Venture Mandates, Cloudflare Wallets, Circle Earnings & Saddam Hussein Patek Philippe
- Rob Hadick's read on crypto funds drifting into robotics and hardware: it can be an adverse-selection trap, not a mandate upgrade. "If you're seeing it, it's because all the real robotics [investors] passed on it," and LPs are already complaining — one brand-name LP asked whether to route money to a big traditional generalist instead. Yano's rule: "you earn the right to be a generalist... if you are everything for everybody, you are nothing to nobody." Sequoia did not go generalist until roughly its eighth fund.
- Post-2021 crypto returns have converged with regular venture — and regular venture is difficult. New Carta data puts the 2022 vintage's 90th percentile at 1.68x TVPI and 12% DPI across venture; a speaker says crypto's numbers are better, but the beta era is over: "long gone are the days of 10x in two years." The remaining investable universe is finance moving on-chain — DeFi, payments, perps and centralized finance — not necessarily NFTs, consumer or social.
- Cloudflare Wallets is a bullish milestone for agentic payments — but the value may accrue to "owning the identity and the authorization layer," not a particular rail. Human-controlled account wallets spawn guardrailed virtual wallets for agents, with x402 micropayments attached to HTTP requests; Cloudflare — 25% of the internet, with bots at 55–60% of activity — stayed stablecoin- and chain-agnostic. The traditional credit-card industry may eventually move to stablecoin settlement, a personal opinion rather than a consensus view.
- Circle's mixed quarter at about $64 per share: revenue +7% YoY, USDC supply down $77B→$73B, and about 62% of reserve income paid away in distribution. The Coinbase agreement renewed on existing terms despite market chatter, though one speaker argues Circle stock would have been more bullish if it had ended. Another speaker notes Circle and Visa at roughly 20–22x forward EBITDA and sees more downside than upside; the proposed long-Circle/short-Coinbase pair trade was immediately caveated, while another speaker preferred long Coinbase.
- Agents may not even need stablecoins — the bear case runs through single-use virtual cards. Every Booking.com transaction already uses a new tokenized virtual card, and TradFi rails plus spending rules may suffice. Stablecoin payment companies partly address chargebacks through batch settlement, while banks are also exploring tokenized deposits for intrabank, 24/7/365 settlement.
- The Western Union bet settles in November, and fieldwork made the bearish case stronger: "people love cash... people love Tether." USDT is dominant in ex-US retail, and stablecoin TAM may materialize more slowly than expected over the next 2–4 years, especially if the CLARITY Act does not pass. The bull case is that WU's roughly $110B of cross-border volume, 50M yearly active customers, 375,000 agents and 200 markets become more monetizable when recipients can hold trusted dollars and spend without FX fees. Dragonfly portfolio company Rain supplies the stack, launched in 37 markets and targets 60 by year-end and WU's 200 markets within 18 months.
- The episode's meta-lesson came via Micky Malka's Founders interview: "your brand and your reputation are the real things that compound value, and everything else is downstream from that." Also discussed: Gavin Baker's Invest Like the Best appearance two days before Situational Awareness blew up, a $26K Saddam Hussein Patek Philippe Ellipse auction that one host bid on and lost, and LiDAR revealing 396 unknown Amazon earthworks. D Asia is planned for Singapore around Token2049; D London is in mid-November, with Hyperliquid's Jeff and Morpho's Paul also mentioned.
1. TradFi has become crypto — and Claude is doing the Messari integration
- The opening riff: traders Jordi and Ansem were watching SpaceX and its unlock timing, describing the setup as "lowflow, high FTV" and saying, "we were made for this singular moment." The show's inversion of the old thesis: "I always thought crypto would become more like finance. I wasn't expecting finance to become more like crypto."
- Jason's post-acquisition report from Blockworks: integrating Messari's Salesforce CRM with HubSpot means consolidating tens of thousands of records — "there used to be these integration teams of like 50 people and now we can do it with a couple people and AI." The crypto-native business is slow, while enterprise, fintech, exchange and brokerage is "the fastest it's ever grown."
2. Dragonfly's answer to mandate creep: crypto-enabled, big-TAM, no adverse selection
- Rob's filter for deals outside pure crypto: "where am I not going to get adversely selected against?" Money is green, but founders with options generally take it from investors deep in their area. Dragonfly therefore focuses on things "enabled by crypto" with large addressable markets: Rain, Polymarket, Sapien and Venice.
- The new AI-spend deal illustrates it: an agent-infrastructure company with Agent Studio plus runtime infrastructure, payments via x42/x402 and stablecoins, and investors including Anthropic, Accel, Gradient, Coinbase and Dragonfly. Its spun-out router cut early clients' token spend by about 80% — "the token-maxing era needs to die."
- The structural claim about the space: if fund returns were open-sourced, the best firms would be multidisciplinary partnerships with deep specialists, because "I don't actually think you can be best in the world" at cryptography, capital markets and network scaling simultaneously. Crypto firms other than Andreessen and perhaps Paradigm are too small to go extremely generalist, and reinvention "takes time... over a period of years," not six to twelve months.
3. You earn the right to be a generalist
- Yano's framing: "if you are everything for everybody, you are nothing to nobody." Sequoia started as a niche semiconductor shop under Don Valentine, who had worked at Fairchild and National Semiconductor, and earned the right to generalize over time. Rob said he did not think Sequoia went generalist until roughly its eighth fund. Dragonfly earned payments by being a top-three crypto fund first; chasing shiny objects is "a real recipe to end up jack of all trades, master of none."
- On Andrew Kang's pivot to robotics: he was early to some deals that are worth more on paper, but the speakers emphasized that time will tell, that the strategy uses prop capital rather than fund capital, and that if Figure is not a good company, Kang is highly exposed to that outcome. A robotics round dominated by crypto investors can be "an anti-signal."
- The LP pressure is real: speakers have heard pushback over mandate expansion, including the existential question from one major LP — if the same deal is available through a great generalist name, "should I just be putting more money into them instead of you guys?" If a fund is unsuccessful, it will not raise another fund.
- The venture-math addendum on why hardware was taboo: a viral Anthropic dilution chart showed $500,000 of options from three years ago becoming worth roughly $20–25M — 40x for the employee while the valuation rose about 1,000x. That can work for Anthropic; for many other hardware companies raising at multibillion-dollar valuations, investors may not make money.
4. The post-2021 investable universe is finance itself
- The attribution argument: pre-2021 crypto fund returns benefited from the industry's expansion from billions to trillions, while "anyone that's made money after '21 has really earned it."
- The Carta data lands the point: the 2022 vintage's 90th percentile across all venture is 1.68x TVPI with 12% DPI. A speaker said crypto returns are better than that, but also expressed surprise at how weak regular venture returns were. Any capital market that accrues value "gets more efficient over time."
- The proposed map of what's left: tokenized assets and stablecoins mean all of finance goes on-chain — "it's in the water supply." That is a broad mandate spanning DeFi, payments, perps and centralized finance, but not necessarily the earlier categories of NFTs, consumer and social. The tweet claiming everything that failed will eventually work was "the most nonsensical tweet" because the industry has learned that some things may not fit this technology.
5. Cloudflare Wallets: guardrailed agents, stablecoin plumbing, agnostic by design
- The architecture: a human-controlled account wallet handles funding, on/off-ramps and policy, then spawns virtual wallets that agents control through API keys. Each has spending allowances, allowlists, maximum transaction sizes and overall caps. Wallets use stablecoins, with x402 enabling micropayments to attach directly to HTTP requests for APIs, MCP tools, content and inference.
- Cloudflare considered issuing its own stablecoin and choosing a single chain, then chose to be agnostic on both, betting that "the best path to transact will be different for every agent." The thesis is that value accrues to "owning the identity and the authorization layer."
- The pushback on the sci-fi framing: "we're never going to be in a world... where all this economic activity happens from agents doing whatever they want and aren't owned by a human." Delegated, controlled agents win, and the people who help manage them capture value. The Syndicate founder Will's team was hired by Aqua, according to the discussion.
6. Do agents actually need stablecoins? The bear case and the settlement thesis
- Santi's bear case, via Column-style banking: agentic spending can use TradFi rails, single-use virtual cards and rules. Most people do not realize that every Booking.com transaction already uses a new virtual card authorized for that transaction and price; Visa has also been working on tokenized payment credentials.
- The personal-opinion counterpoint is that "the traditional credit card industry in and of itself is going to stablecoin settlement over time," which is part of the thesis behind investments in Rain and Velocity. Banks are also exploring tokenized deposits for intrabank settlement: better settlement, 365-day operation and less credit in the system, "whether you call it a stablecoin or not."
- On chargebacks, P2P stablecoin transfers are atomic wallet-to-wallet, but payment companies using providers such as Bridge may batch-settle, providing some hours of recourse and a counterparty to contact. Circle launched an escrow payment protocol for chargebacks, but it "hasn't really taken off yet"; the peer-to-peer side still has trust gaps.
7. Circle earnings: don't bet against Allaire, but watch the distribution line
- The earnings read: total revenue +7% YoY, roughly $50M of net profit, adjusted EBITDA around $140M and slightly down QoQ, USDC circulation $77B→$73B, on-chain transaction volume $21.5T→$15T, Circle Payments Network volume up about 75%, and other-revenue guidance doubled from roughly $150M to $300M. The latter was described as probably reflecting accounting recognition of the Arc token presale because cash was collected in advance. Arc mainnet was scheduled for September 16.
- The bullish stance was: "don't bet against Circle, don't bet against Jeremy Allaire." The counterpoint is that distribution costs keep rising: net revenue is about 60% below gross revenue, with roughly 62% of reserve income paid away, "probably continues to grow" and remains a major headwind.
- Two market-moving details: the stock fell about 17% when USDG, the open consortium, was announced, and the story looked more haphazard after some named partners said they had not agreed to be presented as partners. The Coinbase agreement renewed on existing terms despite widespread rumors. One speaker argued that ending it would have been more bullish for Circle stock, while Coinbase receives 50% of net interest and 100% when the activity is on Coinbase.
- The public-market bull case is that Circle is the only direct way for an ordinary investor to express bullishness on stablecoins. Tether remains powerful in emerging-market retail and FX, while Circle is positioned strongly in institutional and enterprise use cases.
8. The pair trade: Coinbase at $38B versus Circle at $16B
- When forced to choose, the proposed direction was long Circle/short Coinbase, immediately caveated with "I wouldn't do that trade." Another speaker said he was long Coinbase and did not want to short crypto companies; the exchange still produced the headline "Rob shorts Coinbase."
- The valuation discussion: Circle and Visa were described as trading at roughly 20–22x forward EBITDA. Analyst targets clustered around $100 versus Circle's roughly $64 share price, though most had been revised down; even $100 would imply about 32x 2027 EBITDA. The conclusion was that Circle looked fairly priced with more downside than upside, while Visa's team was described as capable of continuing to reinvent the business.
- After a live fact-check, Coinbase was said to have 12 product lines each generating more than $100M annualized; Robinhood was said to have 13 lines. Circle's competition is chiefly the USDG consortium and white-label stablecoins, while Coinbase faces every exchange in the world plus Robinhood — "a little bit more of a knife fight" requiring expertise across many areas.
9. Western Union × Rain: trust and agents versus cash and Tether
- After fieldwork with distribution builders such as Yellow Card, the bearish view was that in parts of Africa and Latin America "people love cash... people love Tether." USDT is dominant in ex-US retail, and a previous Western Union debit card saw little use. The stablecoin TAM may be smaller over the next two to four years than many expect, especially if the CLARITY Act does not pass.
- The concession is that "you kind of only want to buy Western Union because you're buying trust." WU handles roughly $110B of cross-border principal volume across 50M yearly active customers, 200–300M transactions, about $4B of revenue and 375,000 active agents across 200 markets.
- The bull case: recipients no longer need to visit an agent location. They can hold trusted dollars in the app, earn rewards and spend online without FX fees, allowing WU to monetize users in ways it could not before.
- Rain supplies the stack — on-ramp, card issuance and wallet partners — typically for a platform or minimum fee plus potentially user-based fees and a percentage of transaction volume. It launched in 37 markets, expects 60 by year-end and is working toward WU's 200 markets within 18 months. Operating compliantly across those corridors and provisioning the required BINs would have been a major legal and infrastructure problem a few years ago; Rain and a few others, including Stripe, are now attempting it.
10. Saddam's Patek, Malka's compounding, and rewriting the Amazon
- The watch: a Patek Philippe Ellipse from the late 1970s–1980s, never sold publicly, which Saddam Hussein's Iraq used as political currency for military officials, bureaucrats, diplomats and palace VIPs. One host bid on a Loupe This auction and was quickly outbid as the price reached $26,000.
- Content of the week: Micky Malka of Ribbit on David Senra's Founders. Malka described himself simply as an investor rather than limiting himself by stage or sector, and the hosts highlighted his line: "your brand and your reputation are the real things that compound value and everything else is downstream from that." He built a company with Wences Casares and was described as an early Coinbase investor and a Morpho backer.
- Gavin Baker's Invest Like the Best appearance was also highlighted for its discussion of market jitters despite improving fundamentals, recorded two days before Situational Awareness blew up. A campaign to make Rob "the Gavin Baker of crypto" — following the Do Kwon-to-Michael Saylor progression — met with his response that this was not his general demeanor.
- The biggest historical story was LiDAR flown over 4,500 kilometers of the Amazon, which revealed 396 previously unknown earthworks and connected roads, including 15-foot-deep roads, pointing to a pre-Columbian network of cities. Housekeeping: D Asia is planned for Singapore around Token2049, D London is in mid-November, Hyperliquid's Jeff may join an onsite podcast, and Paul from Morpho is scheduled for the following week's episode.
Full transcript
Nothing said on Empire is a recommendation to buy or sell any investments or products.
Still don't have a mic. I still have a mic. But guys, it's tough times. You know, somebody was like, “Try this other room,” with also bad acoustics.
I'm testing out the acoustics, bro. The reality is, I have my Bloomberg screen right here right now, so I'm just seeing the blood bath in memory. Bloodbath in crypto. A lot to talk about.
Gents, how are we doing?
What did I miss last week? How was last week?
It was good. I think Santi is a mini Leopold, and so we had a good conversation last week about situational awareness and Santi trying his portfolio.
Rob, did you not hear from your finance department? I got margin-called, and I used the funds that you provided for Inversion to bail myself out. I'll make it all back. I promise you I'll make it all back. Just one more time, and just give me a breather. [laughter]
You immediately put 400 million into a private company after that, too. [laughter]
Yeah. I need to be less liquid. Definitely need to be less liquid.
Less liquid, for sure.
The discussion was that we thought we were degenerates in crypto, and then we discovered TradFi and Korea. But the moral of the story is, crypto has prepared us for this singular moment in finance. If you've been in crypto, this is our time. [laughter]
Unlocks. Jordi—I think Jordi and some of the guys that are good traders, like Ansem, too, were looking at SpaceX and the unlocks that I think happened today or tomorrow, and they're like, “Guys, this is lowflow high FTV. We were made for this singular moment in time.” [laughter]
I love it. I always thought crypto would become more like finance. I wasn't expecting finance to become more like crypto, and it seems like that's happening. What's going on on your side, Jason?
Life is good. I'm just busy. We're very busy integrating Messari, right? The thing that they don't tell you is there's a lot of work that goes into an acquisition post-deal. Actually, they do tell you, and people do tell you that. A lot of people warned us, but it's good.
People do tell you that. Actually, a lot of people warned you.
Thank God for Claude. There's an enormous amount of work. We just integrated our CRM. They were using Salesforce; Messari was on Salesforce, and we're on HubSpot. You have to integrate tens of thousands of records. We sell to this company, and they sell to this company. They were new with us in September; they were new with Messari in May. You have to consolidate that. There used to be integration teams of 50 people, and now, thankfully, we can do it with a couple of people and AI. There's just a lot of work that goes into integrating, so I'm spending time on that.
I will say our crypto-native business is slow, and our enterprise, fintech, exchange, and brokerage business is the fastest it's ever grown. So I'm spending a lot more time in the rooms that Rob likes to participate in, throwing on the suit, talking to the baby.
B2B. So, you can't blame me.
Oh, we can always come with a no.
I feel like I'm forced to say this because he's a board member. Rob, yes. At the next Inversion board meeting, I would expect you to wear a suit, please.
Who's this? What's this board? Is it literally just you two?
It's just a good reason for Rob and me to sit down and force him to sit down.
We haven't yet done one in person. When are we going to do it in Monaco?
I was going to say, I'll be invited back, fellas. We're going to be back in New York in September, so we should do it in person. It should be good.
What are you coming to New York for?
Just to see the team. There's a couple of deals out there, and I'm just going to go out there.
Nice.
1. Crypto Venture Expands Beyond Crypto
The first or second week of September, I think I'll be there. Sounds like—
Nice. Nice.
What are you guys looking at on the venture side? You just did this AI-spend thing, which I want an intro to, because it's a problem internally. I need to figure out how much we're spending and what we're spending on. I like that deal you guys did. It was a non-crypto deal—or I think it was a non-crypto deal.
Part of this is the infrastructure. They have 3 products. They have this thing they call Agent Studio, which allows you to build your own agents in whatever way you want. You want to build an agent to do a certain thing: How do you build it? What capabilities does it need? What API does it need to plug into?
It also has runtime infrastructure for agents that already exist for you today. That's inclusive of things like payments, and they use x42 for part of their payment stack and stablecoins. Anthropic is an investor, Accel is an investor, Gradient is an investor, and there are a bunch of others, but also us and Coinbase, as well as a few others on that side.
So it is a story around crypto and AI coming together and certain parts of agentic infrastructure. On the thing you talked about, they actually spun out their router business because, in their first few engagements, they were basically just using it internally for clients who were also using the other products. What they found was that, for their first few clients, they were reducing token spend by about 80%.
They're like, “Oh, it's working really well. There's obviously a lot of demand for this product. People are spending way too much on tokens, and the token-maxing era needs to die.” They launched that product, or came out of stealth with that product, yesterday as well. Happy to connect you as a client.
So, is this good as an investor?
Yeah. Rob's the best BD guy for his portfolio companies.
2. Crypto Venture Enters Hard Mode
Are you guys going to start doing more non-direct crypto deals? The reality is, if you look at the portfolio, I think a lot about what we're good at and also where we're not going to get adversely selected against. You're seeing a lot of the crypto venture funds go really far out of their mandate into things that are hardware-related, mechanical-engineering-related, nuclear, robotics, and so on. A lot of these investors are very smart, but they don't have experience in these areas, and there's no real reason for the companies to take their money.
Money is green, obviously, but if you have other options for money, you'd rather take money from somebody who's deep in your area, typically. What we've thought about is: Where are the areas that touch crypto, are enabled by crypto, and maybe have these big addressable markets?
A lot of our companies—the fintech companies, like Rain, for instance—have a payments company and a fintech company that uses stablecoin rails. They announced their thing with Western Union. The prediction markets on the Polymarket side, again, have the infrastructure on-chain but a much bigger addressable market. Sapien is another one of those. Venice is another one, which we had done previously. It's decentralized and private inference.
These things are crypto. They're enabled by crypto. They're applications that sit on top of blockchains, but they're not completely non-crypto. They're in the strike zone of the things that we think we're good at and are tertiary to areas that we've invested in before.
We're looking at—well, obviously I'm spending more of my time talking to traditional businesses and really leaning into just talking more holistically about what releases the most value in businesses. A lot of what I've been thinking about, not just at Inversion but generally in my conversations, even at the Selini Summit, is something Rob mentioned. The key word, the operative word, is “mandate.”
I think of the world more simply now, which is: You want to go where it's the best risk-adjusted return and absolute return. I think there's an element to what Rob is saying, which is: Do you have a right to win? Do you have an edge in a particular sector?
When you think about it, technology is such a catchall phrase. Innovation is happening everywhere. I do think two things can be simultaneously true: A firm can reinvent itself to branch into other sectors and be very successful. What drives success in venture, in my estimation, has always been having a good network and building a reputation of being a very good investor.
Technology has touched every sector in business. So, yes, your mandate is to invest in innovation and technology.
3. Can Crypto Funds Go Generalist?
Some firms have it more loosely. I think of Sequoia. You're just investing in the best founders who are going to do breakthroughs. Now, they don't do biotech as much, but they're doing nuclear, and now everyone's moving to do more hardware. Whereas hardware 5 years ago was taboo. It's like, you don't do hardware; it's super capital-intensive. You leave that to other people.
But I will say, a question maybe more for Rob is: How do you think about mandate, and just going back to people and saying, “Guys, I feel like if you've been investing in crypto, it's been hard mode”? If you're principled and you have a good team, think about crypto: you've had to be multidisciplinary. To win in crypto, you have to have a deep understanding of cryptography and be at the cutting edge of that, which requires its own expertise. You have to be really good on security and risk management. You have liquid tokens; you have to manage a position. Yes, you're in venture, but you've also had to be talking about financial infrastructure—DeFi. You've had to talk about other types of infrastructure, applications, and consumer products.
If you're a crypto-investing firm, you're competing for dollars against other firms. If you're an LP evaluating a particular investor in crypto, you would have had to ask these questions anyway. So I feel like crypto for a really long time has been a very big industry attacking many different verticals. I don't know, Rob, do you feel like you get pushback around being a crypto firm? I know you keep saying fintech and crypto. I would just go as far as saying it's investing in technology, and crypto has just allowed you to cut your teeth; it's been on hard mode. The bigger question is: Can you build a network so that you're the first call?
Yeah. So I think there's also a question around whether you're a very early-stage, pre-seed fund versus a later-stage, maturing business. I think, to your point, being multidisciplinary in crypto was really important, especially since 2021. Pre-2021, I don't think it was that important, because it was more about the trend and the wave of the entire industry. Post-2021, to your point: Do I understand the infrastructure and cryptography? What will work and what won't? Do I understand how networks scale? Do I understand capital-markets infrastructure? Do I understand all of these different things?
I think if we were to open-source all of the returns of all of the funds in the space, what we would find is that the people who have the best returns are those who built their firms to be multidisciplinary and had people who were deep in different subjects, versus trying to have individuals be generalists across all of those different areas. Because the reality is, I don't actually think you can be the best in the world and also be deep in cryptography, and also deep in capital markets, and also deep in some of the other things. That's my general perspective.
Now, to your point around generalist investors doing really well, I'll let you jump in.
4. The Hidden Cost Of Dilution
I think there are people who—we talk about Sequoia being an incredible generalist investor. I think they're the best firm in the world as actual investors. I think they're the best firm in the world, but there is a partner there who is doing the hardware. There's another partner there who's doing something else, right? And so I think that's how you have to build a firm.
The crypto firms, other than 2—Andreessen and maybe Paradigm—are all too small, and they just don't have the capabilities to go extremely generalist. So when we've talked about, “Oh, do we want to be more generalist?” the conversation has also been, “Okay, well, here are the things we are good at as a partnership. Do we need to hire somebody else?” Those are the things. And yes, people can reinvent themselves, but they can't reinvent themselves over 6, 9, or 12 months. It takes time and work, and over a period of years you can do that, but it takes time, right?
I think that's the way that we think about the world. And I think what you're seeing right now is, if I see a robotics company where most of its big investors recently are crypto investors, that doesn't tell me, A, that's a bad signal. B, maybe it's just that crypto people are more first-principles and they're early into this thing. But it also could signal that people who actually understand mechanical engineering deeply are not people who want to invest in this thing.
Yeah. Oh, there's a lot to unpack there. On the last point, I've thought more deeply now about generalists versus experts. Sometimes experts know too much to their detriment, and they won't do a deal. They just go—and they're in academia, right?
But I think it was Benchmark or Sequoia or Andreessen—all 3 of them have had a hardware guy, to your point, or a cybersecurity guy who worked at Palo Alto Networks or worked at Intel in the early days, and just has the network and knows all the founders and stuff.
But generally, on how to build the best investment firm, how much of Sequoia is just the brand that they've built for the signaling? I guess the argument around a robotics strategy, for instance, is that Andrew Kang pivoted away from crypto and started doing robotics deals. So far, I think it's been quite successful. He understands capital markets. He did the whole kind of—I mean, I'm not here to advocate; you're rolling your eyes, I get it. But I hear your point: Why is a crypto investor doing a robotics deal?
Time will tell if they are successful, by the way.
No, no, no, no.
It's about making sure that he accrues value. Now, what happens with the investments? We have time. It's going to take a lot of time.
Putting aside the robotics strategy structure, in his defense, he was early in some of these robotics deals that, at least on paper, are worth much more than what we put in.
Yeah, put in a huge check, to be fair. My point on that is, I think, how do you reinvent a firm? I think you have to do 1 really good deal that really puts you in the position of saying, “Holy shit, these guys were super early here,” and then you just juice that to build a brand and build a reputation.
I think you've seen that in AI, too. There have been some firms that I had no idea who they were, and they were early in Anthropic or early in OpenAI, and just kind of built a name for themselves, then used that and parlayed that into getting another deal. Maybe that's it—you guys have done some AI deals recently. Do you see a world where you're hugely successful in some of your AI deals and then you just say, “Oh, shit, we're just going to go do more AI deals,” that may or may not have a crypto strategy to it?
There's another thing that people are misunderstanding a little bit about some of these deals, and this is happening in AI, too, which is also venture math. The reason people didn't do hardware before is because these things had to raise a bunch of money, and the valuation could go way up.
Yeah, and there was a chart that was going somewhat viral on Twitter earlier this week. I think it was about Anthropic, but it was like, “Oh, if you put—”
The dilution.
Yeah, the dilution, right? And obviously, it's still been incredible because they've had such incredible growth. But it was basically like, if you got $500,000 worth of options 3 years ago, it's now worth $25 million or $20 million, I think. But that's incredible growth, right? You've had 40x growth, but the valuation in that time is up by, like, 1,000x, right?
And so, that's incredible growth. You've had 40x growth, but the valuation in that time is up by, like, 1,000x, right? For Anthropic, that's totally fine; it makes complete sense. Now, all of the other hardware companies that are raising—these venture companies, the vast majority of them will not be fine. And even if you're in a situation where you exit and it's a good company over time, many of those investors who are investing in these rounds at multibillion-dollar valuations are probably not going to make any money, or may lose money, depending on the outcome.
I want to make 1 point here. I've been talking to—I was traveling last week; I was in a hotel when we chatted because I was talking to a bunch of our LPs. I was talking to some LPs the week before, and we share LPs with a lot of different people. Then we talk to new prospective LPs or a lot of different people, and this trend of really expanding the mandate beyond the things that are just within our strike zone, or a little bit tertiary—we have started to hear pushback from the LPs and complaints about other funds doing that right now.
Now, the reality is, no GP should be investing based on what the LPs say. We should theoretically be the subject-matter experts, but—
If you're not successful, you'll never—you won't raise another fund, right? And so that is the risk that people are taking who are funds. Andrew Kang's a little bit of a different story because it's all prop capital, so it's not fund capital. But if you're a fund manager, that's something you have to worry about. And we've even had—I even had one discussion with a really good LP of ours, a really big brand name.
We love them. They have even asked us the question at times: “Well, if you’re doing the same investment with this other person who is a really great name—it looks like a great deal—but should I just be putting more money into them because I’m into this big traditional generalist instead of you guys?” That is the question people ask, and you have to be able to answer that appropriately.
Yeah, I think I have a pretty firm belief here that you earn the right to be a generalist. If you are everything for everybody, you are nothing to nobody.
It’s important—you guys mentioned Sequoia. Sequoia started as a niche firm. They had a thematic focus on technology and electronics. Don Valentine was at Fairchild and National Semiconductor. They were the smartest in the room when it came to semiconductors and chips, and they earned the right to go generalist over many years of having success.
I don’t think they went generalist until, like, the 8th fund, right?
Exactly. They earned the right. I like what you guys are doing at Dragonfly because you did extremely well in crypto. You were a top 3 fund in crypto, you earned the right to go do payments and other stuff, and from the early look of it, it seems like you’re being successful with payments right now and earning the right to go do more things.
What I really don’t like is when funds and companies, too, say, “There are other shiny objects over here. Let me just go try to chase those things. Now I’m a generalist firm.” That’s a real recipe to end up a jack of all trades, master of none.
I also think there’s a brand element here, right? You have 2 customers in venture. You have your LPs, but you also have the founders. You have to think about the brand to the founders, too. Do they want you on the cap table? The signaling is spot-on. Rob, I know exactly the deals you’re talking about with these robotics and crypto funds. To me, that’s an anti-signal. I’m like, “All these other—”
Yeah. It’s an adverse selection problem. If you’re seeing it, it’s because all the real robotics funds passed on it.
Exactly. However, Kang—I don’t know him super well. He seems like he did very well in crypto and is now doing well in robotics. Time will tell. If Figure is not a good company, then he’s not going to do well in robotics because he’s extremely levered to Figure.
But I do like that he has picked another niche. What I wouldn’t like is if he said, “I’m doing robotics, and I’m doing this and this and this.” You have to pick a niche and own the niche. I feel pretty strongly about that.
Yeah. That’s really—
Yeah. And I don’t think this is meant for us to single him out. Obviously, it’s just—
He’s actually going to come on Empire soon. We’re going to have a long chat with him.
And Rob’s strategy was obviously somewhat early to this, but it’s obviously top of people’s minds. I think it’s endemic to your point across the industry.
And so it’s something that I think a lot of people are trying to figure out right now. I like the point, Yano, that you made, which is that picking another niche is important because you have to become an expert on these things, or at least an expert enough. You have to become dangerous enough.
I think, Santi, maybe you made the point earlier about how much of Sequoia today is the brand versus the investors. It’s a little bit of both because they can hire the best people because of the brand, and they can help legitimize companies because of the brand.
Micky Malka was on David Senra’s podcast.
You stole my counter of the week.
Yeah, sorry. One of the other things he said that I really liked—I was listening to this morning when I was at the gym—was that the only thing you really have that compounds value over time is—yes, money compounds value, but your brand and your reputation are the real things that compound value, and everything else is downstream from that. I loved that statement by him, and I think it’s exactly true in all of these funds.
Before we go into the news, I want to get your guys’ take on something I’ve been thinking a lot about: the opportunity set in crypto.
You hear this about the 2021 vintage and after—that they were terrible for venture, and not just in crypto. You said something earlier, Rob, that reminded me of what I’ve seen. I’ve been an investor in many crypto funds, and I could tell you the returns of most of them. In venture, pre-2021, it was a factor. If you do the attribution, crypto went from being a billion-dollar business or industry to, by 2017, a couple of billion, and then it ballooned to a couple of trillion. How much of that was just tracking Bitcoin and the growth of the industry?
I feel like now, anyone who has made money after 2021 has really earned it—really earned it—because it’s incredibly hard to make money. I just think more of the question in crypto is: What is the investable universe? After 2021, there have been only very few pockets that have done well. If you weren’t in those names, you not only underperformed, but you just lost money.
I’m curious how you think about the investable universe. We talked last episode about fund sizes. Is there too much capital in the system chasing fewer deals? I know larger deals are getting done, like Kalshi and Polymarket, but then you’re also competing against other firms that have not historically been crypto firms.
The second follow-up is: Are you getting pushback from LPs around that? Are they saying, “Hey, maybe you shouldn’t be raising as much of a fund of this size”?
Yeah. We’re not actively having those discussions about a new fund size. We announced the 4th fund earlier this year, so we’re not out in the market. We haven’t heard that yet, but we’re still obviously in active dialogue with all the LPs. There is just a lot of interest in AI, and it’s kind of sucking up all of the interest at the moment.
I was thinking about the point earlier, too, about generalist versus non-generalist. I was talking to a venture recruiter because we’re hiring for a few roles right now, and one of these guys was telling me that all of the interest across the Valley is in people with deep-tech expertise. All of the firms are trying to hire people who actually know deep tech. We’re seeing this on the fund side as well.
On the specific question about the investable universe in crypto, I think what is happening is that tokenized assets and stablecoins—and just all of finance—are being put on-chain. If all of finance is being put on-chain and it’s in the water supply, that means that everything that produces and distributes value—all of the applications—is going to touch crypto.
In one sense, that’s a very wide mandate because that’s all of finance globally, and that’s an incredibly big addressable market. But in some sense, it is finance. It includes DeFi, payments, centralized finance, perpetuals, the future of markets, and so on. It doesn’t necessarily include all of those other categories that we used to talk about before, such as NFTs, consumer, and social.
Somebody put out a tweet a few weeks ago that was like, “Everything that didn’t work in crypto before will eventually come back and work.” That was the most nonsensical tweet because we’ve learned something since then. We’ve learned that maybe those weren’t the right things for this modality, this piece of technology, and that’s fine. We should learn from the past.
I would make the more specific point to you, Santi, that everybody says it’s been hard to earn money since 2021. The new Carta data came out earlier this week and was everywhere on Twitter. If you look at it, the 90th-percentile 2022 vintage is a 1.68x TVPI and a 12% DPI—distributions to paid-in capital back to those LPs. This is all venture, not crypto-specific.
It shows you that regular venture is very hard, and making money in this space was actually much easier before. It now just looks more like regular venture, which is also why the deals look more like regular venture and why people are trying to figure out how to change the way that they invest in the future.
I was surprised by how bad some of those returns were on the regular side because I’m hearing crypto and thinking, “We’ve got returns that are better than that.” Then we’re like, “Well, it’s not as good as the 2018 vintage, right?”
But I think any space—any capital market—that is accruing a lot of value, where people are able to make money, gets more efficient over time. That’s just where we are today.
Long gone are the days of 10x in 2 years. Damn, man.
I think so.
Should we pack up our bags, punt the Nasdaq, and call it a day?
Listen, if you could just go 4x, lever long Micron, you can take [laughter] 10 million.
5. Ads (Peaq)
No doubt. No doubt. That, ladies and gentlemen, is a dig at situational awareness. We are aware of the situation. So is Ken Griffin. Not to get too much into that.
Let's get into some of the news of the week.
Robots and machines will outnumber humans onchain and peak is how they get there. Peak OS enables any machine to do business on any chain. It turns machines into autonomous actors and liquid assets with ease, giving robots access to capital and the ability to compound their value. That means funds, allocators, and institutions can underwrite, finance, trade, and route capital to machines the same way they handle any other asset. Millions of new potential consumers for any web3 service, from compute to storage to anything a machine may need, without a human in the loop. Head to peak.xyz or click the link in the show notes to find out more.
6. Cloudflare Bets On Agent Wallets
So, the first one—I don't know if you guys saw this—it was Cloudflare Wallets, so I'm going to share my screen. I actually went and claimed my Cloudflare username, which is not something I thought I would ever do.
Basically, Cloudflare announced Cloudflare Wallets. It is a programmable wallet system. A lot of this is just reading from their blog, designed for the agentic internet. Basically, how it works is the human controls the account wallet, which is your fund storage, the on- and off-ramps, and policy setting.
Then, from your account wallet, you can actually spawn these virtual wallets that are agents, which your agent controls via API keys. Each virtual wallet comes with a guardrail, which is a spending allowance, an allowlist, maximum transaction sizes, and overall caps.
Imagine, Santi, you create your wallet, and then you have a virtual wallet for buying software on the internet, like buying API keys or something like that. Then you have another wallet for trading, which is plugged into Coinbase and Robinhood. Then you have another wallet that is mainly for reading news articles and consuming news or something like that. So, yeah, pretty interesting.
Is this like cookies are dead? Is this the new cookie? You have a wallet, Jason. It knows—everyone knows—what you're spending money on, and marketers love this.
I think it's not just the cookie. I mean, “the cookie is dead” has been something people have been saying for many years. I think it's another data point that people can use to target people. You can use credit-card data; you can use, obviously, your browser history.
I think there's a lot of really, really rich data in wallets, and it's actually something that Blockworks is looking into at the moment, because there's so much rich data on-chain that is really not being used right now.
If you can stitch it. But, I mean, for sure, right? If you're a marketer, crypto's always been pristine data.
Yeah, exactly. And look, it's cool because the relevant thing, and why we're talking about this on Empire, is that each wallet stores and spends using stablecoins. There are these claimable, human-readable handles, but then everything else is agents.
It also integrates the x402 protocol, so micropayments can actually be attached directly to a web browser, like an HTTP request, right? If you want to pay for an API, an MCP tool, content, or AI inference, that's all just done with stablecoins over the x42 protocol.
It's pretty interesting. You can go play around with it right now. I went and set up my handle. And, Jason, just a moment on that, because I know you're so plugged in and you love Robinhood. Have you been poking around with their whole agentic capabilities and instructing agents to buy on your behalf?
I have, and you know who I saw posting a lot about this? Tom Shaughnessy.
Tommy. Yeah, yeah. I want to bring maybe Tommy on and have a conversation about that.
Yeah.
So, any thoughts? Bullish or bearish on this Cloudflare announcement?
No, bullish. I mean, can't be bearish. [laughter] Yeah, I mean, bullish, just generally. I think the future of economic activity is going this direction. Twenty-five percent of the internet runs through Cloudflare, and they've been releasing this data that I think we're all up to between 55 and 60% of all internet activity being bots today, right?
It's very clear that the world is going this direction. I think there's nobody you can talk to today that doesn't think agents will play a real role in transactions in the future. Now, what modality that is, and how and who and all of that, is still very much up for debate.
I think this is very cool. I know the guys on the Cloudflare team who are in charge of this stuff, and they've been working on it for a long time. They've been thinking a lot about, “Do we need our own stablecoin? Do we use other stablecoins?” They eventually decided that they're going to be stablecoin-agnostic.
They've been talking about which blockchain they should use, and they eventually decided they're going to be mostly agnostic there, too, or at least support a lot of the big ones. Their perspective has been that agents are going to do many, many different things. The right way to transact, the best path to transact, will be different for every agent and for a lot of different uses.
They want to be at the center of owning the identity and authorization layer. I think that's actually where the value is going to accrue, because we're never going to be in a world—I know people like to be in this sci-fi area where all this economic activity is going to happen from agents that are doing whatever they want and aren't going to be owned by a human. I don't actually think that's true.
I think it'll have to be owned by a human, and you're going to have to have these delegated agents and controls in place. But there will be a lot of value in the people who help you control and manage those agents, and that's actually part of the thesis of the safety of the investment that we did.
Stablecoins are going to play a big role here. They're not going to be the only thing people or agents transact in, but they're going to be a part of that. I'm excited to see what they do.
Two observations here. I love seeing people who are former founders, like Will, the founder of Syndicate. Smart guy. I think I met him. I ended up investing, but I really liked him. I wasn't aware that he was at Cloudflare, which is crazy. We're going to talk about that later.
Aqua hired the team.
Yeah. Oh, Aqua hired the team. Interesting. Yeah, good for him. One—where was I going with this? I'm curious: We talk about programmable money as uniquely enabled by stablecoins, right? But you can have wallets; you could have credit cards.
Is there a version here that is actually bearish? To your point, Rob, do you believe in a world where agents just transact in stablecoins? I feel like a lot of the narrative has been that agents need stablecoins to transact because, A, they're not human, so you can't open a bank account, and, B, it's 24/7/365.
Is there an argument to be made here along the lines of what Column Bank says? Like, no, wait a minute. We can do this in TradFi. You just set up specific credit cards or debit accounts for the agent, and you have rules and customizations in place. You don't need a stablecoin; this can work with the current rails. I'm just trying to play the bear case here.
I mean, you can do a lot of these agentic workflows with tokenized, single-use virtual cards.
For anyone not familiar, Rob, that means not crypto.
Yeah, yeah, right. It's like when people think about their credit card, they think about this thing they have in their wallet that they pull out, or maybe it's in Apple Pay or whatever. But people don't realize this: In a lot of payment gateways now, depending on the type of business—OTAs specifically—if you think about Booking.com, which is an OTA, every single transaction that goes through Booking.com is actually a single-use virtual card.
They give you a net-new number, and it's specifically allowed to do this transaction at this price, right? You think you're putting in your credit card, but it's actually funding a net-new credit card. That's the way the infrastructure works there.
A lot of transactions work this way, and Visa has been working on tokenized credentials and tokens for payments to do these things—tokens not as crypto tokens, and not tokens as compute, but for payments.
I think where we've come out—and there was all this conversation earlier, and Jeremy, I think, dropped something this week too, about how stablecoins are here for the age of agents—is that there are types of payments where stablecoins make the most sense: direct, peer-to-peer stablecoin payments. And there are types of payments where they won't.
But this is a personal opinion. I wouldn't say this is a thing that's necessarily broadly believed, but I believe that the traditional credit card industry in and of itself is going to stablecoin settlement over time. And I think Visa has shown that they're investing heavily in this. Obviously, that's been part of our thesis, investing so heavily in Rain and some others, Velocity, which we announced recently, because I think that from a settlement perspective, it's way, way better.
And I also think, by the way, to your point, Santi, the counterargument is like, “Oh, well, we can do all of these things.” That might be true for a lot of things domestically, but we're also hearing now from some of these banks that they want to be able to do tokenized deposits. It's for interbank settlement. So it's still—it's so it's still maybe not a stablecoin, but they understand that to get better settlement, to run 365, and to not have so much credit in the system, there has to be some sort of tokenized asset there in the middle. And I do think that is where the world's going, whether you call it a stablecoin or not.
A lot of the—I know we've got to go to other news, and speaking of Rain, the Western Union stuff, of course—but on this point, for people that are coming out, I feel like a lot of the criticism is, “Wait a minute, what about chargebacks? What about, if it's an immutable transaction, how do you figure all that logic, which is not simple?” Has there been a clear, easy solution to all this stuff that maybe the folks at Visa and others have engineered around stablecoins and escrow or something to allow for this? Because chargebacks are a big component of online transactions and whatnot, or just regular transactions.
Yeah. I think the reason Cloudflare has set it up the way they've set it up—which is, you have to on-ramp as a person, then the person can on-ramp the money, and then you can set up the agent in the way that you want to set it up and give it parameters—is to try to protect themselves, insofar as if you need to have a chargeback and the agent misspends $100 or whatever, you've taken on that risk.
Because today, in the peer-to-peer transaction side of stablecoins, that doesn't exist. Now, that's been kind of crudely solved in a couple of ways. A lot of these companies that actually do stablecoin settlement just batch settle. So we think that stablecoin transfers are atomically settled, and that is true if I take my non-custodial wallet and send it to your address, Santi. But if I'm a payments company using one of these service providers, Bridge or whoever else, they're actually batch settling. So if there's a complaint, you have some amount of hours before that batch settlement, and you also have a counterparty that you can go talk to.
The other thing that's happening is Circle did actually launch an escrow payment protocol for chargebacks, and there are a few other people trying to do this. They haven't really taken off yet. This is, again, why I think—and this is also why you pay for a lot of things that you pay for when you pay for credit cards, et cetera.
But it's again why I think that there will probably be a situation in the future where the payments for agents will be done in a bunch of different modalities, depending on size and quantum, and if it's in a closed-loop system or an open-loop system. Is it e-commerce? Is it enterprise? Cards are going to play a really big role in that. Stablecoins will play a really big role in that, both as settlement and as peer-to-peer, but the peer-to-peer side does have further to go in some of these trust factors.
7. Is Circle Stock Mispriced?
Yeah. Let's talk Circle. Circle announced earnings.
They've got to fix the package. Are you going to read through all that, sir?
No, but I'll give you the highlights, like the good host that I am, Santi. I would call these earnings mixed results. I think the stock is pretty flat. Total, though, it outperformed what a lot of crypto-native folks expected.
I think, as I've talked about this before, don't bet against Circle, don't bet against Jeremy Allaire. I think a lot of people like to pile on to the bearish Circle takes, and I think those are generally the wrong takes. I think the business is more diversified than people realize.
But, yeah, some of the highlights: total revenue grew 7% year-over-year. The company's now doing like $50 million of net profit. Adjusted EBITDA is like $140 million, which is a little bit down quarter-over-quarter. Core stablecoin metrics started to reverse, so USDC circulation fell from $77 billion to $73 billion. On-chain transaction volume dropped from $21.5 trillion to $15 trillion.
I'd say bright spots are that the Circle Payments Network volume surged about 75%. I think they raised other revenue guidance from about $150 million. They doubled it to $300 million, though I'm pretty sure that jump, I was reading, is an accounting recognition of the Arc token presale because cash was pre-collected. And then, yeah, the big thing is they have Arc mainnet scheduled for about a month from now, September 16th. So, yeah, curious to get your take—
On Arc or on Circle here.
Well, the stock—just to lay a little bit more color on that—the stock's down like 60% since—well, more than that. It's up 6%. It's down almost 80% in the last year.
Yeah, but they also had that completely absurd IPO.
I'm looking at the candle. Yeah, well, it's still—I'm not going off the Burj Khalifa candle. I'm looking at once it's sort of normalized. It was at like $100, $200 bucks. Now it's like $65. So call it down 70%.
If I look at the analyst targets, some of the analysts came out, like, to your point, the stock kind of is down pretty—it feels like it's bottomed here at like $57. Who knows? I'm not a chartist. I'm not looking to be one. But a lot of the analysts—the commentary was similar to what you're saying, which is, like, I feel like the market got pretty jittery. It went down in May.
Because people were questioning the defensibility of the network. It's only going to get more competitive with stablecoins. Rob, I don't know if you agree with this, but I feel like a lot of the worries were when USDG, the open consortium, launched—
The market got pretty spooked by that.
Yeah. And so I think now it was interesting to see the analysts say, like, “Hey, I feel like the market's overreacted. They do have defensibility. They have other partners. It's not so easy to unseat.”
But I think that continues to be the bear case for Circle, which is that it's only going to get more competitive from here. And, you know, if I look at the analyst targets, a lot of them are in the $100 range. Some of them revised down—actually, most of them revised down, which I found interesting—but the stock's sitting at $64, and most of the analyst targets are at $100. So, okay, you have some upside there, pretty dramatic, like 60% upside if you believe it, but that's still like 32 times EBITDA of 2027. That's not cheap.
I'm curious, on a relative basis, would you rather be buying this versus Visa, Rob?
I think Visa is an incredible business, and I'm actually very bullish over time. Listen, I think on Circle, yeah, obviously the market really spooked after USDG. I think the stock was down 17% the day that that was announced. There has been a little bit of a relief rally, but it hasn't regained all of the loss from that point.
You know, I think it didn't help USDG that it came out—some of the partners were like, “Actually, we don't want to. We're not really partners. We were just told about this, and we didn't know that our logo was going to end up on this thing,” and some other stuff that came out after that. So obviously it made it look a little bit more haphazard than it seemed at first glance.
And so I think they've benefited from that. Listen, I think for Circle, it's the only way to directly express bullishness on stablecoins if you're a normal investor. There are people like me, obviously, private-market investors, and there are certainly other ways, prediction markets and things like that. But the only way to directly express it is to own Circle stock.
And Circle is the leader for all of the use cases people are excited about. Obviously, Tether has this incredible business in emerging markets and with retail, but they're certainly not winning the institutional use case. That is what we're seeing people get really excited about, and also the enterprise use case.
And so it makes sense that people are excited. I mean, I think the one thing you didn't say, Jana, when you talked about their earnings, which I think is still the most important thing on their stock, is that the amount that they are paying for distribution keeps going up every quarter.
Their net revenue is about 60% less than the revenue that you mentioned in the gross revenue, and they're paying on their reserve income—they're paying about 62% of that away today. That probably continues to grow, and that is probably one of the biggest headwinds to the stock right now from a revenue perspective.
At the same time, the good thing has been that we're more constructive about the rate curve at the moment than we've been in a while because of the economy.
But also, the other big thing I missed is that the agreement with Coinbase—which everybody was saying was about to end—has actually just been renewed, I think, on existing terms. That means USDC remains core and central to all of the Coinbase products. I think that was different from what the market expected, so I'd say that's big.
Yeah, there were definitely a lot of rumors about what was happening there. Before USD1 got announced, there was a lot of debate about how involved Coinbase was in the USD1 product as well. It never really made any sense to me that Coinbase would want to get rid of that deal, because they get 50% of the net interest, and they get 100% of it if it's on Coinbase. There was no economic reason for that ever to have happened, so I didn't quite understand that. But, yeah, honestly—and I said this in the post I made about USD1—I think it would be much more bearish for Circle stock if that had ended.
It's much more bullish for Circle stock if that had ended. It would have been way better for them.
Yeah. Okay, Coinbase is trading at, like, $38B or $39B. Circle is at $16B. So Coinbase is trading 2.5 times higher than Circle. If you had to put on the pair trade, what are you both doing?
Neither is an option, Yan, dude. I don't own any HOOD directly. I have some exposure to it in other ways, but I don't like either. I'm a stablecoin bull, so I'm going long Circle.
Well, long Circle, short Coinbase at $38B and $16B.
If you were asking me, if I'm given the choice that I have to do a pair trade, those are the directions I'm picking, but I wouldn't do that trade.
Well, fellas, what is interesting—
Headline? We're making the headline. Let's make the headline: Rob shorts Coinbase.
Yan is trying to get me in trouble right now.
Dude, Rob comes on this podcast and has a couple of lines that he needs to hit to be on the headline, and I think he's been pretty good at it.
I'm taking the long Coinbase. I'm excusing myself from the pair trade that I gave to you both because I don't want to short any crypto companies.
I'm not short, by the way. I'm not shorting anything ever.
Me too. I'm long Coinbase. I disagree with Santi here.
If you want to hear, there are great postmortems on situational awareness. I guess a lot of it is pair-trade related, so I feel like you're going to continue to hear that. One interesting thing on this point: I asked you earlier about Visa versus Circle. What's interesting is that, at the current levels where the stocks are trading, on a forward basis they're about the same—20 to 22 times forward EBITDA. The market just kind of brought Circle down to gravity, and the analyst targets are expressing that Circle's multiple expands from here, which I don't love.
For that reason, I think the stock is rightly priced at the moment, and I see a bit more downside than upside, candidly, because crypto names are more reflexive. Visa, I still think, is going to continue to make inroads into this. Its team is very cracked, and they're going to continue to reinvent themselves.
But the thing about Coinbase, though—can we just talk about that for a minute? Robinhood, for example, has 9 or 10 different business lines that are cranking out, like, $100M each.
13.
13. All right.
But, by the way, Coinbase has 12.
Okay, I was going to ask you that.
Coinbase.
Interesting. Okay.
Wait, not all of those are doing $100M in revenue, which was what Santi said.
Yeah, that's—
Let me look it up. Let me do a quick fact-check on myself. Hold on.
Yeah, I think you're giving the total business.
I would be shocked, dude. Talk about integrations, right? You're going through it right now. I'm sorry. Coinbase's product is not at the level.
No, no, no. Coinbase—we're growing new revenue streams. We have 12 product lines, each generating over $100M annualized.
Okay, there.
Okay. All right.
Good for them. I think I learned something as a business point. Santi is making a very good financial stock point on the multiples. There's also a business point: if I think about competitors today, who are Circle's competitors today? What kind of scale and size do they have? How far along are they?
The thing that is probably most competitive to Circle today is the USDG consortium. Then there's this trend—we'll talk about Western Union later—of people like Western Union launching their own stablecoin, a white-label stablecoin. That's Circle's competition.
8. Western Union Goes Onchain
The competition for Coinbase is every exchange in the world, and Robinhood, because they're trying to be the everything exchange. It does strike me that it's a little bit more of a knife fight on that side of the market. The things they are doing are so much more varied and require a lot more expertise across a bunch of different areas than what Circle's trying to do.
Rob, you want to hear something funny? Speaking of Western Union, you might actually win the bet. The stock is down quite a bit, and we're going to settle in November. I actually feel bearish after doing all this work on remittances, as you guys know.
But to your point around the stablecoin, I came into it because Western Union's CEO talks about working capital and prefunding and all this stuff. I did a lot of work just talking to people in crypto who are building distribution networks—boots-on-the-ground kind of stuff. Folks like Yellow Card and others, really understanding where the demand for stablecoins is.
I don't think that a lot of people in the Western Union customer base want this, because they've had a debit card before and kind of no one really used it. People prefer cash. In my mind, it just comes down to pockets in places in Africa and Latin America. A, people love cash. B, people love Tether. People love USDT.
That's the other thing that we haven't talked about here. Circle, and anyone else who is going to compete against Circle—USDT for retail and for FX is king. You have to then believe: How big is the addressable market today, and how quickly does it materialize? Maybe the CLARITY Act doesn't pass, and the odds are pretty bleak there.
I feel like, in a perfect world, the biggest opportunity for stablecoins continues to be outside the US, continues to be retail, and Tether is capturing the lion's share of that. My point is that TAM is not as big as people want to believe over the next 2 to 3 to 4 years. I can't look farther out than that, to be honest.
Because I need to make a decision whether to buy the stock, I disagree with you generally because of the point I made earlier about settlement infrastructure. Again, I have this perspective that when you talk to the networks and a lot of the correspondent banks, it feels very clear to me that we're going to tokenized asset settlement. That's very unsexy, and people don't love to talk about that in the same way as a peer-to-peer transaction.
I think you make a good point on Western Union's side. Western Union's got—I mean, they do about $110B of cross-border transfer volume right now on a principal basis, across about 50M yearly active customers. To your point, they've tried debit cards and stuff in the past, but it's been hard for them to launch globally and hard for them to roll out in a concentrated way that also incentivizes people to stay on the platform versus outside of the platform.
It's also taken time for them to expand that infrastructure and build it out. What changes here is that people don't really need to know that it's USDP. Is that what it is? Yeah, USDT.
USDT.
Yeah. Because what's happening is that the on-ramp is mostly going to be the same thing today. Now, you can also send a stablecoin—or you have to send USDT, or there'll be swaps for that—and you can just be a customer of the app today.
That's not really what they're focused on today. What they're focused on is that they have these 50M active users who send 200M to 300M transactions a year from places like the US and Europe to emerging markets. Right now, they charge those people, and the person in that emerging market goes to the agent location and gets cash.
Now they don't have to go to the agent location to get cash if they don't want to. They can get it in the application, get revenue share or other rewards for owning the stablecoin, and they can also get rewards in the course of owning or using a Rain card relative to what they were able to do before.
Now, in some areas of the world, they'll prefer cash. In a lot of those areas, they might prefer cash for their spending, but what they will like is that they can hold a US dollar in a form factor and from a location or a company that they trust, and can now spend it online without FX fees. They can spend it in any type of location, and now for Western Union that means they can go and monetize those users in a way that they couldn't before. So that is the story.
They're also talking about treasury-management stuff for saving costs, but that is the bull case for Western Union, even though I might win that bet.
Yeah. No, look, I think, to round this out, you're absolutely right. I think that is the bull case. We just didn't feel comfortable proving that out. I'd rather have Western Union prove that out in the public markets. If that's all well said and done and true, great.
I talked to one of the most notable crypto companies working with Western Union's network, and he said you kind of only want to buy Western Union because you're buying trust. Any other smaller player is a shittier business model with the same regulatory footprint and requirements, but Western Union has a lot of trust. Hopefully they can pull it off.
One of the most recognizable brands in the world in money and remittance. You know, I just gave you a little bit of the data. They did $4 billion of revenue last year.
And so, I think it's 375,000 active agents now globally. They have a network. We talk about networks all the time. They have a network.
To see them doing this with Rain was exciting.
Yeah. Can you talk a little bit about the Rain stuff? I know you guys are investors there, but where does Rain make money here?
Yeah. Rain is actually providing the entire stack for them. They're working with Rain on the on-ramp, they're working with Rain on the card-issuance side, and they're working with Rain's partners on the wallet side.
Each one of these deals is independently negotiated, so I won't talk about it specifically. But the way these usually work is that when they provide all this infrastructure, there's some sort of platform fee, or a minimum fee, or something like that. They also get fees based on potentially the number of users and a percentage of transaction volume.
For a company like Rain, which is uniquely able to do this product because it can go globally, I think on day one—or yesterday—they launched in 37 markets. I think they'll be in 60 markets by the end of the year. Western Union is in 200 markets in total, and so they're working toward all 200 of those over the next 18 months.
The ability to operate compliantly and legally in all those different markets, to do so in US dollars or potentially in multiple currencies, and potentially to use BIN numbers—which, if people don't know what a BIN is, it's the bank identification number required for each individual card that can be used in the US or in other countries—is a huge technological feat.
It's an actual thing that has only been able to happen over the last few years. The ability to do that and open up all these different new geographies is basically a brand-new thing that's happened over the last couple of years. That's why you see people like Western Union so excited about it, because it would have been an absolute compliance, legal, and infrastructure nightmare a few years ago.
Rain, and now a few others who are trying to do this—Stripe is trying to do it too—is uniquely able to allow people to open up all these different corridors with these products.
All right.
You know that I'm an investor, so bullish.
I asked—I said, Rob, what do you want to talk about today?
He goes, "I don't know. There's a Koshi [?] thing, which is like the Kawhi lawsuit, and Rain." I was like, "Oh, you don't say. You don't say, Rob."
I respect it, Rob. You're out here.
[Laughter] You're out here hustling.
Look, we saw there was—we announced this deal in Velocity and this kind of payments company that's doing this acquiring stuff. The founder went on Twitter, or one of these podcasts, and called me his CMO. So, you know, I'm just saying—
High praise. When was the last time Rob's talked about inversion out here, ladies and gentlemen? [Laughter]
He's going to flip it back. He's like, "Well, do noteworthy stuff," you know.
I almost said something like that, but I wanted to let you do it.
9. Watches And Content Picks
Before we do content of the week, Santi, I have a watch you should buy.
Okay. Oh—oh, that, the Saddam Hussein Patek.
How did you—
Did I not read your mind? Oh, there you go.
Look at how incredible. There's a Patek Philippe Saddam Hussein collaboration. So, for folks who are not watching on YouTube, there's a Patek Philippe Saddam Hussein collaboration.
I—you know, it turns out in the 1980s—
It's an Ellipse, not a Travel. My bad.
Yeah, beautiful watch. I mean, the Ellipse is—if you want to get into the watch game, I really dislike this idea that you have to buy and spend a ton of money. You can find beautiful vintage watches. This included, obviously putting aside the Saddam Hussein attachment, but the Ellipse is a fantastic watch, in my opinion.
So, from the late 1970s through the 1980s, when Saddam Hussein was leading Iraq, they were a huge corporate client of Patek Philippe. Patek would give these watches; they were never actually sold to the public. They were given as political currency to Iraqi high-ranking military members, bureaucrats, diplomats, and VIP visitors to the presidential palace. Hussein would be able to give out these watches to VIP visitors.
Yeah, there's actually this—it's an amazing website, Loupe This.
It's like a new watch auction website, but anyway, they've got—
Yeah, they've got one for sale, so I thought of you.
This was not owned by Saddam. It was probably owned by some dignitary or something. Someone was hosted, and then they got it.
So I bid. I've quickly been outbid. It is now the current bid—
Oh, yeah.
$26,000.
Did you actually—
Oh, yeah. Yeah.
Ladies and gentlemen, Jana's printing. I bid at the very beginning, when it was not $26,000.
It's all right.
After we get off, you're going to put in another bid.
I'm tapped out. Can't be doing $26,000 in a bear market for—
That's not founder—
For a Saddam Hussein Patek Philippe watch. [Laughter]
He's like, "Sir, you're welcome to the United States. What do you have there on your wrist?" [Laughter]
Well, actually, officer—
He's not going to be allowed back in. He's—
I mean, at least allow me to go back to America, and I want to record some impersonations, guys. You can't be flexing around a ton.
Can't be doing that. Can't be doing that. All right, content of the week, folks. What do we got?
Ah, you said it. Rob stole it from me. Micky Malka, Ribbit founder, phenomenal investor. I think he's one of the best investors. The Founders podcast. Worth a listen. I won't say more than that.
They talk about his art collection?
Yeah, they talk about NFTs. They talk about his art collection and the idea behind it. I haven't finished it. I'm three-quarters of the way there, but really good stuff.
Yeah. He talks a little bit about—they did this digital art. He and his wife funded this digital art gallery.
Oh, yeah.
And, yeah, they talked a little bit about people. He talked a lot about crypto, actually.
They're very long on crypto. They were seed investors in Morpho, and they're huge crypto investors. I think they were late investors in Morpho, but they were early Coinbase investors. I mean, they've been around for a long time. He's done an incredible job.
Well, he built a company with Wences Casares, who is obviously a Bitcoin OG, OG, OG, you know.
Yeah, he's done an incredible job. To your point, he talked a lot about, "People ask me if I'm a seed investor, and I say yes. They ask me if I'm a late-stage investor, and I say yes. They ask me if I'm a fintech investor, and I say yes."
"Nuclear investor, I say yes." His point was, "I'm an investor, and I figure out how to do all these things." So he really backed up your point earlier.
Well, that's—and I think that's really what you said earlier: reputation compounds. I feel like he had a total outpouring of founders who retweeted David's post, saying things like, "Yeah, Micky Malka—the guy's just going to battle with you. They're amazing to work with, and they really listen." It's actually pretty impressive to see. Amazing, amazing. We should have him on the podcast if he wants to come on.
He's been on as well.
Yeah. I can only hope that someday people talk about me the way they talk about him. I love the stuff that he said about brand and compounding, and I think the thing about him, too, is that he opens his doors to everybody and is just an incredibly nice person as well. There aren't really bad things you can say about him. No doubt.
Rob, content of the week.
Gosh, I don't know how I follow that up.
You pick your random Rotten Tomatoes horror movie. I have now a Rotten Tomatoes content-of-the-week predictor, man. I have 3 theories here.
I'm seeing new horror movies this week. There is a new movie I'm seeing this week, but I won't pick it. I was going to say Gavin Baker was back on Invest Like the Best, talking about a lot of the AI jitteriness in the market. I think that was very timely, following what happened with—he was like, “I can't explain what's going on. Fundamentals are getting better.” I think he didn't know, but it was 2 days before Situational Awareness blew up, and it was very interesting to listen to it after what happened and in the context of the 2 of them together.
Rob, we need—
We need that kind of energy from you because I'm going to be bearish—people know that, no surprise. Any opportunity I get, I can be bearish. Yano is sort of in the middle; he needs to be neutral. You are the guy that needs to be the Gavin Baker of crypto: the market is wrong, Solana back to 10,000, Ethereum to 50,000.
We do need that. We have to dream. We need that Tom Lee sort of spending all his cash, man. We need a new figure to arise.
It could be you, dude. It could be you. It could be you.
We went from that. That is not my general demeanor about anything.
Do Kwon to Michael Saylor to Rob Hadick.
Rob, I'm telling you, man, you are under no obligation to be the person you were 2 minutes ago. It could be you. I'm just going to leave it at that.
So, Rob, you could be this guy and Santi can be the Tony Robbins. That helps you get there, you know.
10. Amazon Discovery Rewrites History
Rob, I charge advisory fees.
All right. Speaking of bullish announcements, I have my content of the week. Do you guys know what you're looking at here?
Dude. All right.
Saddam Hussein.
All right, Saddam Hussein—no, no. Okay, this is content of the week, way bigger than Micky Malka's podcast on David Senra. For basically forever, archaeologists believed that the Amazon could only support these tiny little communities because its soil was too bad for big, complex civilizations. Then we started ripping all the trees out of the Amazon through deforestation, and they started exposing these big circles and squares carved into the ground. We basically just did a bunch of LiDAR, using all this advanced LiDAR laser technology, and found a huge pre-Columbian network of cities, which essentially rewrites our understanding of South America. They flew LiDAR over 4,500 kilometers, digitally removed all the trees, and found 396 unknown earthworks surrounded by 15-foot-deep connected roads through all these different cities. Anyway, it completely rewrites history.
Dude, aliens are real, I'm telling you. You go down this rabbit hole, and no one can explain why they built the pyramids in Egypt—the engineering marvel of the precision of those laser-cut stones—and the Maya. Is this validating this? These look like alien structures, man.
It looks like old civilizations. I'm reading this on Science.org.
Yeah, this could just be old but sophisticated civilizations that we've known.
I think, as a Mexican, Rob, I'm here to tell you that that looks very sophisticated, though. I wouldn't put it past the conspiracy theories to say this is some alien civilization.
All right. Well, the next Blockworks event is going to be a cruise on the Amazon. We have D Asia coming up in Singapore around Token2049, and D London in mid-November. Jeff from Hyperliquid is coming, so we should do a podcast on site with Jeff. We also have an episode with Paul from Morpho coming out next week.
Nothing said on Empire is a recommendation to buy or sell any investments or products. This podcast is for informational purposes only. The views expressed by anyone on the show are solely their opinions, not financial advice or necessarily the views of Blockworks. Our hosts, guests, and the Blockworks team may hold positions in the companies, funds, or projects discussed.