Modern Capital: The Private Markets Podcast

Henry Ward argues private markets are structurally inverted and predicts AI will drive consolidation or fragmentation by 2040

57 min Listen to the original Episode page
Private MarketsVenture CapitalPrivate EquityFund AdministrationArtificial IntelligenceMarket StructureRegulationFinTech

Summary

Henry Ward describes private markets as structurally inverted compared to public markets, where liquidity is viewed as a bug and selling large blocks is easier than small ones. He explains that Carta aims to replace fragmented spreadsheets with an integrated ERP for private capital, noting that private companies prefer illiquidity to maintain control and avoid competition for primary capital. Ward predicts a continued decline in public listings, with only top companies remaining public while others stay private to avoid regulation. He highlights the 'retailization' of private markets, which requires daily NAV calculations that current manual processes cannot support. Ward also details Carta's rapid adoption of AI, with 90% of journal entries now handled by agents, and discusses the acquisition of Avantia to automate legal workflows. He concludes by predicting that AI will either consolidate private capital into a few dominant firms or fragment it into millions of small funds by 2040.

Key takeaways

  1. 1
    Private funds operate on a fragmented stack of 15+ providers and spreadsheets, lacking a unified ERP equivalent like NetSuite for operating companies.

    Unidentified speaker (speaker_0) · 2:35

  2. 2
    Venture capital companies can command premium valuations because they are the sole sellers in the market; allowing secondary sales would turn shareholders into competitors.

    Unidentified speaker (speaker_0) · 11:48

  3. 3
    Companies stay private specifically to avoid the regulation, transparency, and liquidity required by public markets, preferring to pick their own shareholders.

    Unidentified speaker (speaker_0) · 14:20

Episode sections

  1. 0:00 – 1:35

    The Inversion of Private Markets

    Henry Ward describes private markets as inverted compared to public markets, where selling large blocks is easier than small ones and liquidity is a bug rather than a feature. He introduces Carta's vision as the ERP for private capital, aiming to replace fragmented spreadsheets with integrated software.

  2. 1:35 – 8:22

    Carta's Evolution from Cap Tables to Fund Accounting

    Ward explains that Carta started by digitizing cap tables to create a single source of truth, similar to how ERPs like NetSuite run operating companies. He notes that private funds lack such a platform, relying instead on 15+ providers, and Carta is building the equivalent for fund operations.

  3. 8:22 – 10:29

    The Failure of the Private Stock Market Thesis

    Ward recounts his decade-long attempt to build a private stock market, which failed because the problem is structural, not technological. He argues that private companies prefer illiquidity to avoid competition for primary capital and maintain control over their cap table.

  4. 10:29 – 15:14

    Market Structure and the Case for Staying Private

    He asserts that price discovery and liquidity are detrimental to private companies, which prefer to be the sole seller in auctions. He believes that if private markets became liquid, they would effectively become public markets, losing their primary value proposition of control and privacy.

  5. 15:14 – 24:46

    Public vs. Private Company Trends

    Ward predicts that the number of public listings will continue to decline, with only the top 20-30 companies remaining public. He argues that mediocre companies will stay private, which he views as a positive selection mechanism for public investors, despite concerns about access to winners like SpaceX.

  6. 24:46 – 26:46

    Retailization and the Need for Daily NAVs

    Ward highlights the 'retailization' of private equity and credit, where funds must move from quarterly to monthly or daily NAV calculations to serve retail investors. He states that current manual processes cannot support this frequency, creating a demand for software solutions.

  7. 26:46 – 29:22

    The 1933 Act and the 'People Business' of Private Markets

    He traces the lack of a central exchange in private markets to the 1933 Securities Act, which prohibits soliciting strangers. This creates a 'people business' with negative selection bias for outsiders, whereas funds can mitigate this through positive selection and retailization via interval funds.

  8. 29:22 – 30:24

    Transaction Costs and Operational Inefficiencies

    Speaker_0 explains that private market transaction costs are high because they rely on human time for networking and deal sourcing, which cannot be digitized like public markets. He notes that management fees (the 'two' in two-and-twenty) cover these operational costs, which are substantial over a fund's life.

  9. 30:24 – 33:03

    AI Disruption in Fund Administration

    Speaker_2 asks about AI's role in fund administration, citing incumbent resistance. Speaker_0 compares current human-heavy processes to NASA's 'human calculators' before electronic calculators, arguing that technology should replace humans for complex math, not the other way around.

  10. 33:03 – 34:01

    Speed of AI Adoption and Carta's Current Metrics

    Speaker_0 highlights the rapid adoption of AI at Carta, stating that by the time of recording, 90% of journal entries and 200,000 monthly customer emails are handled by AI agents, a significant shift from zero AI usage in January of that year.

  11. 34:01 – 39:45

    Generational Shift and Strategic Value of Accounting

    Speaker_0 argues that while older CFOs view accounting as a static back-office function, younger CFOs see it as a strategic tool. He claims that solving accounting via AI allows GPs to build planning and strategy on top of accurate, real-time data, similar to how CEOs use ERPs like NetSuite.

  12. 39:45 – 44:27

    Data Ownership and Carta's Role as Custodian

    Speaker_2 asks who owns private market data. Speaker_0 asserts that Carta acts as a custodian, not an owner, of captable data. He argues that being a software company precludes selling data, unlike data companies like Prequin that rely on surveys rather than a system of record.

  13. 44:27 – 46:53

    Acquisition of Avantia and Legal Automation

    Speaker_0 details the acquisition of Avantia, a law firm with ~200 lawyers. He explains that high-volume, low-complexity legal work like NDAs is now being automated by AI agents, reducing turnaround time from hours to minutes and shifting billing from hourly to output-based.

  14. 46:53 – 56:38

    Standardization vs. AI Flexibility in Legal Docs

    Speaker_2 asks if AI will drive standardization in private market docs like LPAs. Speaker_0 suggests AI allows for 'snowflake' customization without the operational nightmare, potentially reducing the need for standardization. He also notes that lawyers have a perverse incentive to avoid standardization to maintain billable hours.

Claims worth verifying

In private companies, it is easier to sell $100 million of stock than a single share.

Unidentified speaker (speaker_0) ·0:00

How to verify: Verify by analyzing private market transaction data from sources like Preqin or Carta’s own market reports, which show that secondary transactions are predominantly large-block trades. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Venture capital price discovery and liquidity are a bug, not a feature.

Unidentified speaker (speaker_0) ·0:19

How to verify: Verify by reviewing venture capital term sheets and shareholder agreements, which often include right of first refusal (ROFR) and co-sale rights that restrict liquidity to maintain control. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Private funds use 15 different providers for their stack.

Unidentified speaker (speaker_0) ·2:55

How to verify: Verify by surveying fund administrators or reviewing industry reports on private fund tech stacks, such as those from KPMG or PwC, which detail the fragmented nature of fund admin. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Carta’s AI agents book about 90% of journal entries.

Unidentified speaker (UNKNOWN) ·1:17

How to verify: Verify by requesting Carta’s technical whitepapers or case studies detailing their AI implementation in fund accounting, or by checking their public product updates. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

The 1933 Securities Act prohibits selling stock to people you don't know.

Unidentified speaker (speaker_0) ·27:35

How to verify: Verify by reading Section 5 of the Securities Act of 2026 and the definition of 'public offering' under SEC rules, which restrict general solicitation for unregistered securities. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Primary source

There are fewer than 4,000 public companies in the US, down from 8,000 in the 90s.

Unidentified speaker (speaker_2) ·15:13

How to verify: Verify by checking the NYSE or NASDAQ annual listing statistics, which track the number of listed companies over time. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

SpaceX has only 2,000 shareholders with all their SPVs.

Unidentified speaker (speaker_2) ·16:36

How to verify: Verify by reviewing SpaceX’s cap table disclosures in SEC filings (if any) or credible reports from financial media like Bloomberg or Reuters detailing shareholder counts. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Primary source

Private equity firms are moving to monthly or weekly NAVs due to retailization.

Unidentified speaker (speaker_0) ·25:43

How to verify: Verify by examining prospectuses for interval funds or semi-liquid private equity funds, which specify valuation frequencies, and comparing them to traditional quarterly reports. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Carta raised a billion dollars on the vision of a cap table stock market.

Unidentified speaker (speaker_0) ·21:09

How to verify: Verify by reviewing Carta’s historical SEC filings (Form D or Form D-A) or press releases from 2026-2026 detailing funding rounds and stated use of proceeds. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Primary source

The best companies will stay public, while average companies will go private.

Unidentified speaker (speaker_0) ·15:42

How to verify: Verify by analyzing IPO performance data and delisting trends, comparing the market cap and performance of remaining public companies against those that went private. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Speaker_0 states that in private markets, $100 million of stock is easier to sell than a single share, explaining why a private stock market is hard to create.

Unidentified speaker (speaker_0) ·29:22

How to verify: This is a conceptual claim about liquidity dynamics in private vs. public markets; verify via academic literature on private market liquidity or Carta's public blog posts on this topic. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Speaker_0 claims that at the start of the year (January), Carta had zero AI doing fund accounting, but eight months later, 90% of journal entries booked on Carta are done by AI agents.

Unidentified speaker (speaker_0) ·33:03

How to verify: Verify via Carta's official blog or press releases from January 2026 to September 2026 detailing their AI adoption milestones. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Speaker_0 states that Carta processes about 200,000 individual emails from customers monthly for fund accounting services.

Unidentified speaker (speaker_0) ·33:23

How to verify: Verify via Carta's operational disclosures, investor presentations, or press releases from 2026 detailing their volume metrics. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Speaker_0 asserts that before electronic calculators, NASA employed 'human calculators' in rooms of cubicles to calculate moon orbits and rocket trajectories.

Unidentified speaker (speaker_0) ·31:12

How to verify: Verify via historical records from NASA or biographies of the 'human computers' (e.g., Margot Lee Shetterly's 'Hidden Figures'). For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Speaker_0 claims that Avantia, acquired by Carta, is a law firm with about 200 lawyers and paralegals, started by a former private equity GC.

Unidentified speaker (speaker_0) ·44:46

How to verify: Verify via Carta's press release announcing the acquisition of Avantia or Avantia's own website/about page. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Speaker_0 states that with AI, NDA markups are now done in two minutes instead of two hours at Carta.

Unidentified speaker (speaker_0) ·46:25

How to verify: Verify via Carta's case studies or product documentation for their AI legal automation tools. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Primary source

Speaker_0 claims that Prequin and similar data companies get data via survey and do not own a system of truth.

Unidentified speaker (speaker_0) ·41:20

How to verify: Verify via Prequin's methodology disclosures or public statements about their data collection processes. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Speaker_0 predicts that by 2040, private capital will either consolidate into two firms (like KKR and Thoma Bravo) or fragment into a million funds.

Unidentified speaker (UNKNOWN) ·51:01

How to verify: This is a forward-looking opinion/forecast; verify by checking if this prediction is cited in future industry reports or if the market structure aligns with this dichotomy by 2026. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Speaker_0 states that Carta bills for legal services by output/project (e.g., per NDA) rather than by the hour.

Unidentified speaker (speaker_0) ·49:03

How to verify: Verify via Avantia's or Carta's legal service pricing pages or client engagement letters. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Primary source

Speaker_0 claims that Opus V (likely referring to OpenAI's o1 model) was the inflection point for AI acceptance in January.

Unidentified speaker (speaker_0) ·38:29

How to verify: Verify via OpenAI's release notes for the o1 model (released Sept 2026, but speaker may be referring to a specific update or misremembering the date/model name; verify OpenAI's product timeline). 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

  • Review the text of the 1933 Securities Act regarding solicitation rules and exemptions to verify the 'people business' constraint.
  • Check Carta's latest S-1 or 10-K filings for revenue breakdowns between software subscriptions and data services to assess the 'custodian vs. data seller' claim.
  • Analyze the acquisition terms of Avantia to evaluate the integration of legal automation and the shift from hourly to output-based billing.
  • Compare Prequin's data collection methodology (survey-based) with Carta's system-of-record approach to understand potential data biases.
  • Review the performance data of interval funds and daily NAV private products to assess the feasibility of the 'retailization' trend.
  • Examine the public listing trends from S&P Global or Bloomberg to verify the claim that only the top 20-30 companies will remain public.
  • Look for counter-arguments from private equity firms regarding the operational costs of daily NAV calculations versus quarterly reporting.
  • Verify the AI adoption metrics cited by Carta against independent third-party audits or case studies if available.
Full transcriptexpand
0:00
Unidentified speaker (speaker_0): The private markets, it's like that Netflix show. It's like the upside down. Everything's inverted. In the public markets, it's easier to sell one share than a billion dollars with a stock. In private companies, it's the opposite. It's easier to sell a hundred million dollars of a private company than it is to sell one chair. We generally believe price discovery and liquidity is a feature of
0:19
Unidentified speaker (speaker_0): markets. And venture capital, specifically private assets, is not a feature. Price discovery and liquidity is a bug. And that's what people miss about this problem.
0:34
Unidentified speaker (UNKNOWN): This
0:34
Mark Andrew: is Modern Capital. Conversations with the people accelerating the next generation of private markets.
0:42
Mark Andrew: I'm Mark Andrew.
0:43
Unidentified speaker (UNKNOWN): The
0:44
Mark Andrew: infrastructure of private markets is taking shape. Let's unpack
0:48
Unidentified speaker (UNKNOWN): it.
0:55
Mark Andrew: Henry Ward knows private markets. He's thought about them as deeply as pretty much anyone. And he's reached conclusions much of the industry hasn't. Henry built Carta. It started with cap tables. Now it's moving into fund To accounting for venture, private equity, and private credit.
1:12
Unidentified speaker (UNKNOWN): And
1:12
Mark Andrew: he's building the system a fund runs on. accounting, reporting, legal,
1:17
Unidentified speaker (UNKNOWN): with
1:17
Mark Andrew: AI doing much of the work.
1:19
Unidentified speaker (UNKNOWN): In
1:19
Mark Andrew: this conversation, we cover Carta's and Henry's journey through the varied hills and lanes of private markets, and where that road is leading next. Please enjoy this conversation with one of the pioneers of private markets technology, Henry Ward.
1:35
Unidentified speaker (speaker_0): Henry Ward, welcome to the Modern Capital Podcast. Great to be here, Mark. Thanks for having me. Henry, I can think of really no one that's more interesting to speak to about, you know, the world of private capital, the background in private shares, the future of private markets than yourself. So I'm thrilled to welcome you today. And I thought we could start with a fairly big picture look. You know,
1:55
Unidentified speaker (speaker_0): you've described Carta as the ERP of private capital, or of that as a vision for what you're trying to build. For those who aren't deeply familiar, with what an ERP is and what the vision is, let's start there. What are you building a carta? Yeah, so ERP is an old word from the old days of the 90s when SAP financials figured out how to build
2:15
Unidentified speaker (speaker_0): software that CEOs and CFOs could use to run their businesses. And so, you know, there was this era in the 80s, 90s, and early 2000s where every CEO and CFO of a Fortune 1000 company was emblematic. implementing an ERP, enterprise resource planning software. But it was really how do you run the business? And we have one at Carta. You know, our CFO uses NetSuite, which is
2:35
Unidentified speaker (speaker_0): sort of the standard mid-market ERP. Well, we learned as we got into fund operations, whether it's private equity, private credit, venture capital, that there is no equivalent of Net Suite or SAP financials for private funds. They're running things on spreadsheets and outsource outsourced the accountants, and they have 15 different providers that do different pieces of the stack
2:55
Unidentified speaker (speaker_0): for them. But there isn't one platform where they can run these funds on. And we said, hey, that's something the world should have. And one of our arguments that venture capital had difficulty scaling before Carta existed was just managing the cap captables, tables, which is like the ERP of shareholders was very hard at the time.
3:16
Unidentified speaker (speaker_0): And we'd like to take a little bit of credit for thinking like because of cap tables, we can increase the number of companies that existed in the world. Same on ERP is that if we can build a platform for funds to run funds better, more efficiently, smarter, there'll be more funds in the world. And we think there should be more capital in the world uh investing in private assets. And that's what we've been working on for the last decade.
3:37
Unidentified speaker (UNKNOWN): So
3:37
Unidentified speaker (speaker_2): we're gonna come back to that in depth. But I want to step back a bit and learn more about Henry as a as an individual. Where did you grow up? How did you come across starting a business focused on captable management?
3:47
Unidentified speaker (speaker_0): Yeah, I'm a Michigan boy, grew up in the Midwest, went to University of Michigan, got a job down down in Texas as a sales engineer. So I was always kind of straddling as a computer scientist by training, but but always sort of did I did a lot of the demos and built a lot of software. as prototypes and you know in the 2000s for large insurance carriers. I worked for a few companies that
4:07
Unidentified speaker (speaker_0): sold a software to insurance companies. And then you know fast forward ended up in France for for business school. And then my at the time fiancee, you know, got a job in Palo Alto. So I followed her to Palo Alto. I didn't have a job, didn't know anybody, decided to start a company like one does when you show up in Mountain View as
4:27
Unidentified speaker (speaker_0): a as a coder, but you know, not not knowing anybody. My first company died. It was like an investment. It was like Wealthfront. It was a robo advisor, wealthfront or betterment. But the they just did a better job than I did. My my company failed while theirs succeeded. But then out of the ashes of that, I worked with an investor on this idea for the second one, which is
4:47
Unidentified speaker (speaker_0): hey,
4:48
Unidentified speaker (UNKNOWN): uh,
4:48
Unidentified speaker (speaker_0): this cap table problem is a big problem. You
4:50
Unidentified speaker (UNKNOWN): know,
4:50
Unidentified speaker (speaker_0): it's all managed in Excel. It's all managed, you know, in paper stock certificates. We should digitize paper stock certificates. And at that time, you know, this was in the mid-2010s you know, 2012, 13, 5, 14, 15, PayPal was picking up a lot of this financial infrastructure was starting to take off. And we really thought of it as like
5:10
Unidentified speaker (speaker_0): PayPal but for equity. So can you transfer shares via email? And that was our starting point. And that turned into a cap table product, which turned into a fund accounting product, which turned into a car to law, which we now do, and a bunch of different things.
5:24
Unidentified speaker (speaker_2): And talk more about that problem that you were solving. Okay. So you have a bunch of different investors and the cap tables get distorted, right? It's being traded through paper. People don't know what they own anymore. When something gets transferred or gets bought or goes public, you may have a an a ledger that that's wrong, or you may have different owners who have different versions of the same truth. So you're helping to solve a common
5:44
Unidentified speaker (speaker_2): truth for private shares of companies.
5:46
Unidentified speaker (speaker_0): Correct. And I I like that you said sort of common truths. You know, the technical term we often use is system of record, but it's more fundamental to say like what is the common truth that we all rely on. Because pre-CARTA, there were three versions of the cap table. The company held their version, the law. firm held their version, and the investors would hold their version. There may be
6:06
Unidentified speaker (speaker_0): multiple investors that said multiple versions. And then when it came time to do a transaction, that's when you had to sort of make sure reconcile all these different cap tables. And it was an incredibly time-consuming and expensive process to do that. And so what we said was, hey, instead of cap tables being an adversarial process of multiple people owning different records of it, we're gonna put it in one
6:26
Unidentified speaker (speaker_0): place. and then we're gonna have everybody log in and verify their amount. So the company will log in and say, hey, we issued these shares for 100 or 100, we issued 100 shares of our stock to this person. That person would then receive them via email and they'd log in and they'd confirm I own a hundred shares. And then you had this this bilateral confirmation between and then the lawyers
6:46
Unidentified speaker (speaker_0): would also confirm, hey, we did all the documents correctly. So that these three parties would come into the same database and say, yes, this hundred shares is correct. And that's how you knew the cap table was always correct. And that was really the unlock. And it seems like an obvious idea today. It wasn't obvious back then. It was very odd, this idea of multiplayer systems of
7:06
Unidentified speaker (speaker_0): record. And I I tell that story because if you look at everything Carta does, it's all in that taking what used to be adversarial processing. processes, like where two people had different versions of the truth and then have to reconcile that and saying, hey, we're actually gonna provide the system of truth for multiple parties. So if you look at fund accounting, fund
7:25
Unidentified speaker (speaker_0): administration, that is a multiple truth problem. The LPs have their version of the truth, the GPs have their version of the truth, the fund accountants have their version of the truth, and everybody has to reconcile this. They do it quarterly to make sure everything's right. And we built some software that allows this stuff to be tracked in one place and everyone can go verify their amount.
7:43
Unidentified speaker (speaker_2): So one can see the connected line from the focus on private shares to begin with to private funds and the broader capital connected vision you have. Of course, what's within these private funds is private shares of companies. And so often finding a common truth of that is a key piece of it. But talk to us a bit. You've had an amazing journey there
8:01
Unidentified speaker (UNKNOWN): from
8:01
Unidentified speaker (speaker_2): sort of the founding of Card. Carta and then into this broader discussion that was Carta X for a while about the vision of private shares and the burgeoning still today, incredibly burgeoning trade and pre-IPO shares. Can you talk to us about that vision of that problem you had, where it is today? Because you were very eloquent for years there about how to better trade private company
8:21
Unidentified speaker (speaker_2): shares.
8:22
Unidentified speaker (speaker_0): Yeah. So the original vision of Carta CARDA was to build the we at the time we called it Nasdaq for private markets, or the NYSC for product private markets. It was the the stock stock market that didn't exist that shouldn't. And it was for private companies. It's not a new idea. A million people have had this idea.
8:45
Unidentified speaker (speaker_0): And it worked. And our thesis was in 2015, the reason it hadn't worked is that everybody who tried to build the stock market didn't do the hard part first, which is digitizing the shares. So even in you know the New York Stock Exchange, it used to be a bunch of traders on the floor, you know, with pieces of paper trying to do trading. And it wasn't until they built a depository trust
9:04
Unidentified speaker (speaker_0): corporation where the government came in and said, Hey, we got to digitize all these shares. And they created the master database that, you know, if you sell 100 shares at Google, that's the database that gets updated. And then you could build the exchange, electronic exchanges on top of the central clearinghouse. There was no equivalent of a clearinghouse in the private markets. And that's what we decided to
9:24
Unidentified speaker (speaker_0): build via the cap table. So if you could digitize all these shares and now you had the cap table, you are now not just the source of truth, but you were also the transaction platform. You were the clearinghouse. Because if somebody wanted to transfer shares from A to B, we had the database. that would get updated for. And that would unlock liquidity. It didn't work.
9:44
Unidentified speaker (speaker_0): There's a lot of my economic reasons why I it didn't work, and I don't think it ever will. But that was the original thesis. But what we found was the infrastructure we had would built, trying to build trading, allowed us to actually go do more with cap tables, but then go into private funds, private equity, private credit. Because the infrastructure that I allowed the
10:03
Unidentified speaker (speaker_0): digitization of private assets that was supposed to unlock liquidity didn't unlock liquidity, but it unlocked all these other assets that we could digitize.
10:12
Unidentified speaker (UNKNOWN): So
10:12
Unidentified speaker (speaker_2): you've spoken at length, I think, about why it didn't work and other podcasts and so forth. But can you just give us a bit of a fraction of those economic reasons it didn't work now? And then provide us, yeah, some maybe more novel insights. We were chatting a
10:23
Unidentified speaker (UNKNOWN): bit
10:23
Unidentified speaker (speaker_2): offline, Henry, about some of the pieces you think that maybe aren't that they're in there the right record, but you're thinking on this private shares dialogue?
10:29
Unidentified speaker (speaker_0): Yeah. I think one of the things that a a lot of people miss, they people still come to me and like, hey, you know what you should do? You should go build a private stock market for companies. And the problem is it's not a technology problem. A lot of companies too are like, hey, we'll do it on blockchain, all these things. And I think what people are missing is it's the reason we don't have a private stock market is not a
10:49
Unidentified speaker (speaker_0): technology problem. It's a market structure problem. The private stock, people think it's, oh, it's it's equity. So it should work just like the public markets. What they what they miss is the private markets, it's like that Netflix show. It's like the upside down compared to public. Like everything's inverted in the private markets. So for example, in the public
11:09
Unidentified speaker (speaker_0): markets, it's easier to sell one share than, you know, a billion shares or, you know, a billion dollar. dollars with a stock. In private companies, it's the opposite. It's easier to sell a hundred million dollars of a private company than it is to sell one share. In the public markets, the price is set by the last buyer, you know, the last person to do a train. And the private is set by the first.
11:29
Unidentified speaker (speaker_0): You know, in the public markets, there is no comp competition for primary capital. Facebook just runs its business and then all the trading, all the stock transactions are done at external Facebook as secondary trades. In private, all by sort of definition, all venture back companies are always raising more money and trying
11:48
Unidentified speaker (speaker_0): to sell their stock to investors. And so if they allow shareholders to sell, the shareholders start to be competition. And one of the reasons venture gets get such premium prices, right? Everyone's like, well, oh my gosh, how did so-and-so so and so company raise it that at value? valuation? That's crazy. One of the reasons they they can do that is cause they're the only seller in the market because
12:08
Unidentified speaker (speaker_0): nobody else is allowed to sell. And when we did that, we ran an auction. One seller, many buyers. So by definition, you can top tick the price to get the money on your balance sheet. And so there's all these like market structure problems where what a lot of people miss is they is you generally believe, we generally believe price discovery and liquidity is
12:28
Unidentified speaker (speaker_0): a feature of markets. We believe that. And venture capital, specifically private assets,
12:34
Unidentified speaker (UNKNOWN): and
12:35
Unidentified speaker (speaker_0): is not a feature. Price discovery and liquidity is a bug. And that's that's what people miss about this problem.
12:40
Unidentified speaker (speaker_2): That is very interesting. Okay. So we are at this moment when PIPO shares are in high demand. There is a lot of focus among national exchanges, not just in the United States but around the world. to create a more of a share of private markets. For example, in London there's something called Pisces. The Australians have competing stock exchange to their Australian stock exchange. I'm curious what you think of when I if I
13:00
Unidentified speaker (speaker_2): was to ask, like what does this look like at maturity? It feels like the tr the exchange in private shares remains nascent to to the points you've made. It doesn't have very much market structure. But assume we reach maturity where there's a functioning exchange for private shares alongside healthy, vibrant public market exchanges. What does the market structure of those private shares look like?
13:20
Unidentified speaker (speaker_0): I think it's very difficult because the public markets, NISE and NASDAQ, are pretty tough competition. You have to build a better stock market than they have. And they've been doing this for a long time. And so w one way to ask this is if you're a private company and you want liquidity, is a private company stock market a
13:40
Unidentified speaker (speaker_0): better product for you than the public markets? And I I think the public markets, for companies that want liquidity and real-time trading, public markets do that pretty well. So then you might say, well, why does somebody want to stay private? Well, the reason you want to stay private is you don't want liquidity. That is the feature of private markets. And by not wanting liquidity, that also means you don't
14:00
Unidentified speaker (speaker_0): want regulation, because if there's gonna going be to liquidity there's got to be some level of regulation to protect investors because there's a lot of secondary trading. So you want liquidity, you need regulation, you need uh information transparency, you need information symmetry, you need all these things that that exist in public market infrastructure to allow for liquidity. And so if you
14:20
Unidentified speaker (speaker_0): if you make private markets liquid, you're sort of taking away the only reason to be private. The reason we want to stay private is because we don't want liquidity. We don't want real time trading. We don't want, we want to cat to pick who's on our cap table versus just the random day trader in Idaho, Idaho is now on my cap table. And so I I actually don't believe if
14:40
Unidentified speaker (speaker_0): it's true, if there's a true statement, if you sort of say, Henry, I can see the future, there is a world where there is a liquid private market, I would say, oh, I think that's just the public market, but they called it private.
14:53
Unidentified speaker (UNKNOWN): Do
14:53
Unidentified speaker (speaker_2): you think so we talk a lot about how there used to be 8,000 public companies in the United States in the 90s, today there are less than four, something like 3,700? Do you think that trend continues? Are we going to see fewer and fewer? And then you have this argument that public markets are so concentrated, partly because of that reason, that it's harder to go public for companies at an early stage. Does that trend continue, or what does that
15:13
Unidentified speaker (speaker_2): look like?
15:14
Unidentified speaker (speaker_0): I think so. I think the number of listings, it it's interesting, the public markets, the listings go down, number of companies and the public markets.
15:22
Unidentified speaker (UNKNOWN): But
15:22
Unidentified speaker (speaker_0): the total market cap keeps going up and is really driven by the top, not just maybe top maga seven, but the top 20 or 30 companies because companies at Power Long, law. they're just the best companies are 10,000 times better than the average. And so the best companies will stay in the public markets. Mediocre
15:42
Unidentified speaker (speaker_0): and average companies will go into the to the private markets. It's this sort of interesting thing where we often think like, oh, venture has the best companies and the old, you know, age companies go public. But it's actually the opposite. The best companies in the world will be public. And then the average companies will be private. And that will just continue to be, continue to be true, which if you
16:02
Unidentified speaker (speaker_0): you're a regulator, you kind of want, right? Because public markets allow access to everyday people to access securities. You want them to only be able to access the best ones. And then then the average companies will go into the private asset class.
16:16
Unidentified speaker (speaker_2): I guess one of the places I'd push on, you'd been as eloquent as anyone, maybe more so than anyone, about building the ownership economy. And now you have a situation where all of these companies are going public quite a bit later than. for example, when Amazon or Apple went public. And in those days, you were able, as a general investor, to be able to share in the upside and growth of that company. Today SpaceX goes public
16:36
Unidentified speaker (speaker_2): quite late, and so only 2,000 shareholders with all their SPVs are those who are getting benefit to the early stages of growth of that company. That seems like a societal problem. What is your thought on that?
16:46
Unidentified speaker (UNKNOWN): Yeah,
16:46
Unidentified speaker (speaker_0): it it's interesting. I I go to DC twice a year and then go meet with the House and Senate and SEC and we talk about these issues. And the the general view is yeah, you know, it's a big problem that average American can't access SpaceX, you know, before they go public and all the wealth gets concentrated by the people who are able to access it. I think
17:06
Unidentified speaker (speaker_0): what gets missed in the conversation is there is a positive selection bias on it, right? We're all focused on the SpaceX. What we don't talk about is the hundred other companies that people invested in lost all their money that that weren't SpaceX.
17:22
Unidentified speaker (speaker_2): If you get general investors doing that, it's not going to be as useful.
17:25
Unidentified speaker (speaker_0): That's right. And there is this question of like, hey, I think, you know, I have this view. Everybody is right. You know, smart people are on both sides of this argument. It's which aperture? How How do you want to think about this problem? The Republicans are right. Like, people are missing out on space. SpaceX 100%. But the Democrats are also right. In exchange for missing out on SpaceX, we're protecting
17:45
Unidentified speaker (speaker_0): from Americans, we're tech protecting Americans from investing in the not SpaceXs that don't go anywhere. And there is tons and tons of those. You just don't read about them. You don't see them because people are successful, you know, publicly and fail privately. And I I think that's the the challenge. And so by definition, the public markets is sort
18:05
Unidentified speaker (speaker_0): of the selection criteria that says these are the companies that are good enough to invest in. And you're right, Americans might miss the first 10 or 15 years of growth, but hopefully they get the next 50. And you know, we can certainly debate a lot about like, hey, is it a good selection mechanism? Um We do have the Enrons and Comcast, Comcasts, you know, those do happen. So
18:25
Unidentified speaker (speaker_0): is it a good selection and regular and regulatory framework to make sure Americans get good companies? Is it overprotective? Should there be more SpaceXs that we don't get access to that we should? And that a lot of Republicans are working on like, hey, make it easier, make it less selective to get these best companies in the public markets. So I think these are all worthy conversations, but I I
18:45
Unidentified speaker (speaker_0): I don't think it's just, oh well, Americans missed messed out on SpaceX, so the system is broken. You you have to look at the other side. too.
18:51
Unidentified speaker (speaker_2): This has been a fantastic introduction. You know, it and it's cont as context even for you, Henry. When I I've called this company, I run the private markets forum. When you're in the Valley, private markets means pre-IPO shares. But most of our work, I would say 90% of our work, has been much more about private equity and private credit in particular and improving the information infrastructure for funds and their distribution
19:11
Unidentified speaker (speaker_2): into the wealth channel. We have looked hard at doing a private. shares type of summit, and thus far haven't really launched it, partly for all the reasons you've just outlined here, is it feels like a conversation or an area that's still taking shape, still categorizing. I haven't figured out quite how to how to structure it as a conference, for example, versus something like building the infrastructure underlying
19:31
Unidentified speaker (speaker_2): private wealth, where there's a richer conversation about what's taking shape. And it's just an ironic kind of piece when most of the valley thinks private markets is. just that. But for us in the private markets forum, it's really much more so about the funds aspect and the information infrastructure underlying asset classes like private credit. So before we move on to that funds world, which is where Carta is
19:51
Unidentified speaker (speaker_2): really taking your business, I want to just touch on and going back to that, you you were so eloquent. I mean, to your point, it was like the biggest problem many people articulate around a private shares exchange. You've said it didn't work. How did that feel? What was that like, Henry, over that time? And then kind of repurposing Carta for its next phase after that. I'm really curious.
20:10
Unidentified speaker (speaker_0): Yeah. Oh, it was really hard. It's like fail privately. Carta X was a fail publicly phenomenon. And everybody was rooting for us to do it and wanted us to do it. You know, we tried. I spent almost a decade of my life working on different arc incarnations of it. The the last one that failed was was the one everybody, you know, knew most because we publicly said we're not even
20:30
Unidentified speaker (speaker_0): gonna start trying anymore. But I had multiple a attempts at it fail that just didn't work. And then I re retried it again with a different approach. And so it was, you know, almost a decade of my life and and hundreds of millions of dollars that we spent trying to make this work. And then to go back to the employees and investors and not not just say like it failed, but like we're not even gonna
20:49
Unidentified speaker (speaker_0): try anymore. And then we're gonna go do this other thing called the ERP for private capital, which is accounting, not nearly as cool as like a stock market. And to get both the investors and all the investors, you know, we'd raised a billion dollars on the on the vision of it. So all the investors like, wait, but we invested in a cap table
21:09
Unidentified speaker (speaker_0): turned into a stock market company. And now you want us to do fund accounting? Like I don't even get what that means. And and to to like turn the company around to get people on board with it. It was a very difficult sort of twenty-three, two thousand twenty-three, twenty-four, trying to get and then the customers as well. Like, hey, this is what we're going to start working on.
21:29
Unidentified speaker (speaker_0): It was a tough time, but but I would say we we've done a really good job the last two or three years to to figure it out.
21:34
Unidentified speaker (speaker_2): And so talk about now the opportunity you see going forward. You touched on this off the top too with defining what it means to be an ERP for private capital. When I sit back, you know, I think the sort of fund admin space, the operations layer of private capital globally is a massive arena to go and play in. And that may have taken some convincing, but once you start locking in on that vision, it is a
21:54
Unidentified speaker (speaker_2): fantastic market with all kinds of problems to solve through technology. And so where do we stand today on this infrastructure underlying private funds and private capital firms?
22:04
Unidentified speaker (speaker_0): Yeah, it's an incredibly big market. And I'll say, like, you might ask, like, what why does somebody need to get into this space? Like, what thought it was a solved problem? What's the disruption here? And what's interesting is I kind of reflect back, I thought we were a captable stock market business. What I realized is we're we're actually an accounting company from the beginning. We just
22:24
Unidentified speaker (speaker_0): didn't know it. So if you think about like what accounting is for an operating company, a startup or any company, any operating company, the accounting system, the Nutsuite Zero, you know, whatever solution, QuickBooks takes revenue minus expenses. That money goes is fungible money that goes into retained earnings. So that's the profit of the business. It goes into
22:44
Unidentified speaker (speaker_0): retained earnings. And then what happens with those retained earnings is it gets divvied up to the shareholders via the captain. cap table.
22:50
Unidentified speaker (UNKNOWN): So you
22:50
Unidentified speaker (speaker_0): could argue. as a cap table company, we were the second half of the accounting equation. The first half is revenue minus expenses, very well-solved problem. There's a million companies that be back, but nobody solved the second half, which was now that I had the retained earnings, how do I distribute it? And that was the problem we solved. So then we started look at looking at funds. And
23:10
Unidentified speaker (speaker_0): funds don't do operating company accounting, they do partnership accounting. And partnership accounting. is different because there is no such thing as retained earnings in a fund. Instead, PL is tracked at the partner level because you have to generate K1s for each partner. And so it's very different. Where in an operating company you have one PL and retained earnings. In
23:30
Unidentified speaker (speaker_0): a fund, you have thousands, hundreds, thousands of PL. You have a PL and an income statement for every OP. LP. And that means in fund accounting, the cap table in the accounting and system. have to be combined. You can't separate them like you can with operating companies. And nobody has solved that problem. There is no net suite for partner level accounting or partner level PNL. And
23:50
Unidentified speaker (speaker_0): that's our reason to exist is we went and decided to build that because of our cap table roots. And that product has taken off. And we're now going after all the basically anything with partnership accounting, which obviously the biggest markets are private. equity, private credit. But
24:04
Unidentified speaker (UNKNOWN): did you
24:04
Unidentified speaker (speaker_2): I was gonna ask, did you see traction first among venture capital and then move into PE and private credit? Or
24:09
Mark Andrew: yes.
24:10
Unidentified speaker (speaker_0): Exactly. And and because venture capital is like in our neighborhood, I live on a Sandhill Road, like we're right here and where the captables our problem was it kinda saturated the cap table business and then the f fund admin business. And we're like, oh, can we take this playbook outside of venture? And that was the big unlock for us is was the playbook of Net Suite for funds
24:30
Unidentified speaker (speaker_0): works in private equity, private credit. We think it'll work in energy, infrastructure, anything where you have closed end funds, where you do PL at the partnership level. And so that dramatically expanded the TAM for us. That's what's that's what's really been driving the last two years and I think the next 20 for us.
24:46
Unidentified speaker (speaker_2): Fantastic. So you're building a product for the partners at those closed-end closed end funds to be able to better serve. service and understand their final accounting, their operations. Do you also see the growth of Evergreen funds and always on funds, which is a huge topic across a lot of these private credit funds in particular, as a growth area? Is that something you're servicing?
25:03
Unidentified speaker (speaker_0): Yeah. So we service Evergreen. We've definitely seen a lot of growth there. And the large end private equity, I I call it the retailization of private equity, where private equity firms are trying to access retail capital. Uh this is going a little bit back to like public versus private market infrastructure. Here's where these start to hit. Because if you have a mutual lack fund, like an
25:23
Unidentified speaker (speaker_0): interval fund or a semi-liquid fund that where you can access retail money, you have to strike nabs more often than once a quarter. So in in our world of private equity, private credit, It the cadence is a is a quarterly cadence. We all operate on quarterly cadence in and private funds. Public markets is daily, and these interval funds are starting to split
25:43
Unidentified speaker (speaker_0): the difference. And so we work with a lot of private equity funds to feel like, hey, we used to do valuations and reporting once a quarter. Now we have to do it once a month because it's a registered act fund. So we need to do it like a mutual fund. Some of them are moving to once a week. I think if they if they really go all in on it, they're going to gonna have to post daily nabs. navs. Right now they just can't do it. They literally cannot. do
26:03
Unidentified speaker (speaker_0): it because they don't have the human power to do these spreadsheets every, every day to calculate the nav for hundreds of portfolio companies. And that's where we think software will not just solve a lot of the pain points for them, but it will allow the asset class to scale because it's been a pretty sleepy asset class, at least in terms of operations, for a long time. And
26:23
Unidentified speaker (speaker_0): now if they're going gonna to access retail money, they're after they're gonna going to have to get better at it.
26:26
Unidentified speaker (UNKNOWN): So,
26:27
Unidentified speaker (speaker_2): given your thoughtfulness and experience on the private shares into the average person's portfolio dialogue, let's talk about private funds into those people's portfolios and where it makes sense. What is your perspective on this as to who private markets and by this I mean funds, distribution BDCs, private credit into an average portfolio, average person's portfolio? When is that appropriate? Who Who's is it
26:46
Unidentified speaker (speaker_2): appropriate for? And how big will that space get? I mean, how big will will it? every individual have a certain allocation to private markets across their portfolio?
26:55
Unidentified speaker (speaker_0): Yeah, I I hope so. I I think so. I think, you know, and if you're doing direct stock picking, the problem with direct stock picking for many individual investors, a lot of people say, Oh, they don't have the expertise, they don't have the money, you know, to lose all of those things. But the biggest problem is this goes back to this difference between private and public. You may ask, like, what is the
27:15
Unidentified speaker (speaker_0): difference between private and public? And you'll talk about, oh, well, some is regulated, others not, blah, blah, blah. I would actually say all of those are symptoms of the most fundamental underlying difference between public and private markets is that in the public markets, you are allowed to transact with people you don't know. In private, you're not. When in 1933,
27:35
Unidentified speaker (speaker_0): when the SEC created the Securities Exchange Act, they said hey, you can do anything you want in private, but you are not allowed to solicit to people you don't know. You can only sell stock to people you do know. And so that's why it's it's a broker business. That's why it's a people business. There is no central exchange because it is illegal to transact with people
27:55
Unidentified speaker (speaker_0): you don't know. You can't solicit or sell to people that you haven't met. And so the problem is, is if you're trying to buy an individual company, you have to know that CEO
28:04
Unidentified speaker (UNKNOWN): for
28:04
Unidentified speaker (speaker_0): them to sell it to you. And if you don't know them, which most people don't, you're not going to get access to it. And so by definition, you always get the worst companies because the companies where everybody already knows the person get to invest first and you don't. So there's a very strong negative selection bias. I think funds can be different because the Is the retalization retailization of funds, like if you're KKR
28:24
Unidentified speaker (speaker_0): and you're going gonna to you know open an enterprise. interval fund, you don't need that through the fund. Like if you want to invest in a KKR company, you need to know the company, but you don't need that for the fund via an interval fund. And so you can have positive selection bias. So I'm a big fan of that. And I think a lot of people will should diversify the portfolio and have exposure to private versus public as a piece of their
28:44
Unidentified speaker (speaker_0): portfolio. I think the thing that I'd love to see is that some of the fees could go down because it is high fee for retail.
28:51
Unidentified speaker (UNKNOWN): Yeah,
28:51
Unidentified speaker (speaker_3): there is a vision where there's almost an ETF-like exposure to these different funds
28:55
Unidentified speaker (speaker_2): where you can start having an open rails where you've got selections of different managerial managers' funds within a common ETF according to a certain asset class, and it just drives more diversification. But your point on the fees, if that's eating away a lot of the returns, then it may not make sense for the average investor. And and you know, a lot of that requires a lowering of transaction costs, a better infrastructure. So, how
29:15
Unidentified speaker (speaker_2): would you characterize Carto's platform for that discussion of funds and the lowering of transaction costs?
29:22
Unidentified speaker (speaker_0): Yeah. So a big part of the transaction cost is just that you have to know the people. And so so much of a two in 20 you're paying to a mentor fund or a private equity fund is their time just flying around the country meeting companies and like figuring this stuff out because they can't just digitize it and put up on a bullet. bulletin board like we can in the public market. So
29:42
Unidentified speaker (speaker_0): that there's that like pay for time cost structure that exists. And then the other is just like the management fees. And one would hope that the pay for time would be in the carry. And so the incentives are aligned. But then there's the other, the the two and twenty. The two part of it is the operations piece of it. And that's a big, you know, 2% two percent a year over 10 ten
30:02
Unidentified speaker (speaker_0): years. There's a lot. is of money. And so ca and that's where CART is coming in, which is how do we operationalize it? Can we get that two to one to to half a point by making it faster, keeper, better, smarter, less less employees, like all the things to run a a fund better, faster, keeper. And especially now with AI, like the amount that we can reduce the cost structure of
30:22
Unidentified speaker (speaker_0): those, I think can be can be quite dramatic.
30:24
Unidentified speaker (UNKNOWN): So
30:24
Unidentified speaker (speaker_2): let's talk about the fund administration business, which is effectively what Carta is now. This is a longstanding business. This business has been around. It's got some serious incumbents in it. As a disruptor, what do you see when you look at the fund admin business and how do you see technology playing a role in it? Many will say, and I'll set this up a bit further, when you kind of I've written about this on LinkedIn, for example, and been critiqued, right?
30:44
Unidentified speaker (speaker_2): Like the application of AI, and then some of the incumbents will say, well, this is a very high high-touch, touch, high services business. It's got multi multi-currencies, currencies. it's incredibly intricate and important with different cash changing hands and capital calls and so forth. And you can't automate too much of it. I've heard that said, literally. So tell us what you see when you look at that business.
31:01
Unidentified speaker (speaker_0): Yeah. Well, it it's interesting. So a lot of people do the like it's really complex, technical, detail Detail-oriented. oriented,
31:11
Unidentified speaker (speaker_2): human.
31:12
Unidentified speaker (speaker_0): And yeah. And so so people would say, so you need humans. And I would say, so you really shouldn't use humans to do this. It's a little bit like, you know, the term calculator came from there was a job called calculator. Before they had calculators, that was a job. You got trained and you did the math equation. And if you were at NASA, you know, there was the head calculator that had rooms full
31:32
Unidentified speaker (speaker_0): of cubicles full of hundreds of calculators, human calculators that did a piece of the math puzzle and they'd aggregate this, and this is how they calculated orbits of the moon, rocket trajectories to the moon. And then when the calculator, you know, Casio came out, or Texas Instruments, there were a bunch of humans that were like, we can't trust the little electronics
31:52
Unidentified speaker (speaker_0): thing to do this massive calculation. People's lives are at stake if we miscalculate the trajectory to the moon. And now we know how absurd that is. The argument that something is very difficult is the argument to be made that we need to build tooling for it versus let humans continue to do it. And so the problem is I have a lot of
32:12
Unidentified speaker (speaker_0): conviction I'm right over time. The question, and we've always been right. I mean, that has always been the direction of human progress, is everything that's hard for humans, we build technology to make it easier for humans. It's Whether it's excavators to dig shovels, whether it's forklift. forklifts, whatever it is. That's the path of human progress. The question is always the timing. When does
32:32
Unidentified speaker (speaker_0): it happen? Like how quickly can you move human calculators to Cassios and Texas instrument calculators? You know, how quickly can you get people off shovels and into excavators and train them on how to use the excavators and deliver the excavators? And so now it's how quickly can you get people, how quickly can the AI get good enough that people trust it? And how quickly can you get the behavior
32:51
Unidentified speaker (speaker_0): change for people to go, okay, you're right, the AI is better. I think what is different in this technology transformation than in past is the speed that it's happening is
33:03
Unidentified speaker (speaker_0): unbelievable. It is completely remarkable. At the beginning of this year in January, we had zero AI doing fund accounting. Today, about 90% of the journal entries booked on CARTA, and we book millions of our journals a month. We process about 200,000 individual emails from our
33:23
Unidentified speaker (speaker_0): customers to ask them to do something as a fund accounting service. I mean, we operate this at massive scale. 90% of all of that now is done with agents eight months later from January. The speed is astounding.
33:36
Unidentified speaker (speaker_2): Given the pace and the technology, what is the biggest connection? constraint? And I suspect, I mean, often you'll hear about just the human reaction in these funds. And maybe tied to this, earlier you had said, which is a very common thing to say, this is an industry that's well known for, let's go say private equity, private credit, is an industry well known for being a technology laggard, despite often investing and owning in technology businesses. What is
33:55
Unidentified speaker (speaker_2): changing? And then what is the next step to implementing AI across private capital firms?
34:01
Unidentified speaker (speaker_0): Yeah. I think a couple of things are changing. It is a sleepy, laggard uh industry when it comes to like their own transformation, even though they're in the business of transforming businesses. Their own transformation, they tend to tend to delay on.
34:15
Unidentified speaker (UNKNOWN): Um,
34:15
Unidentified speaker (speaker_0): I think two things are are changing. I think, you know, if all you came into, if I go to a private equity firm today or or two years ago, and I'm like, hey, I'm gonna transform accounting for you. You are gonna like, the accounting is gonna be amazing. They're like, they don't care. Like nobody cares. Just like nobody cares. I I don't care that much
34:35
Unidentified speaker (speaker_0): about the accounting of my own company. But my CFO is the most important person that I talked to about any material physician, not because he does the accounting, but because the accounting is done and he has time to do other stuff. And he has time to do FPA and strategy and planning. and da da da da. And so I can't make a decision now without him because he owns all the
34:55
Unidentified speaker (speaker_0): systems that run the company. And so what we do is we go in now to GPs and say, hey, you thought accounting was a back office function. Like you just didn't care about it. But let me show you what it looks like if you if you actually had accounting solved and then you could build planning, strategy, finance, all of the tooling on top of it.
35:15
Unidentified speaker (speaker_0): that CEOs have, but GPs don't. And CEOs have because we have NetSuite and we have Anaplan and we have Ramp and Brex and Coupa and all these systems that run the business. So now we can drive strategies on top of those systems. And then the second is: I do think there's a generational change that's happening. If you look at CFOs
35:35
Unidentified speaker (speaker_0): that are at the tail end of their career, they're like, look, I just implemented an accounting system in 2014. I'm not doing this again for another 20 years. But they're getting replaced by the young, you know, the CFOs that are in the early part of their career. They're like, I want to be a partner to the GP. I want to use the best technology that's out there. I want Binance
35:55
Unidentified speaker (speaker_0): and a firm to contribute to alpha, not just be an expense, but to be an alpha generator for the firm. And when we talk to those CFOs, the conversations are very different.
36:04
Unidentified speaker (UNKNOWN): Okay,
36:04
Unidentified speaker (speaker_2): let's speak specifically. That's exactly where I was going to go next. How do AI native private capital operations, quote unquote, support GP's job, which is finding alpha, creating alpha?
36:17
Unidentified speaker (speaker_0): Yeah. So one of the biggest things that we work with CFOs on is they're like the GP, I have to do the Monday morning tear sheets. I I I have to do the, you know, account reviews. I have to do the valuations. I have to do all these things. I spent all my time doing the work and none of it actually interpreting and helping my GPs understand it. And so one of the things you tell them is like,
36:37
Unidentified speaker (speaker_0): let the software and the AI and us just do the work for you. It is not a good use of your time to review debits and credits. Like that's crazy, right? You should just assume all that stuff works. And your job is to interpret and help understand those things for the GP. The second thing is most of the time, that like how the workflow works when the GP is like, I
36:56
Unidentified speaker (speaker_0): really have to understand what we invest in this company, if we model out an exit, what that looks like, what the fun waterfall does if we sold this company in six months,
37:05
Unidentified speaker (speaker_0): like all the things that go around it. And they have to ask the CFO that, and that's like a three three-week week project versus I don't have I just log into, you know, Anoplan. AnnaPlan with my my CFO and we just look at all the possible outcomes, how to make decisions. And so can you get to that level where the GP has one question and they don't even want to ask the questions because they know
37:25
Unidentified speaker (speaker_0): it'll take three weeks to get an answer. What are all the questions you can ask for the GP? Uh and so when you start thinking about like forward looking and strategic, it changes. It's it that that's the biggest challenge we have is like getting people to understand that. kind of behavior change. Your job isn't to do debits and credits anymore. Your job is to be strategic finance.
37:44
Unidentified speaker (UNKNOWN): And
37:45
Unidentified speaker (speaker_2): we're recording this in September 2026. Where do we stand today? It's been a big year for dialogue and circumspection about the impact of AI on software and services in particular. Where do we stand right now in your vision? And for those who haven't read Henry's pieces, whether it's you know on your blog or on LinkedIn, Henry about AI, you're as thoughtful and knowledgeable as almost anyone on this topic. So
38:05
Unidentified speaker (speaker_2): how is AI impacting software today? And how is it going to impact services over the next five years?
38:09
Unidentified speaker (speaker_0): Yeah. So all we're building now is agents at Carta. And I I think so I think this year is the year, it'll bleed into next year that I think the world is accepting AI. In January, I would say the inflection point was Opus V. When Opus V came out, things changed. And we could do stuff we
38:29
Unidentified speaker (speaker_0): didn't, we wanted to. We imagined maybe one day we would be able to do, and it suddenly became real. And you can see a world, Carta, where almost all the work and email correspondence with our and support of our customers is now agent driven. Like our humans no longer do dubits debits and credits. Like all they're doing is training our customers on
38:49
Unidentified speaker (speaker_0): AI and how to use. the AI that we we give them. I think more and more the surface area of the product, you know, we have a a million pages on Carta.com that our customers log in and do cap calls and, you know, run reports and do whatever, all of that's moving into our cloud plugin. So they can just fire up cloud. They're already in cloud doing stuff. They access
39:09
Unidentified speaker (speaker_0): the Carta plugin and MCP. And anything they could do on the web, they can just now do and in and cloud. And I think more of the service area of the product is going to go into these harnesses. And then more of the work that's going to be done is in our harnesses that we custom build for customers.
39:25
Unidentified speaker (speaker_2): And one of the big picture questions that I've sometimes wondered at, maybe in an abstract way, is who owns the private market's data as this evolves toward maturity? Because you've done some work on the institutional LP side. Obviously, with the rise of wealth, you're going to have gonna new types of LP. LPs. The GPs own their own data, but they've often exported it to the fund admin historically. Curious what you think of this. Is there
39:45
Unidentified speaker (speaker_2): gonna be an emerging back to this general ledger of private shares point of view on the private funds? Like, will it be companies like Carta in the fund admin space who have this data? GPs obviously have ownership of the data, the LPs expect ownership of the data. Where does the data sit in private capital?
40:01
Unidentified speaker (speaker_0): Yeah, it's a great question. I think it does. I think like captable data sits with us, but it's owns. We're the custodian. We are not the owner. So we custody the data on behalf of our customers. We have the ability, we can't say anything about any customer, but we have the ability to say, to aggregate this stuff. So we can say, oh, you know, the total number
40:21
Unidentified speaker (speaker_0): of rounds that were done, you know, in venture capital were this, or you know, there's so many, you know, this is the average valuation of Series A companies. We can talk about in aggregate anonymous terms. The same is true in private capital. So, and I think there's a lot of value in having a system of record where all the data is in one place, or custodied, I should say, in
40:41
Unidentified speaker (speaker_0): one place, but owned by the constituents that all participated, going back to the very beginning, right? The lawyers own their piece of the data, the companies own their piece of the data, the investors own their piece of data. the but they all trust us to be the custodian of it. You may ask, like, why doesn't Carta have a data business? And it's exactly for that reason. I have this view you can
41:01
Unidentified speaker (speaker_0): you can be a software company or you can be a data company, but you can't be both. Because if you're a software company, you don't own the data. You are custodian of the data, which means you can't sell it. And if you're a data company, you can sell it, but that means your customers will never give it. to you. And so if you look at the data companies in this in this space, like prequin and so on, they all get data via
41:20
Unidentified speaker (speaker_0): survey, right? They they don't actually own system of truth. It's all via via some form of collections.
41:26
Unidentified speaker (UNKNOWN): I
41:27
Unidentified speaker (speaker_2): speak to quite a few CFOs, COOs who are working in this space who are building out their AI. We're doing an AI summit in November. They often just want to know where they're where they rank among their peers. And many are still getting started. What is your advice to the private capital firms you speak to who are who who would ask, you know, where do I stand today on AI, even if I'm just getting going with cowork or uh not using agents
41:47
Unidentified speaker (speaker_2): yet? How far behind am I if I'm not implementing agents, for example? And and where do I get started other than of course calling CARTA or something like that?
41:54
Unidentified speaker (UNKNOWN): Yeah.
41:54
Unidentified speaker (speaker_0): Well at the risk of being self-serving, I would say the first you could start with CARTA because we actually have a program where we train CFOs on how to use AI in the fund space. Like it's literally for that. And it's very self-serving for us because we think most of the work is moving into AI. And if our customers don't know how to use AI, they're not gonna know how to use CARTA because like
42:14
Unidentified speaker (speaker_0): everything is now being delivered via AI. And so we do a free training for all our customers on here's here's how Claude Cloud works. And it starts with what's a skill? What's the difference between cowork and code? What's Cloud M? MD for? What's a working directory for? Because we need them to have the basics in so that when they use the Carta plugin or MCP, they
42:34
Unidentified speaker (speaker_0): understand the basics of how these cloud, you know, Cloud Desktop work so they can interact with us. If you're not if you want to do it on your own, it sounds kind of funny, but man, YouTube is an incredible resource. There's a million videos on how to get started on Cloud Desktop. The biggest challenge that most CFOs have in these firms is even if they learn
42:54
Unidentified speaker (speaker_0): like the technical skills of how to use Cloud Desktop, like what a skill is, how to deploy a skill, how to share a skill, like all uh what an artifact is. The the hardest part most people have with AI is, well, what do I do now? Like it's so powerful, it's almost hard to start. Like, what is the use case that I can use it for? And that's where we find most CFOs struggle
43:14
Unidentified speaker (speaker_0): is the application. of AI.
43:16
Unidentified speaker (UNKNOWN): For
43:16
Unidentified speaker (speaker_2): years it was get your data right first and maybe talk to a snowflake or databricks. Is that still the case or can is have we have we started moving away from even just a focus on data as a first layer of application?
43:27
Unidentified speaker (speaker_0): So for if you're working with like a a good system of record, the data, data tends to be good. So like with us, we have a data warehouse, our customers can plug in directly to the cloud, they can pull the data, the data's good, and then it's like what do you do after that? Same like Salesforce, you
43:41
Unidentified speaker (UNKNOWN): know,
43:41
Unidentified speaker (speaker_0): if you you can access Salesforce data, you you hope the Salesforce data is good, and then it's what can you do after? I think the concerns of a year ago where it's like, oh, hallucinations, does it get the right data but do the wrong thing? Does it, you know, read the document wrong? Is it making stuff up? That's largely solved. That doesn't
44:01
Unidentified speaker (speaker_0): exist anymore. I think the the big challenge where a lot of CFOs now have is like the tools. Okay, so I can get the data, but what can I do with the data? Like what's interesting to do with the data? And that just requires more and more tooling and infrastructure that goes into these, you know, whether it's codex or or or claw desktop. And that's what these plugins are for,
44:21
Unidentified speaker (speaker_0): right? And that's why like if you look at CARTA, most of engineering is now building plugins versus web pages.
44:27
Unidentified speaker (speaker_2): Henry, you've made a number of acquisitions over the last 12 to 18 months. Of most interest, what jumped out at me was you literally bought a law firm. Almost every private market's operation runs through the legal function at some stage. Tell us about Avantia and it wasn't just legal tech, it was a literal law firm.
44:45
Unidentified speaker (UNKNOWN): Yeah,
44:46
Unidentified speaker (speaker_0): it was a law firm. It's such a it was about 200 lawyers and paralegals paralegal, started by a former private equity G. GC, very forward-thinking GC, who wanted to build a tech-enabled law firm, but started with the legal side. And his thesis was hey, you know, there's a lot of legal work that private equity and private credit firms are overpaying for and getting a not a great level of service. So I'll give you
45:06
Unidentified speaker (speaker_0): an example NDAs. You know, we have lots of customers that do NDAs. They pay us to negotiate the NDAs. We have lawyers and paralegals trained on the customer playbook. If you're a large private equity firm, you might do 15, 20,000 NDAs a year. And you don't want to pay, you know, a top-end $2,000 an hour law firm just to negotiate NDAs. And so they pay
45:26
Unidentified speaker (speaker_0): us to do that. And we do LP transfers and all these other things that I would put in the category of high volume, low complexity, fixed outcome work. right? I just get this NDA done against a playbook and a and sign. sign. You know, just get this LB transfer done, just get this KYC inquiry filled out. And they did an amazing job of scaling the business.
45:46
Unidentified speaker (speaker_0): And then when they came to us, it was like, hey, we have so many customers, CFOs and private, you know, PE firms and G and GCs and GPs that have to deal with this. You know, we should be able to offer legal services to them along with our financial apple, you know, software and platform. But also we're deep deep on the automation of
46:06
Unidentified speaker (speaker_0): services via AI, whether it's fund admin services, legal services. And they were earlier in their tech journey. They were doing really well commercially, but earlier in the tech journey. And we're like, hey, can we take our AI playbook and start applying it to the legal? So an NDA DA doesn't get negotiated by a paralegal anymore. It gets negotiated by an
46:25
Unidentified speaker (speaker_0): agent. And then if it becomes instant, right? It's the turnaround, the the markups are done in two minutes instead of two hours.
46:33
Unidentified speaker (speaker_2): That is fascinating. Tied to this question of LPAs and legal docs are is the question of standards across private capital. It's something you hear a lot of how to standardize a subdoc process, LPAs, because the industry, you know, if it's looking at public markets as inspiration, thinks, well, maybe it's got gotta to figure out a common standard. for all these docs in order to scale. What is your
46:53
Unidentified speaker (speaker_2): take on this? There's a lot of debate as to whether AI can help standardize the process because obviously prior s efforts at standardizing key documents in private markets haven't worked. Do you think the industry will eventually, as it matures now, and certainly a lot of focus around this, will it move towards standardized docs Docks or will AI just solve that problem?
47:11
Unidentified speaker (UNKNOWN): Yeah,
47:11
Unidentified speaker (speaker_0): well, I think that the the question is why hasn't
47:15
Unidentified speaker (speaker_0): standardized docs? And the argument, one argument is the market doesn't want it. You know, it's like the market doesn't want standardized.
47:21
Unidentified speaker (speaker_2): There have been lots of efforts at this. Yes.
47:23
Unidentified speaker (speaker_0): Totally. Many, many people have tried. And L LPs all want their own version of the LPA. They want their special side letters. You know, GPs want to do things a little differently. So one argument is the market doesn't want it. And in that case, if if the answer is it's the market, it's a market product. problem, then AI is actually wonderful because it doesn't require standardization.
47:43
Unidentified speaker (speaker_0): Standardization is a human problem. We want standardization as these service providers because it's too complicated to track everybody's unique, you know, and if everybody's a snowflake, it's really hard to put that stuff in software. But with AI, now everybody can be a snowflake. It's actually wonderful. And so one could argue AI will allow
48:03
Unidentified speaker (speaker_0): us to actually create more, have more creativity and more broadness and less drive for standardization in the in the uh in private markets, which could be a good thing if that's what the market wants. Another argument is actually the market wants, you know, LPs don't want to negotiate every single LPA. They would love a standard you know template that they use
48:23
Unidentified speaker (speaker_0): every investor. So the market actually wants standardization. So the question is, well, why hasn't there been standardization? And that the argument I would say in that case is is probably the lawyers. Because the lawyers get paid to not standardize, right? The whole point of standardization is you don't need lawyers to do it. And so the lawyers have a perverse incentive to not standardize and to tweak everything and
48:43
Unidentified speaker (speaker_0): rewrite stuff and do it their own way and so on. And that's what keeps the legal industry going. And so in that world, us owning Carta Law can really help because we actually have strong incentive. Like we don't bill by the hour, we bill by the output, by by the project. You know, we bill by per NDA or per LP transfer. So then we have an interest in standardizing. And and my
49:03
Unidentified speaker (speaker_0): hope is we will help drive the industry to standards that the market wants, but then also have AI that when the market doesn't want standardization, we can accommodate those those edge cases as well.
49:14
Unidentified speaker (UNKNOWN): Well,
49:14
Unidentified speaker (speaker_2): and the industry is shifting its target market, right? We've talked about this wealth piece and and you know, an industry that was that had taken shape around bespoke institutional agreements is now moving towards scaled wealth types of structures and agreements. And so how that impacts the standards debate, I suspect, will be pretty profound.
49:32
Unidentified speaker (speaker_0): Yeah, 100%. A hundred percent.
49:34
Unidentified speaker (speaker_2): So all right, wrapping up then, Henry, I'm curious as someone who's such a power AI user. at work, what are the most interesting personal applications you're using AI for these days? I'm sure it's walking through every aspect of your personal life too.
49:46
Unidentified speaker (speaker_0): Yeah. Well, the thing I've been playing with most, and it maybe it's like personal professional blend a lot a lot for me. You know, I think I do all the standard stuff. I use it help my son with his math homework and, you know, recipes for dinner. Like I do all the normal stuff. We we have an internal tool that we use for engineering, but we've now expanded into business
50:06
Unidentified speaker (speaker_0): cases. And I'm trying to figure out how to apply it in more personal cases, which is multiplayer uh AI. So the the big problem, of course, with cloud code is it's only for me. I'm the one managing the session. We now have tooling where we can have sessions run where multiple people log in. And for example, you and I, Mark, could be in one session trying
50:26
Unidentified speaker (speaker_0): to figure out this podcast. Instead of you, you know, running a session and exploring giving me the questions and me loading it into my session and figuring out the answers, like we can just have one session and we can work on it together. It's almost like a Slack thread, but with an agent running in the background that's talking to us. And so finding all these applications of multiplayer
50:46
Unidentified speaker (speaker_0): AI, I I think that's the the next huge thing.
50:50
Unidentified speaker (UNKNOWN): Well,
50:50
Unidentified speaker (speaker_2): Henry, there's been a this has terrific conversation. Thank you for your time. And we look forward to further conversations. You know, if I was to close with one uh last one, uh assume the year is 2040.
51:01
Unidentified speaker (UNKNOWN): How
51:01
Unidentified speaker (speaker_2): does the average private capital firm look different?
51:04
Unidentified speaker (speaker_0): My I I think one one of two ways. One is the consolidation that's happening continues because capital begets capital and there's like there's like two firms left. You know, it's like Toma Bravo and KKR and and there's nothing else. You know, they just dominate just like there's four banks that matter, you know. The other
51:24
Unidentified speaker (speaker_0): version is because of AI, because of operational efficiency, because of technology, all the tools that we're building, all those things, there's an explosion load in where yeah, you yeah. know, the the other extreme is everyone has their own private equity firm. Like we all know like the guy down the street that's running, you know, whatever business, and you're like, look, I could probably run that a little better.
51:44
Unidentified speaker (UNKNOWN): The
51:44
Unidentified speaker (speaker_2): long tail, just like social media is in back, for example. Yeah. Focused
51:47
Unidentified speaker (UNKNOWN): on the
51:48
Unidentified speaker (speaker_2): typical small niches. Yeah.
51:50
Unidentified speaker (speaker_0): Totally. You I would have thought, you know, social you know, and if you asked me this question twenty years ago, I'd be like, you know, there used to be when I was a kid, there was like three television channels, right? And the I would have
51:59
Unidentified speaker (UNKNOWN): been
51:59
Unidentified speaker (speaker_0): like, there's gonna be one, you know, when I grow up like like you know, it's gonna consolidate and it said it it f it fractured, it's splintered into a million channels. And then will will that be true of private capital? There'll be a not just we're already fracturing splintering into a million companies, will it be a million funds? And we find that idea really exciting.
52:18
Unidentified speaker (speaker_2): That's very interesting. As someone who does a ton of writing too, to express your thoughts, how do you use Clode Cloud to help you write? Do you do you still write first drafts yourself? Are you using it a lot?
52:26
Unidentified speaker (speaker_0): Uh I do all my own writing. I I've just found Clyde Code doesn't, you know, it doesn't capture the essence, you know, of the writing. Well, I should bucket, two types of writing. There's like generic writing that you just have to put informational content out there. And then there's writing where you're trying to, you know, explain an idea that no one's thought of before, right? It's a
52:46
Unidentified speaker (speaker_0): new concept in the world. And that one you just have to do yourself.
52:49
Unidentified speaker (speaker_2): Writing is the act of writing is the act. of thinking. There's nothing else that draws it from you. And I often, you know, yes, for business writing, AIs can be great as long as you don't take the first prompt and then use it, by the way. But I'd never suspect I'll ever read fiction by AI. And you can always tell it is. And I I'll be curious, I guess this is maybe the AGI debate. Could you ever see a point
53:09
Unidentified speaker (speaker_2): where you could be fooled by fictional AI writing? Probably. It's hard to bet against, but I think the human experience lends itself to that fiction. So
53:17
Unidentified speaker (speaker_0): this is an interesting idea. So I think most people at this point believe AI is going to be smarter than most humans, and I would argue all humans, perhaps all humans combined. So I think most people are on that trajectory of believing that. Just like twenty years ago, we didn't believe AI could beat us at chess, and now it's like obvious it will beat us at chess. So it will be smarter than us at
53:36
Unidentified speaker (speaker_0): everything. There's a site parallel of question. Will it be more creative? And I think the the question, I think one way to try to understand that question is will we will we be listening to AI music over human music? And I think most people have a visceral reaction to no. But it's already started because we're w for a long time, we're listening
53:56
Unidentified speaker (speaker_0): to AI chosen music for us, right? We all have that Spotify track, right? It's a it's computer. the music's human, but the track is is computer generated. And all you know, there's this next question is well, if the the AI is picking the music, what if the m AI just starts going one level more primitive and picks the sounds and starts
54:16
Unidentified speaker (speaker_0): generating music? And it is kind of this weird thing. We all listen to the same music. Why don't I listen to custom music made just for me? Because it knows my preferences. And and I think that will be the my guess is that's the beginning of where people go, oh, actually AI can be better than humans at creativity too.
54:33
Unidentified speaker (speaker_2): Just to put it on the record, Jan Lakun, who is obviously one of the godfathers of AI in this space, has as I was just reading over the weekend his comments on why we won't achieve that type of AGI that surpasses humans. And not just to be a contrarian about it, but his specific point is that all of the text in all of the world doesn't rival what a human sees by the age of four in terms of
54:53
Unidentified speaker (speaker_2): their visual and sensory experience. And so obviously this is partly why they're building world models, different types of models, and I'll never bet against AI. I do wonder, though, if there's something about the full intake of sensory experience that humans can achieve that AI will struggle to ever achieve. And that that is where creativity is embedded, for example. So just a closing thought on that question.
55:11
Unidentified speaker (UNKNOWN): Yeah.
55:12
Unidentified speaker (speaker_0): Well, i interesting thought I that argument seems to silly silly to me because all you had to do is we could just put a camera on one child.
55:18
Mark Andrew: Exactly. And
55:19
Unidentified speaker (speaker_2): that's what's being
55:19
Unidentified speaker (UNKNOWN): worked on.
55:21
Unidentified speaker (speaker_0): The difference though is we only get one child's sensory. Like we could put a camera on every child, and AI could get every child's sensory input. And that's the difference. Like I I that that argument is a little bit like, well, the same argument then applies for chess. Well, they won't see every game that one person plays. So that person is going to be better because
55:41
Unidentified speaker (speaker_0): they saw all the no, he'll see every game that every human has ever played in in the history of humanity. So it's just it's a question of do you think uh yeah, I that argument doesn't make a ton of sense.
55:51
Unidentified speaker (speaker_2): And by the way, I that is still why I would feel bullish on the need for more compute, if anything. So anyway, Henry, thank you for this conversation and look forward to the next one. Appreciate your time.
56:01
Unidentified speaker (UNKNOWN): Thank you,
56:01
Unidentified speaker (speaker_0): Mark. We really appreciate it. Thanks for having me.
56:03
Unidentified speaker (UNKNOWN): me.
56:05
Unidentified speaker (speaker_2): Thanks for listening to Modern Capital.
56:08
Unidentified speaker (UNKNOWN): The
56:08
Unidentified speaker (speaker_2): next generation of private markets is being built right now.
56:12
Unidentified speaker (UNKNOWN): And
56:12
Mark Andrew: it's reshaping business, technology, and society.
56:16
Unidentified speaker (UNKNOWN): I'm
56:16
Mark Andrew: Mark Andrew. Learn more about our work at the Private Markets Forum.
56:21
Unidentified speaker (UNKNOWN): In
56:21
Mark Andrew: each of our upcoming events at private-markets.com.
56:27
Mark Andrew: See you next
56:28
Unidentified speaker (UNKNOWN): time.
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.
All insights