unidentified speaker (speaker_0): have been trying to raise money for this idea, and they just couldn't sell it to raise the money necessary to start out. Whatever the reason, founders with a chip on their shoulder.
42:03
Unidentified speaker (speaker_0): are unbelievably dangerous in a good way. When they latch on to
42:11
Unidentified speaker (speaker_0): a great vision and are,
42:15
Unidentified speaker (speaker_0): I'll say, have a non-negotiable vision, but highly negotiable missions. They're very adaptable as they get information from the market on how to adjust tactics
42:26
Unidentified speaker (UNKNOWN): in order to
42:27
Unidentified speaker (speaker_0): to get to their vision.
42:30
Unidentified speaker (speaker_0): And where we have a thesis about the space that they're building in,
42:37
Unidentified speaker (speaker_0): which is not believed by the rest of the community. And as such, we have the opportunity to buy ownership at attractive prices to back these visionary, pissed-off founders. to build something truly transformative. Literally, every one of my most successful companies have those three
42:56
Unidentified speaker (speaker_0): characteristics.
42:58
Unidentified speaker (speaker_1): What would you say? Are there any differences between the traits and skill sets that make an excellent hedge fund investor, call that fundamental equities, macro, whatever that may be, um, that don't translate very well to venture. Um and
43:18
Unidentified speaker (speaker_1): yeah, talk a little bit about that.
43:20
Unidentified speaker (speaker_0): That's great. That's a great question. Hedge fund managers and and venture managers by their nature are very, very different. Especially l let's be clear.
43:31
Unidentified speaker (UNKNOWN): I
43:31
Unidentified speaker (speaker_0): do think there's convergence at with late stage venture and public equity trading, which is why you've seen uh the the Chase Colemans and people like that be very successful at the intersection of those two, um, or
43:47
Unidentified speaker (UNKNOWN): Don
43:48
Unidentified speaker (speaker_0): Sunheim, whatever.
43:50
Unidentified speaker (UNKNOWN): Um
43:53
Unidentified speaker (speaker_0): I think the characteristics that make a public
43:59
Unidentified speaker (speaker_0): equity hedge fund investor
44:02
Unidentified speaker (speaker_0): bad at very early stage venture is A impatience, B
44:12
Unidentified speaker (speaker_0): lack of
44:14
Unidentified speaker (UNKNOWN): I
44:14
Unidentified speaker (speaker_0): guess to a degree I'll call it lack of vision, because if your time horizon is so short, you don't really need much vision.
44:23
Unidentified speaker (speaker_0): What you need is
44:25
Unidentified speaker (UNKNOWN): a
44:25
Unidentified speaker (speaker_0): better understanding of how others will perceive an opportunity over a fairly short span of time. As you lengthen holding period, then it starts to look more and more like venture. So if you think somebody like Warren Buffett, Warren Buffett would have been a great early stage venture capitalist if he wasn't so focused Focus on Ben
44:45
Unidentified speaker (speaker_0): Graham margin of safety. But in terms of his ability to withstand volatility, and to look dumb at times in a single position over years and years and years, but then eventually to be vindicated, that is much more the mindset of a very early stage
45:05
Unidentified speaker (speaker_0): investor than it is a long, short hedge fund manager.
45:09
Unidentified speaker (UNKNOWN): No,
45:09
Unidentified speaker (speaker_3): It's it's funny the thing you mentioned about the the short time short time horizons, you don't need vision. One hundred percent I think most of the claw pod analyst game is just positioning. I mean you don't really need to at some point it's not even really fundamental analysis. It's a positioning game to
45:29
Unidentified speaker (speaker_3): think about fundamental analysis in the context of where everyone else is gonna be and how things are gonna move.
45:35
Unidentified speaker (speaker_1): Are there elements of that within venture? Um and forgive me if this is a dumb question,
45:41
Unidentified speaker (UNKNOWN): but
45:41
Unidentified speaker (speaker_1): um
45:43
Unidentified speaker (speaker_1): surely there must be moments
45:45
Unidentified speaker (speaker_3): where
45:45
Unidentified speaker (speaker_1): you plow capital into a company
45:49
Unidentified speaker (UNKNOWN): and
45:49
Unidentified speaker (speaker_1): you're thinking, okay, we have a pre we have a certain reputation for say backing defense companies. Um we've just backed Back a to defense company. We know this is gonna get marked
45:59
Unidentified speaker (UNKNOWN): up.
46:00
Unidentified speaker (speaker_1): We can exit a little bit at the secondaries and then we do also believe in the company, but if we do that over and over again,
46:06
Unidentified speaker (UNKNOWN): uh
46:07
Unidentified speaker (speaker_1): we know that the fund is the returns of the fund are a lot more guaranteed and we're kind of
46:12
Unidentified speaker (UNKNOWN): um
46:13
Unidentified speaker (speaker_1): locking in some of those gains. I'm curious how much of that thinking can translate from the hedge fund business to venture.
46:21
Unidentified speaker (speaker_0): I really think the hedge fund game and the very early stage venture game are radically different games. I do not think there's any room for that kind of thinking in early stage venture for the reason that
46:34
Unidentified speaker (speaker_0): is basically batting average and the fact that what you really care about is slugging percentage. Applying that mindset of
46:47
Unidentified speaker (speaker_0): risk reduction, liquidity optimism. optimization in the context of venture, you are going to be truncating those outliers that are those things that are going to generate the 5 to 10x net generational fund type returns where you're generating the 30 to 40%
47:07
Unidentified speaker (speaker_0): IRRs.
47:09
Unidentified speaker (speaker_0): You need to let winners run. This is why I'll come back to Warren Buffett.
47:14
Unidentified speaker (UNKNOWN): He
47:14
Unidentified speaker (speaker_0): and I have way more in common because of this. Long holding period, finding great businesses, long holding period, let capital compound.
47:26
Unidentified speaker (speaker_0): It compounds tax deferred. And
47:31
Unidentified speaker (speaker_0): ultimately, when you compound at 30, 40, 50 percent rates for long periods of time, you want end up building these generational companies and generational fund returns because of them. So the buy the buying and selling, unless it's done at prices that are effectively
47:51
Unidentified speaker (speaker_0): what look like enterprise full value,
47:54
Unidentified speaker (UNKNOWN): and
47:55
Unidentified speaker (speaker_0): you're effectively exiting in a way that would enable you to return multiples of a fund to your LPs, it's just not worth it.
48:03
Unidentified speaker (UNKNOWN): I
48:03
Unidentified speaker (speaker_1): was talking
48:04
Unidentified speaker (UNKNOWN): to
48:04
Unidentified speaker (speaker_3): uh um uh
48:06
Unidentified speaker (UNKNOWN): a
48:06
Unidentified speaker (speaker_3): friend recently who's also a guest of the podcast, and he was
48:12
Unidentified speaker (speaker_3): talking about a very, very prominent venture investor whose brother is very senior at one of the top um multimanager hedge funds.
48:23
Unidentified speaker (UNKNOWN): And
48:23
Unidentified speaker (speaker_3): he was saying that the um the the the venture investor at you know one of the largest
48:30
Unidentified speaker (speaker_1): venture funds um
48:32
Unidentified speaker (speaker_1): is worth multiples of his brother.
48:37
Unidentified speaker (UNKNOWN): And
48:37
Unidentified speaker (speaker_1): I wanted to ask you a question about
48:41
Unidentified speaker (speaker_1): how you think about long-term wealth creation.
48:46
Unidentified speaker (UNKNOWN): You know
48:46
Unidentified speaker (speaker_1): in the in in the you know call it fundamental investor, quant trader, whatever that is, hedge fund guy. Um, I think the game is often framed as uh
48:56
Unidentified speaker (speaker_1): I have X amount liquid, I compound at a certain percentage every year.
49:02
Unidentified speaker (UNKNOWN): Um,
49:02
Unidentified speaker (speaker_1): and my goal is just to maximize that, hopefully with minimal drawdowns, but just compound and over a long time horizon make a big pile of money.
49:13
Unidentified speaker (UNKNOWN): Um,
49:13
Unidentified speaker (speaker_1): whereas for venture, you have these upswings, you can choose not to mark your portfolio or not mark fewer of your position positions on the downswing, um, but
49:22
Unidentified speaker (UNKNOWN): you know,
49:22
Unidentified speaker (speaker_1): hopefully ride a a a generational wave and ride tailwinds like AI, um, and and everything else that's happening.
49:31
Unidentified speaker (UNKNOWN): Um, how
49:32
Unidentified speaker (speaker_1): do
49:32
Unidentified speaker (speaker_3): you personally think about
49:34
Unidentified speaker (UNKNOWN): the
49:34
Unidentified speaker (speaker_3): wealth
49:34
Unidentified speaker (UNKNOWN): creation
49:35
Unidentified speaker (speaker_3): game? How do you try
49:36
Unidentified speaker (UNKNOWN): to
49:37
Unidentified speaker (speaker_3): accumulate a, you
49:38
Unidentified speaker (UNKNOWN): know, you know
49:39
Unidentified speaker (speaker_1): how
49:39
Unidentified speaker (speaker_3): do you how do you think about growing your wealth and year after year compounding?
49:43
Unidentified speaker (speaker_1): Because I imagine after
49:45
Unidentified speaker (speaker_3): all your positions,
49:46
Unidentified speaker (speaker_1): you
49:46
Unidentified speaker (UNKNOWN): know,
49:47
Unidentified speaker (speaker_3): uh have ex
49:48
Unidentified speaker (speaker_1): have you've exited all your all of your positions after a
49:51
Unidentified speaker (speaker_3): liquidity
49:51
Unidentified speaker (speaker_1): generation event, um I'm sure you have skill and I'm sure you could bet on the public markets and companies there are are are amazing. You have the best founders there. How do you think about the game personally?
50:04
Unidentified speaker (speaker_0): Wealth maximization is interesting framing. And I do think that
50:10
Unidentified speaker (speaker_0): much of the hedge fund industry is geared around that principle. I do not feel that same way in venture. I think it's partly a culture issue. I think it's partly the different psychologies of the people. that go into these businesses. When I
50:30
Unidentified speaker (speaker_0): left Deutsche to kind of change what I wanted to do,
50:35
Unidentified speaker (UNKNOWN): I
50:36
Unidentified speaker (speaker_0): could have gone into the hedge fund game. I could have raised a hedge fund. I could have gone out and raised billions of dollars.
50:44
Unidentified speaker (UNKNOWN): I
50:44
Unidentified speaker (speaker_0): could have raised billions of dollars at IA ventures. You know, we were very, very successful. We could have raised way larger funds. than we did.
50:52
Unidentified speaker (speaker_3): Why didn't
50:52
Unidentified speaker (UNKNOWN): you?
50:54
Unidentified speaker (speaker_0): Because that's not the game that I want to play. It's just not. I'm not
51:01
Unidentified speaker (speaker_0): my objective function is not wealth maximization. It is impact optimization.
51:11
Unidentified speaker (speaker_0): Wealth creation is part of it. It's not the goal.
51:17
Unidentified speaker (speaker_0): Building,
51:18
Unidentified speaker (speaker_0): enabling great builders to build great companies that do good things that employ tens of thousands of people.
51:28
Unidentified speaker (speaker_0): That makes me feel good. That's part of the goal. Making money for my LPs, ILPs who themselves are mission-driven and do good things. That's important to me. So as I think about
51:42
Unidentified speaker (UNKNOWN): this
51:42
Unidentified speaker (speaker_0): kind of third phase of my career in firm building, there was kind of firm building within the context of Deutsche,
51:50
Unidentified speaker (UNKNOWN): there
51:51
Unidentified speaker (speaker_0): was IA Ventures, and now there's GCV,
51:55
Unidentified speaker (speaker_0): it is very much with the eye towards building something sustainable
52:01
Unidentified speaker (UNKNOWN): in
52:01
Unidentified speaker (speaker_0): terms of my teammates and the culture and the way we do business and our brand. It is backing that next generation of incredible founders that with their success, they will have life-changing outcomes for themselves and their families. We will have no, it'll be life-changing for my partners, not really
52:21
Unidentified speaker (speaker_0): life-changing for me at this juncture, but
52:23
Unidentified speaker (UNKNOWN): I'm
52:24
Unidentified speaker (speaker_0): ultra-competitive. ultra competitive. I want to win. I want to prove. to people that
52:30
Unidentified speaker (speaker_0): we don't need to be all in on the AI trade in this narrow way with gigantic AUM to be successful. We can be contrarian and still generate generational returns.
52:43
Unidentified speaker (speaker_0): That that to me is the game. It is not a wealth creation game. That is a byproduct of doing great work. enabling great founders to do great things, and it all flows from that.
52:56
Unidentified speaker (speaker_1): We don't need to be all in
52:58
Unidentified speaker (speaker_3): on the AI trade in order to make great returns.
53:01
Unidentified speaker (UNKNOWN): I
53:01
Unidentified speaker (speaker_3): want to ask a question
53:02
Unidentified speaker (speaker_1): about that and how you see the
53:09
Unidentified speaker (speaker_1): Sounds a little vague, but the
53:11
Unidentified speaker (speaker_3): economy
53:11
Unidentified speaker (speaker_1): evolving.
53:13
Unidentified speaker (UNKNOWN): I
53:13
Unidentified speaker (speaker_1): was
53:14
Unidentified speaker (UNKNOWN): walking
53:14
Unidentified speaker (speaker_3): around New York the other day. I think it was,
53:17
Unidentified speaker (UNKNOWN): I was
53:17
Unidentified speaker (speaker_1): having
53:18
Unidentified speaker (speaker_3): brunch in the West Village on a Saturday, right? And
53:21
Unidentified speaker (UNKNOWN): I'm
53:21
Unidentified speaker (speaker_3): eating there and I'm thinking about,
53:25
Unidentified speaker (UNKNOWN): um,
53:25
Unidentified speaker (speaker_3): you know, I'm with friends,
53:27
Unidentified speaker (UNKNOWN): but
53:27
Unidentified speaker (speaker_3): I'm looking around. Um,
53:29
Unidentified speaker (speaker_1): and New York is a city, especially right now, where you can just feel the prosperity.
53:36
Unidentified speaker (UNKNOWN): You
53:37
Unidentified speaker (speaker_1): know, it's it's unlike anything. I've ever felt. I've lived in Hong Kong, I've lived in London. When I go back, I visited London over the summer, Hong Kong as well.
53:48
Unidentified speaker (UNKNOWN): It
53:48
Unidentified speaker (speaker_1): doesn't feel the same way. And I think the reason for that is the US has been able to ride the AI wave in a way where no other country has, and it's across every single industry.
54:04
Unidentified speaker (UNKNOWN): You
54:04
Unidentified speaker (speaker_1): you know, you think about finance, you know, the deals getting done, it's all data centers, it's all related to AI.
54:12
Unidentified speaker (UNKNOWN): You
54:12
Unidentified speaker (speaker_1): know, I I even think about a podcast business. My beta to AI is really high. Right? If this thing tanks, no one's sponsoring a a hedge fund podcast.
54:23
Unidentified speaker (UNKNOWN): You
54:24
Unidentified speaker (speaker_1): know, people are gonna downsize their expenditure across everything. You know, I'm probably three X levered to it, uh laundering the VAL, right? Um and it just feels like everyone is all in, all the billionaires, every single person is all in. And it the reason why I would bet on it, um, maybe this
54:43
Unidentified speaker (speaker_1): is dangerous, but it's it's kinda kind of like it has to work out. Yeah, it feels like it's AI or bust sort of thing.
54:51
Unidentified speaker (UNKNOWN): Can
54:51
Unidentified speaker (speaker_1): I get your take on that?
54:53
Unidentified speaker (speaker_0): There's a lot to unpack there, Ethan. That was quite a question.
55:04
Unidentified speaker (speaker_0): So
55:07
Unidentified speaker (speaker_0): AI or bust, AI has to work out, I think is different than
55:14
Unidentified speaker (speaker_0): investing in AI.
55:18
Unidentified speaker (speaker_0): There is no doubt that AI
55:22
Unidentified speaker (speaker_0): has and will continue to disrupt, transform, optimize industries in ways previously unimaginable. Like that's a done deal. That's happening.
55:33
Unidentified speaker (UNKNOWN): And
55:33
Unidentified speaker (speaker_0): then you roll back, okay, well, what does that necessarily mean from an infrastructure perspective? Compute token prices. So that's where all the models get built.
55:45
Unidentified speaker (UNKNOWN): And
55:45
Unidentified speaker (speaker_0): that's how we have this kind of daisy chain of chips, data centers, public and private credit providers, et cetera, um, that are very interconnected, which is why
56:02
Unidentified speaker (speaker_0): from a
56:05
Unidentified speaker (UNKNOWN): I
56:05
Unidentified speaker (speaker_0): won't even call it bust, because AI long-term wins. No question about that.
56:12
Unidentified speaker (UNKNOWN): But
56:12
Unidentified speaker (speaker_0): again, coming back to the volatility point,
56:15
Unidentified speaker (UNKNOWN): oh,
56:15
Unidentified speaker (speaker_0): there's going to be some vol. We've already seen it, and we're gonna see a lot more because of the interconnectedness. Connectedness And there have been several kind of market commentators who have said, you know, this is more a akin to 07-08 than it is 99-2000. And I do think that's right because of how interconnected everything is. And like you're you were
56:35
Unidentified speaker (speaker_0): saying, prosperity in New York, you kind of haven't felt that before. It is true that every industry is being either disrupted by, but then winners are winning bigger, or they're service servicing the AI industry, Wall Street
56:55
Unidentified speaker (speaker_0): consultancy, law firms. There is so much money because of this.
57:03
Unidentified speaker (speaker_0): My hypothesis is that we are going to achieve a higher plane because
57:11
Unidentified speaker (speaker_0): we will be more productive and get to do these things that we couldn't do. previously.
57:17
Unidentified speaker (UNKNOWN): Dru
57:17
Unidentified speaker (speaker_0): you know, whether it's drug discovery, space travel, everything.
57:23
Unidentified speaker (UNKNOWN): And
57:23
Unidentified speaker (speaker_0): that's great. But
57:27
Unidentified speaker (UNKNOWN): we
57:27
Unidentified speaker (speaker_0): are heavily levered to the phenomenon right now.
57:33
Unidentified speaker (UNKNOWN): And
57:33
Unidentified speaker (speaker_0): I'm not saying this is why we invest, how we invest, I would say it's partly a byproduct of it, not the raison d'etre, which is because so much focus is on the fundamental AI infrastructure trade,
57:49
Unidentified speaker (speaker_0): much of the investment community doesn't care about anything else. Those are the times in my life as a market participant, be it on Wall Street or in the venture space, that I have captured the greatest opportunity is when things are not legible, when things are not in
58:09
Unidentified speaker (speaker_0): favor, and kind of when no one cares. And that's literally that's what I get today. And I do feel like this is something that's going to go on from for many, many years. Like it's not, oh, oh, the AI trade, there's a there's a bust and now no more AI, and all of a sudden money flows back at. in everything else. I don't believe that. I think there's
58:29
Unidentified speaker (speaker_0): so much that is going to continue to happen in AI for a generation that there's going to be this capital starvation of all things that are not that. That makes me excited.
58:42
Unidentified speaker (speaker_1): You're a junior today.
58:44
Unidentified speaker (UNKNOWN): How
58:44
Unidentified speaker (speaker_1): would you bet your career?
58:46
Unidentified speaker (UNKNOWN): Um,
58:46
Unidentified speaker (speaker_1): and I asked ask that with the context of
58:51
Unidentified speaker (speaker_1): so many of the previously extremely valuable skill sets are
58:58
Unidentified speaker (speaker_1): not just getting automated
59:01
Unidentified speaker (speaker_1): but
59:02
Unidentified speaker (UNKNOWN): are
59:02
Unidentified speaker (speaker_1): going down in value. Stuff that used to be learn this and you're good.
59:09
Unidentified speaker (speaker_1): Skill sets are getting commoditized, and the incentive structure is to replace those skill sets first. Um and so if you were a young person today, how would you think about building competitive advantage and motes around your skill set that won't be
59:28
Unidentified speaker (speaker_1): commoditized by the models?
59:30
Unidentified speaker (UNKNOWN): That's
59:30
Unidentified speaker (speaker_0): a really great question, Ethan. Young people today
59:35
Unidentified speaker (speaker_0): have it harder in many ways
59:39
Unidentified speaker (speaker_0): certainly than than I did, even though it seems uh the world is so much seemingly better with so many more opportunities, but because of the singular force that is so powerful, it's made those safe careers, like being a coder, not safe anymore.
59:58
Unidentified speaker (UNKNOWN): If
59:59
Unidentified speaker (speaker_0): if I'm a junior
1:00:03
Unidentified speaker (speaker_0): I think the competitive motes are around human interaction, creating your own brand, being relevant to other humans, and not necessarily trying to compete head on with the world of AI. There are going to be, and there are
1:00:22
Unidentified speaker (speaker_0): obviously top technical talent that will get hired by the frontier companies and the component makers who are unbelievable scientists, and that's that is the highest and best use of their time. That is not most people. That is a teeny fraction of the people that are graduating from top universities
1:00:42
Unidentified speaker (speaker_0): today.
1:00:44
Unidentified speaker (speaker_0): I think
1:00:46
Unidentified speaker (speaker_0): the broadest set of winners in this society are going to be able
1:00:51
Unidentified speaker (speaker_0): to talk to other people, service other people, generate content,
1:00:57
Unidentified speaker (speaker_0): are themselves compelling and can build a brand because the intersection of technology and customers is going to become even more important. Technology
1:01:09
Unidentified speaker (UNKNOWN): or
1:01:10
Unidentified speaker (speaker_0): a set of power tools.
1:01:11
Unidentified speaker (UNKNOWN): It
1:01:12
Unidentified speaker (speaker_0): is not a panacea.
1:01:14
Unidentified speaker (speaker_0): You still need people. You need people to develop new products. You need people in services businesses. You need people to build the next great set of companies.
1:01:29
Unidentified speaker (speaker_0): At the same time, I feel like social media and our
1:01:35
Unidentified speaker (speaker_0): still not having figured out how to. best use that
1:01:41
Unidentified speaker (speaker_0): for young people.
1:01:46
Unidentified speaker (speaker_0): So many young people I talk to are almost afraid to talk to other people, afraid to put themselves out there. Everything to them feels like a video game.
1:01:55
Unidentified speaker (UNKNOWN): But
1:01:55
Unidentified speaker (speaker_0): that's not real life. Real life is not a video game. So
1:02:01
Unidentified speaker (speaker_0): focusing on
1:02:04
Unidentified speaker (speaker_0): the intersection between you and other humans,
1:02:08
Unidentified speaker (UNKNOWN): and
1:02:09
Unidentified speaker (speaker_0): everything that that entails, that's where I would spend my time. That's how I advise young people building your brand, building awareness, writing, creating. It's so easy to create. Build stuff, put it out there.
1:02:25
Unidentified speaker (UNKNOWN): Honestly,
1:02:26
Unidentified speaker (speaker_0): I mean. as a
1:02:27
Unidentified speaker (UNKNOWN): as
1:02:28
Unidentified speaker (speaker_0): a venture investor, again, this is a very, very small industry, obviously. But I am continuously impressed by young people that are builders and that use these tools. They they're not engineers,
1:02:41
Unidentified speaker (UNKNOWN): but
1:02:41
Unidentified speaker (speaker_0): they have figured out how to use the tools to build really cool stuff.
1:02:45
Unidentified speaker (UNKNOWN): I
1:02:45
Unidentified speaker (speaker_0): mean, the opportunities are endless, but you need that
1:02:49
Unidentified speaker (UNKNOWN): kind of
1:02:50
Unidentified speaker (speaker_0): fearless. fearlessness of colliding with humans. You can't just sit in your room and code. You need to get out there.
1:03:05
Unidentified speaker (speaker_1): And so this is a question I haven't really asked much on the podcast, but I do ask a lot to guests where
1:03:14
Unidentified speaker (UNKNOWN): off
1:03:14
Unidentified speaker (speaker_1): the record
1:03:16
Unidentified speaker (speaker_1): I
1:03:18
Unidentified speaker (speaker_1): think about the conditions that are in place in the economy. You know, AI is this massive tailwind or headwind, or tailwind or headwind, depending on who you
1:03:27
Unidentified speaker (UNKNOWN): are.
1:03:29
Unidentified speaker (speaker_1): And
1:03:30
Unidentified speaker (UNKNOWN): I
1:03:30
Unidentified speaker (speaker_1): think one of the most important things that's come out of it, I mean that everyone can see, is it's not just the top 1% of talent that's dominant. dominating, it's not just the top 0.1% of talent that's dominating. It's the if you're say the 50,000, one of the 50,000 most talented people in the world, you are winning way
1:03:50
Unidentified speaker (speaker_1): more than everyone else. It's winner take all at a completely different level. Um and if you just
1:04:00
Unidentified speaker (speaker_1): plot that chart out
1:04:01
Unidentified speaker (UNKNOWN): of
1:04:02
Unidentified speaker (speaker_1): the amount of value. that's accruing to the top versus the bottom, right, and you compare this year to five years ago, five years ago to the previous to ten years ago, right, you just see it. Hold on, wealth is gonna
1:04:14
Unidentified speaker (UNKNOWN): get
1:04:15
Unidentified speaker (speaker_1): way more concentrated than we have ever seen. You know, we're gonna have, call it a five trillionaires in the next couple of years, maybe a couple more, or maybe those trillionaires becoming like even. you know ten trillion, right? Um
1:04:30
Unidentified speaker (speaker_1): what do you think that means for the social contract and for what government policy has to do to call it take care of people? And where do you think we evolve as a society, as a as an economy
1:04:45
Unidentified speaker (UNKNOWN): on
1:04:45
Unidentified speaker (speaker_1): the back of that?
1:04:47
Unidentified speaker (speaker_0): A discussion about the social contract?
1:04:50
Unidentified speaker (speaker_0): We could probably fill five podcasts up with that
1:04:55
Unidentified speaker (speaker_0): very important question.
1:04:58
Unidentified speaker (UNKNOWN): I
1:04:58
Unidentified speaker (speaker_0): will give you a few thoughts and then we can continue this another time.
1:05:05
Unidentified speaker (speaker_0): I believe that
1:05:09
Unidentified speaker (speaker_0): government has
1:05:13
Unidentified speaker (speaker_0): a role to play. in how we adapt to this abundance that's been created. I'm not sure that the way policymakers are thinking about it now is correct. I actually think that the social contract, in many ways,
1:05:33
Unidentified speaker (speaker_0): needs to be with the beneficiaries of
1:05:38
Unidentified speaker (speaker_0): this. outsized wealth and the rest of the societies in which they operate.
1:05:45
Unidentified speaker (UNKNOWN): And
1:05:45
Unidentified speaker (speaker_0): that this is it's less about taxing wealth and it's incentivizing these companies
1:05:55
Unidentified speaker (speaker_0): to
1:05:56
Unidentified speaker (speaker_0): educate, retool, invest in local economies, things that ultimately benefit. them by virtue of benefiting society. But when I think about the education complex,
1:06:11
Unidentified speaker (UNKNOWN): I
1:06:11
Unidentified speaker (speaker_0): mean this has been talked about a lot recently, just kind of like the uh
1:06:17
Unidentified speaker (speaker_0): if you have large government subsidy, than then supply and demand when you've got tremendous demand and constrained supply, what happens prices go up inexorably, education and healthcare probably being the two most notable examples, the systems need to be changed. They
1:06:37
Unidentified speaker (speaker_0): have to. It's not you can't have double-digit compounding of cost to customers and have this be democratic and have this benefit broader society. So it's going to take breaking some eggs and doing some really creative thinking. So if you really believe that
1:06:56
Unidentified speaker (UNKNOWN): these
1:06:56
Unidentified speaker (speaker_0): companies that are creating trillions of dollars of value and themselves are benefiting to the tons of trillions of dollars should be investing that money back, but doing so in a way where they can help reshape education.
1:07:10
Unidentified speaker (speaker_0): I don't see why we shouldn't be looking at at new models of education. I also think that the healthcare industry is completely broken. And when you look at large companies that have now basically self-insure,
1:07:24
Unidentified speaker (speaker_0): I'm not sure that that's not the answer to a lot of the problems because of how much money is spent on acute and end-of-life care and how little is spent on prevention. It's literally the inverse
1:07:35
Unidentified speaker (UNKNOWN): of
1:07:35
Unidentified speaker (speaker_0): the way it should be. But if you're
1:07:38
Unidentified speaker (UNKNOWN): a
1:07:39
Unidentified speaker (speaker_0): Walmart or you're an anthropic or you're a Costco, right, the rational thing to do is to self-insure
1:07:48
Unidentified speaker (speaker_0): and to invest
1:07:51
Unidentified speaker (speaker_0): most money into prevention while providing acute services at a fraction of the rate that the way society is currently pricing things, which is what skews healthcare premiums. So
1:08:05
Unidentified speaker (UNKNOWN): I
1:08:05
Unidentified speaker (speaker_0): think if we're able to kind of break out of the legacy mindset, which we're going to have to do to deal with this tectonic shift that AI is bringing, there is so much value being created. This is the perfect opportunity to take some of these sacred cows and get rid of them and say,
1:08:25
Unidentified speaker (UNKNOWN): let's
1:08:25
Unidentified speaker (speaker_0): talk about reinventing. healthcare, but from the standpoint of these companies being models of how to lower cost,
1:08:37
Unidentified speaker (speaker_0): improve quality of care, but doing so in a way that is market driven and then having money that is always there for market failures where certain segments of the society aren't part. of one of these companies that are doing this, but there's already this ethos and
1:08:57
Unidentified speaker (speaker_0): this
1:08:59
Unidentified speaker (speaker_0): kind of rational thinking that is divorced from PACs and all of that, where
1:09:09
Unidentified speaker (speaker_0): prevention, early child childcare, care, all of these things are provided on a much broader. basis because that reduces the cost to everybody.
1:09:18
Unidentified speaker (UNKNOWN): So
1:09:19
Unidentified speaker (speaker_0): to me, there's this golden, this golden opportunity that the social contract, rather than creating this win-loss, or you've made this money, we're going to take this all away from you. How about instead of taking it away, you need, you know, you need to develop programs that actually solve some of these societal problems like job training, education.
1:09:39
Unidentified speaker (speaker_0): and healthcare. I'd much rather see that.
1:09:42
Unidentified speaker (speaker_1): Definitely an exciting time to innovate on on
1:09:46
Unidentified speaker (speaker_1): on the social contract.
1:09:49
Unidentified speaker (speaker_3): And uh
1:09:51
Unidentified speaker (UNKNOWN): I
1:09:51
Unidentified speaker (speaker_3): think an exciting time to take risks. Thank you so much for coming on the pod. Roger, this is awesome.
1:09:57
Unidentified speaker (speaker_0): Thanks so much, Ethan. Appreciate you having me.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.

More from Odds on Open

All insights