Animal Spirits Podcast

"Talk Your Book: Public & Private Opportunities in AI"

Date: November 10, 2025
Hosts: Michael Batnick & Ben Carlson
Guest: Jonathan Schellen (COO, CraneShares)

Episode Overview

This episode explores the rapidly evolving landscape of artificial intelligence (AI) investing with a special focus on both public and private market opportunities. Michael and Ben interview Jonathan Schellen, COO of CraneShares, about the launch and unique structure of their AI-focused ETF, AJAX. The discussion centers on how the fund integrates private AI companies into its portfolio, the challenges and methods of valuing these entities, and how investors can thoughtfully capture the multi-decade AI opportunity—while navigating hype cycles and market bubbles.

Main Discussion Points & Insights

1. The Challenge and Opportunity of Investing in AI

[00:45-03:35]

• The "fast forward button" concept: The AI market is evolving rapidly, making it difficult to predict long-term winners and losers.
• Differentiating between public (Mag 7/big tech) and private (emerging, foundational) AI companies.
• AJAX ETF is designed to bridge this gap and offer exposure not just to giants, but key private players.

Quote:
"AI is no different in that it'd be fascinating to in 5, 7, 10 years, what does this look like in terms of not only just AI in our lives, but from a market structure perspective." – Michael Batnick [00:45]

2. The AJAX ETF: Bridging Public and Private AI

[02:04-04:14]

• CraneShares collaborated with AI-native investors (ex-Sequoia and Cruise Automation) to understand the AI landscape.
• AJAX is structured to capture the full AI value chain: hardware, infrastructure, and applications.
• Private foundational model companies (e.g., OpenAI, Anthropic, XAI, Perplexity) are structurally critical, but traditionally inaccessible to public investors.

Quote:
"If you wanted to build a proper AI portfolio, you have to move beyond Mag 7. You have to move beyond the Nasdaq and you have to move beyond public securities." – Jonathan Schellen [04:14]

3. Mechanics of Including Private Companies in an ETF

[05:50-08:40, 21:05-22:56]

• ETFs can allocate up to 15% to illiquid (private) securities; AJAX targets ~10% for risk management.
• Fair valuation is required continuously, using secondary market data and qualitative overlays.
• Direct participation in private company funding rounds (not just via SPVs) is key to accessing "cap table" economics.
• Example: AJAX acquired Anthropic shares at $60B valuation; later marked at $180B as the company raised another round.

Quote:
"There are now data sources, and we rely on them, that give you secondary market data with regularity... That gives us a pretty good proxy. And then we do a lot of our own kind of qualitative overlays." – Jonathan Schellen [06:44]

4. Global and Thematic Composition of the AJAX Fund

[08:40-11:26]

• 25% Mag 7, 50% non-Nasdaq companies, 23% hardware, 49% infrastructure, 30% applications.
• U.S. and international exposure; holdings include less obvious AI names like Duolingo, ServiceNow, Salesforce, and non-US server/data center company Nebious.
• AJAX is constructed for evolution—expect shifting allocations as the AI landscape matures.

Quote:
"You're getting private company exposure up to 15%. And you're getting a number of companies...in this application space that are going to be very important as the cycle progresses that we're accessing globally." – Jonathan Schellen [08:51]

5. AI's Physical Future: Robots and Embodied AI

[11:26-13:42]

• Discussion of the "glide path" from hardware to infrastructure to application.
• Huge anticipated growth in physical AI: humanoid robots, automation, and manufacturing.
• CraneShares launched KOID (humanoid/embodied AI ETF), focused on actuation tech, rare earths, and automation applications.

Quote:
"AI right now is what we think of as digital AI. ...Eventually AI becomes physical AI. It becomes embodied in things like robo taxis...humanoids...equipment manufacturing. That is very much the future." – Jonathan Schellen [12:03]

6. Is AI in a Bubble? Lessons from Tech Booms

[13:42-18:40]

• Comparison to railroad, dotcom, and fiber optics booms.
• This cycle’s capex is driven by world's largest, most operationally sound companies.
• Bubble signs: Not evident via classic markers yet; current AI investments often backed by strong revenues and real business models, unlike late ‘90s tech fads (e.g., Webvan).
• Basket approach is used to manage risk and uncertainty.
• Historical example: investing in Intel (hardware), Apple (infrastructure), and Amazon (applications) at $50B–$160B market caps in 2010—diversification across roles produced vastly different, and in some cases, enormous returns.

Quote:
"But the metrics that we had in other bubbles, like I remember eyeballs and page views and all that...which really wasn't tied to revenue...we don't see as much of that yet in the AI space. We're seeing tangible metrics." – Jonathan Schellen [17:40]

7. Picks, Shovels, and Applications: The Portfolio's Core

[19:55-21:05, 23:04-25:03]

• "Picks and shovels" (hardware/chip producers): Nvidia, SK Hynix, ASML, TSMC (~23% of portfolio).
• Growing emphasis on infrastructure and applications (Cloudflare, Palantir, Datadog, Nebious).
• AI Relevance and AI Readiness scores guide portfolio construction—how central AI is to revenue and strategy, and depth of AI penetration in operations.
• Turnover is quarterly, subtle, and focused on thoughtful, research-driven adjustments.

Quote:
"The easiest way to think about [AI Relevance and Readiness] is...how much sensitivity is there to AI in this company?...how much penetration has AI had within that organization?" – Jonathan Schellen [23:04]

8. Private Markets: Size, Access, and Turnover

[25:03-27:48]

• Many private AI firms would be among the Nasdaq 25 in size if public.
• AJAX is, per Schellen, the first ETF with direct shares of private AI companies—not only through SPVs.
• Preference for direct ownership, but SPVs or secondary market purchases considered when necessary.
• Secondary market provides liquidity when trimming is required due to overweighting (crossing 15% illiquidity cap).

Quote:
"If you took the five largest private companies globally, they would fall within the NASDAQ 25. So it's a huge opportunity cost not to invest in private firms." – Jonathan Schellen [25:32]

9. Screening and Future Portfolio Evolution

[28:53-31:25]

• AI Relevance is key for inclusion: e.g., XAI ranks high for LLM intelligence, Duolingo for deep AI integration.
• Application layer exposure is expected to dramatically increase in the next 5–10 years, both from emerging private companies and newer public entrants.
• AJAX’s future will pivot increasingly toward applications as infrastructure matures.

Quote:
"I think the major differences you're going to see is that our application sleeve...is going to be significantly larger...I also think you're going to see a lot of the newer private companies that we invest in are not going to be just in infrastructure...companies that may not even exist today." – Jonathan Schellen [31:25]

10. Liquidity & Exit Strategy for Private Holdings

[32:28-34:06]

• AJAX holds flexibility to sell private positions via tender offers or secondary markets if premiums arise or portfolio limits are breached.
• Primary goal is to hold until IPO, as most such companies are too large to be acquired.

Quote:
"We have that flexibility...the preferred route is for these companies to go public, and we are a beneficiary of what that means. But another approach is, if the company has a tender offer, we can participate...our desire is to hold them until the companies go public." – Jonathan Schellen [32:55]

Notable & Memorable Moments

• Anthropic’s Value Surge:
  "We invested around 60 billion. They were more recently marked at 180 billion." – Jonathan Schellen [06:44]

• ETF Innovation:
  "We're the first ETF that I'm aware of that has direct shares of private companies...not anti-SPV, but our pecking order is to be participating in these rounds." – Jonathan Schellen [27:48]

• On Private Market Opportunity:
  "Not having access to these private firms means you're really giving up a meaningful part of the universe." – Jonathan Schellen [25:46]

Key Takeaways for Investors

• AJAX is engineered to capture the long cycle of AI by balancing hardware, infrastructure, and especially application layer exposure as the industry matures.
• It uniquely accesses private market leaders by participating in primary rounds, not just secondaries or SPVs.
• The fund’s construction is deliberately global and broad, focused on both present giants and future disruptors—public and private.
• AI investing will likely replicate historical tech cycles: "picks and shovels" win early, applications win big over time.
• Stringent valuation and monitoring processes exist for private holdings, with an emphasis on disciplined rebalancing and exit management.
• The ETF’s mix is expected to shift significantly within 5–10 years as AI applications dominate.

Timestamps of Important Segments

• [00:45] — AI’s future: winners, losers, and investing challenge
• [04:14] — Rationale for including private AI companies
• [05:50] — ETF rules & mechanics for private holdings
• [06:44] — Secondary market valuation & direct cap table access
• [08:51] — Global reach & breakdown of AJAX portfolio
• [12:03] — The rise of embodied/physical AI (robots, KOID ETF)
• [14:36] — AI bubble, historical analogs, and risk management
• [17:40] — Tangible business metrics vs. bubble signs
• [19:55] — "Picks and shovels"—what hardware AJAX buys
• [23:04] — AI Relevance/Readiness scoring
• [25:32] — Size of private AI companies vs. public giants
• [27:48] — Direct private share ownership (not via SPV)
• [31:25] — Future shift to AI applications in the portfolio
• [32:55] — Exit strategy for private holdings

Further Information

• Learn More: kraneshares.com/ajax and kraneshares.com/coid (for humanoid/robotics ETF)