The Six Tokens To Own This Cycle | Yan Liberman
Yan Liberman is long the rally but does not see near-term Bitcoin highs without materially easier liquidity. The squeeze began from “bottomy” positioning plus the prospect of long-end curve control resembling QE; subdued leverage and Saylor’s absence are constructive, but Yan sees Bitcoin reaching the high $90,000s before needing a more accommodative Fed or favorable inflation data. “It kind of has to be in this order”: slow Bitcoin grind, pause, then alts.
The opportunity set has narrowed into a stock-picker’s market of perhaps 10–15 investable tokens. Obvious revenue names are already crowded and need growth to expand their multiples, so Yan wants assets that “look and smell like those” without the positioning. HYPE remains a comfortable long-term hold because of non-crypto trading activity and a prospective USDC bid supporting buybacks, although $200 within four months “would be really surprising.”
His six-token portfolio comprises HYPE, Zcash, Lighter, Venice, Grass and AERO, likely Aerodrome, with sizing undisclosed and positions split between spot and leverage. The preference is for productive revenue, especially demand not wholly dependent on crypto activity; Yan typically adds leverage early for “torque,” then unwinds it into spot after a substantial move. He distinguishes durable earnings from assets whose fundamentals are tightly linked to market reflexivity.
Grass is Yan’s clearest forgotten-business repricing thesis: roughly $70 million of expected revenue and $40 million of indicated operating profit against about $330 million FDV. Its roughly eight-million-node residential-IP network supplies continuously refreshed web data to frontier labs, while a planned Live Context Retrieval subscription could unlock real-time sites blocked to ordinary agents. Investor unlocks are scheduled to be done in October, roughly 30% of supply sits with the foundation, and there is no parallel equity claim: the token is the value-accrual mechanism.
AERO is an event-driven trade with a dated catalyst and explicit invalidation. Its delayed expansion is slated for the second half of September, with Ethereum and Arbitrum announced and Robinhood unconfirmed; Yan expects broader distribution to lift TVL, volume and fees quickly. Planned dynamic emissions and new revenue lines—estimated to add 40% without an emissions offset—could answer the longstanding criticism that token emissions neutralize staking revenue.
Delphi finds more opportunities in liquid crypto than new crypto venture, while its sole recent crypto startup investment targets a dollar-hedged Turkish-lira carry trade yielding about 10.5%. Tory Finance offers organic yield without token incentives and potentially large capacity. The conversation flags the dollar hedge and bank-counterparty structure, while users could add another carry layer by borrowing USDC against the position. Only one of Delphi’s five current venture investments is crypto, reflecting a broader-adoption thesis that has “fallen by the wayside” as AI absorbs talent, attention and capital.
The hardest pushback is opportunity cost: productive tokens remain among the highest-beta assets and must compete with Nvidia, SpaceX and other scalable businesses. Yan wants cash available for a brutal economic unwinding and asks what tangible flows and adoption would justify HYPE at $79 clearing a 2–3x return. The hosts’ corrective is blunt: “Do you want to be right? Do you want to make money?” Both AI private markets and crypto tokens can be simultaneously frothy.
1. The rally has legs, but new highs still need a buyer
Yan is “pretty long,” has not taken profit and thinks the move retains legs. The catalyst hit while consensus already viewed the market as “bottomy,” turning prospective long-end bond-curve control—effectively some form of QE—into the spark for a large short squeeze.
The tape still looks measured: leverage has not rebuilt aggressively, and Yan wants Bitcoin to grind higher slowly before pausing and handing leadership to alts. If everything rises too quickly together, he doubts the move will persist.
Saylor’s absence has been favorable in Yan’s reading. He has been “cash-maxing” and paying preferreds, while STRC is moving toward 100. Yan thinks the related arbitrage is difficult because the derivatives structure permits effectively unlimited selling when downside is capped at 100.
Yan cannot easily see fresh all-time highs without material new QE. Saylor supplied the last obvious buyer, although the counterargument is constructive: older holders sold into stronger hands, potentially improving the holder base even if rebuilding the required demand takes time.
2. A thinner alt market rewards uncrowded productive tokens
The hosts count perhaps 10–15 broadly investable coins, repeatedly owned by the same liquid funds. Yan accepts the crowding problem but argues that a shallower alt universe lets sidelined capital coalesce around fewer assets, creating “a bit of a stock-pickers market.”
These tokens are “not necessarily cheap” and require meaningful growth for multiple expansion. HYPE earns a higher multiple because investors trust its earnings and anticipate a durable USDC bid supporting buybacks; Pump trades cheaply because the market questions whether its cash flows endure.
Yan’s preferred setup is a token that “look[s] and smell[s] like those” revenue names but has not been bid up. Lighter could still appreciate substantially; HYPE is easy to own over a long horizon, but a rapid move to $200 would surprise him.
The discussion turns to HYPE near $79 and the flows needed for a 2–3x return. Yan says very large outcomes are possible—even a much higher long-term valuation—but asks where the volume and demand would come from: U.S. adoption, additional trading activity, fee increases and an “everything exchange” role.
3. The six-token book pairs crypto beta with outside demand
Yan’s six-token book is HYPE, Zcash, Lighter, Venice, Grass and AERO, likely Aerodrome. He declines to disclose weights because prices move faster than he rebalances. As of the September 11 recording, some positions were spot and some remained levered.
The discussion favors “productive rather than reflexive assets,” while recognizing that trading-fee revenue will still decline in a broad selloff. AI demand or non-crypto markets can nevertheless make a token’s business less dependent on speculative turnover.
Venice fits through private inference: periodic privacy-related events can remind users why the product matters, while revenue does not rely primarily on crypto activity. A financing round exposed the awkward pecking order between equity and token holders, temporarily stalling the token.
Yan nevertheless expects Venice’s token to become the long-term value-capture mechanism. More proprietary hardware and greater scale should improve margins, and he is less optimistic about investors exiting through equity; importantly, there is no separate equity value-accrual mechanism competing with the token.
4. Grass is the forgotten-business repricing bet
Grass solves a concrete data bottleneck: many websites block data-center IP addresses, so its browser plugin uses spare residential bandwidth for geographically distributed scraping. Yan stresses that it does not inspect browsing or keystrokes; it merely uses excess bandwidth for ad hoc requests.
That network has grown to roughly eight million nodes, with frontier AI labs as customers. Yan says labs continually need new data because web content changes and they do not retain everything internally, producing an underlying business expected to reach about $70 million of revenue and, based on company indications, $40 million of operating profit.
A planned product called LCR—Live Context Retrieval—could arrive this year. It would sell subscription access to live information from sites that current agents cannot reach, including pages shielded by Cloudflare, adding a higher-level product atop the existing collection network.
Against roughly $330 million FDV, Yan sees a business that might command a $1–2 billion private-market valuation if it were ordinary equity. Investor unlocks are scheduled to be done in October and about 30% of supply remains with the foundation; there is no parallel equity claim, making the token the sole value-accrual vehicle.
5. AERO offers a catalyst that can confirm or invalidate quickly
Yan frames AERO as a catch-up trade to UNI, whose performance followed moving to Unichain and activating value accrual. AERO’s own migration was expected in mid-July, then delayed, punishing the price.
The new timetable was the second half of September. Ethereum and Arbitrum had been announced, while Robinhood remained unconfirmed; Yan thinks rapid multichain expansion could materially lift TVL, trading volume and fees.
The historical objection was that emissions paid liquidity providers while revenue went to stakers, making apparent value accrual illusory. The proposed fix makes emissions dynamic and one-to-one with trading volume and revenue, producing “positive reflexivity” because higher prices reduce the required emissions.
Management also estimates new business lines could add roughly 40% to revenue without corresponding emissions, with proceeds directed to token holders or buybacks. Low open interest relative to FDV gives Yan a lightly positioned catalyst with a clear event, confirmation and invalidation.
6. Liquid crypto now clears a higher bar than venture
Yan sees more opportunities in liquid tokens “purely [as] a numbers game,” with much shorter realization periods and therefore attractive IRRs. Forgotten legacy assets can keep building while attention disappears, allowing investors to re-enter at unusually compressed valuations.
Delphi has made roughly one crypto venture investment in the past year; among its five current investments, only one is crypto. AI has diverted talent, attention and capital, while experience has narrowed which crypto businesses appear defensible and capable of capturing value. The majority of Yan’s venture book is now outside crypto.
That exception, Tory Finance, tokenizes a dollar-hedged Turkish-lira carry trade yielding about 10.5% organically—“no juice,” incentives or subsidies. Private-wealth relationships secure high-tier counterparties, while the trade offers large capacity and yield originating outside crypto. The company does not yet have a token.
The conversation highlights the dollar hedge and its banking counterparty; the host describes the bank as backed by Turkey’s central bank and says it has historically held up during unwinds. Users could add another carry layer by borrowing USDC against the position, while higher base yield lets the protocol earn more per dollar of TVL.
7. Gaming and prediction markets exposed venture’s moat problem
Yan has cooled sharply on crypto gaming because short game lifecycles make long-duration NPV difficult. Gunzilla was a high-quality, AAA-caliber bet comparable in ambition to Call of Duty, PUBG or Fortnite, but he attributes its failure largely to management and spending.
After multiple smaller-game attempts, Gunzilla was intended as Delphi’s final large barbell swing. Yan still expects isolated games to work, but doubts gaming will return as a broadly investable crypto category; Delphi’s remaining interest is concentrated in “financialized apps.”
One rejected concept used paired impact markets—for example, Trump-BTC and Kamala-BTC contracts whose prices represented Bitcoin under each election outcome. The team liked the mechanism but passed because adoption was uncertain and distribution, rather than protocol defensibility, could determine the winner.
Yan views Kalshi primarily as “regulatory arbitrage on sports betting”: peer-to-peer matching removes the bookmaker’s traditional vig, but avoiding state gaming taxes by classifying income as trading revenue could eventually backfire. The hosts consequently struggle to underwrite the floated $40–80 billion valuations.
8. Missed deals changed the process more than the appetite
Delphi saw Pump’s earliest round but passed because the available allocation was only $25,000–$50,000 and did not feel like a venture deal. One host passed around a $20 million valuation; Yan had previously missed Axie and says that experience helped him avoid repeating the mistake with Pump.
Yan says AI has become a “talent vacuum.” Crypto no longer receives the concentration of ambitious founders it attracted in 2019–2021, leaving fewer new companies and making established teams—and their already-liquid tokens—relatively more compelling.
To learn AI, robotics and, more recently, defense technology, Delphi has reviewed roughly 500 emerging managers and backed seven or eight. It seeks first- and second-time funds across China, San Francisco, Europe, South America and India, using them as both investments and distributed sources of market intelligence.
The decisive manager trait is a repeatable “right to win”: taste, hunger, differentiated sourcing and principles that identify trends early. Founder diligence now carries more weight too—Delphi walks through a founder’s life history and looks for “spikiness,” exceptional prior achievement and the relentless energy Yan sees in Ethena’s Guy.
9. Cross-asset opportunity cost has replaced crypto exceptionalism
The hosts challenge whether any token’s possible upside compensates for liquidity risk and violent downside when Ethereum already sits near $300 billion and many venture-era “10x” assumptions no longer work. Yan increasingly wants cash available for a brutal economic unwinding and asks what concrete flows and demand would justify a large HYPE outcome.
The hosts argue that recession precursors could initially help crypto: rising yields might force the Fed or Treasury to respond, reviving the debasement trade and demand for harder assets. Yan agrees that the buildup could be bullish, while keeping the hedge that once “it all goes to shit,” everything falls.
The opportunity-cost example is Nvidia after the ChatGPT moment: investors tried to find “the next Nvidia” because simply buying the leader felt too easy, despite Nvidia gaining more than 2x at virtually unlimited size. Yan also cites copper’s roughly 2–2.5x move and concedes that most of Delphi’s venture book is now outside crypto.
Both sides retain valuation humility. Miro reportedly fell from a roughly $17 billion last round to a $1.3 billion acquisition, while Kraken secondaries once available around $3–7 billion preceded a roughly $21–22 billion investment. Momentum can reverse either way, and AI froth can coexist with roughly 99% of tokens trading below their pre-TGE valuations.
Full transcript
Nothing said on Empire is a recommendation to buy or sell any investments or products. All right, everyone. Welcome back to Empire. We've got Santi, myself, and our good friend Yan Liberman, who's one of the co-founders of Delphi Digital. He kind of grew up in the industry alongside Delphi. I remember the day you guys launched, Yan. Yan is also the managing partner of the venture fund at Delphi. Yan, welcome to Stoutman.
Thank you for having me. Excited to be here.
How's life?
Good. Good. Can't complain. Markets are better. And yeah, PUMP kind of solves everything, doesn't it?
1. Does Crypto’s Rally Have Legs?
Yeah, PUMP does solve everything. Okay, let's talk markets. I'd love to start broad and high-level and get your view. I think you had some of the biggest wins that I have seen in the last cycles, and I would like to hear how you're thinking about this cycle. I'd actually love to spend time in this episode talking about your actual portfolio and the tokens that you hold.
But before we do that, let's talk about how you're viewing this market. We've had a couple of weeks of number go up, so I'm curious how you're thinking about the market today and how you're allocated.
Yeah. No, I'm pretty long. I haven't really been taking profit on any of the positions yet, just holding steady. It feels like there's still legs.
I think the market got a bit caught off guard by the news that there's going to be some form of QE through long-end-of-the-bond-curve control. That was a catalyst at a time when everyone agreed that the markets were kind of bottomy, where nothing was really happening and Bitcoin was just chilling. If you had one light to ignite the fuse, you had this big short squeeze.
So I think after that, everyone was kind of wondering, does this continue to have legs? What happens from here? I think it's great that Saylor hasn't really been buying, right? That's usually been a bearish signal. He's been kind of cash-maxing and paying off some preferreds, and you're starting to see STRC go to 100.
I think that arb is going to be really tough with the derivatives market that exists on that. There's infinite selling that can basically happen because your downside is capped at 100. I don't know if he'll come back into the market anytime soon, but I think that's honestly been pretty favorable for it.
Overall, you haven't really seen leverage build up too much, so it seems pretty measured. These kinds of slow grinds higher, and then we pause and alts run—I think if you really want to see this continue, it kind of has to be in this order, where it doesn't all happen very quickly.
I personally have a hard time seeing new all-time highs anytime soon without some material new form of QE. I think Saylor was the one that took us to those highs, and that took quite a bit of capital. The counter is that you did have a lot of old holders sell out, right? And so it kind of went into much stronger hands in the form of Saylor.
You can argue that that's constructive toward future all-time highs, but I think it's just going to take quite a bit of time because you need a lot of money to get there. There are going to be skeptics along the way, and so you don't really have this obvious buyer, which we've historically had in these markets.
In the meantime, I do think the alt market isn't as deep as it used to be, and that's constructive in an environment where there's just not that much cash sloshing around in this market. It's easier for people to coalesce around a handful of assets, and that makes it so you don't necessarily need a whole new collection of buyers to come in and a lot of new capital.
I think there's a decent amount of sidelined capital that can move out of assets that are less attractive, and it becomes a bit of a stock-picker's market, so to say. I think it's been pretty evident with just the amount of outperformance you've seen in a handful of names while a bunch of others haven't really done as well.
Yeah. Is that a good or a bad thing, though? Because I feel like, let's say there are 15 investable coins right now. Any liquid fund that comes on Empire names the same 5 coins over and over again. It's Hyperliquid; it's some of the same things. It's a lot of the things that generate revenue. There aren't many of them right now, so let's call it 10 to 15 total.
2. How Delphi Evaluates Founders
The good thing about that is, yeah, maybe when fresh capital comes in, there's only 15 things to buy. The other side of that, though, is all the crypto funds are already allocated to these 15 tokens. How do you think about whether that's a good or a bad thing?
Yeah. No, it's reasonable. Especially if you look at valuations, they're not necessarily cheap. They're not overly expensive, but you do need a decent amount of growth for the multiples to expand. And so you can start to make individual cases for each of them.
They are trading at different multiples because of the conviction in those earnings, right? Hyperliquid is much higher because you're also going to have the USDC bid that comes in. And so that becomes a pretty stable source of buyback. On the other hand, you have the PUMPs of the world, where everyone's skeptical of the durability of those earnings, and it trades pretty cheaply.
It is a problem—or not a problem, but I think the setup there is that you kind of want to look for tokens that look and smell like those but aren't necessarily as crowded a trade, right? And so I think the setup is there where there's appetite to look for compelling alts that haven't really been bid up, and I think that's where the real opportunity is.
3. Is AI Outshining Crypto?
Lighter could still go up quite a bit. I think Hyperliquid on a longer time horizon is easy to be bullish on. Do I think it goes to 200 in the next 4 months? That would be really surprising. But I think on a longer time horizon, it's something you can sit in comfortably. If you're trying to do shorter-term outperformance, I think you kind of have to look beyond some of the more obvious ones.
Hey, Yan, get your perspective, because I know you guys operate across asset classes. One of your partners, Tom, is—I feel like we get a lot of really good content on AI and robotics. A lot of what we've felt is the risk-reward there in crypto. People talked about coins being down 60% to 70%, but some, like we've talked about, have gained some—they've sort of bottomed and rebounded a bit.
How do you guys think about this more broadly? Is this an interesting category to invest behind versus AI, which is incredibly exciting?
Yeah. No, it's a great question. I think the question is the risk-reward, right? When things heat up in crypto, the markets move harder than anything else, right? And so you can see the upside when things are hot. I think in those periods, it's really worth paying attention to crypto.
Then I think you probably shift your weight a little bit away from it as the market cools off. We do spend quite a bit of time—everyone kind of varies, but I do as well—on AI, just because that's where there's a lot more interesting things happening.
Even our venture fund right now, for the most recent one, we have 5 investments, and only 1 of the 5 is in crypto. And I think we've done 1 crypto venture deal in the last year, I'd say.
The venture environment, I think, makes it a lot more clear what can accrue value and what's defensible on the crypto venture side. And so what that allows you to do is screen out a lot of the stuff pretty quickly.
Unfortunately, the broader adoption of crypto across a bunch of different verticals—that thesis has kind of fallen by the wayside for a handful of reasons. I think AI is definitely a contributing factor in terms of attracting talent, attention, and capital.
But it is the case, and so you narrow your focus on the crypto venture side, and that allows you to do more on the AI venture side. And then also, I think liquid crypto has a lot of interesting opportunities when markets are trading well.
I'm curious what that one crypto investment is, if you're willing to share.
Yeah. Yeah. It's Tory Finance. They're live right now. They don't have a token yet, but it's a tokenized carry trade, dollar-hedged. So you get a Turkish lira carry trade, basically, and it's in dollars, and you're getting about 10.5% organic yield. There's no juice. There's no incentives.
The idea is you have some private wealth and ultra-high-net-worth folks who have the right relationships to run this trade, because you need to find the right counterparties. And the counterparties are the ones that they would use. They're all kind of the highest tier, with the least amount of credit risk.
And so that trade doesn't really exist in crypto at all. The idea is you can bring this on-chain, and the benefit is that there's a massive amount of capacity for the trade. And really, the risk—
Obviously, you have the crypto-specific one, and then the market one is the dollar-hedge component. And so it's your counterparty, and it's basically a bank that's backed by the central bank of Turkey. They've never really had issues, even during the unwind.
So that’s a durable yield that’s kind of exogenous to crypto but can be brought in. And then, if you want to get cute, you can carry trade by borrowing USDC against this tokenized position.
Yan, is this you mentally mapping your guys’ investment in Ethena into trying to find the Ethena model in other areas?
Yeah, I mean, it makes sense, right? Yield is one of the areas that makes sense. I think, depending on how you structure it, I’m less interested in the net-interest-margin-type businesses, at least unless you’re doing it from a venture position and really early on. Ethena now is obviously going to do well, but you’re not really going to get venture returns.
If you can do that, the higher the yield, the higher the take rate, right? Each dollar of TVL can be more accretive to the top line when you have higher yield, and so it doesn’t need to get as big to generate the same degree of returns.
4. Liquid Tokens Or Venture Bets?
It’s funny you mentioned this. One of my buddies is a pretty big sovereign-debt trader, and he was telling me, “You’ve got to do this Turkish lira thing.” I’m like, “I’m too busy for this stuff,” but hearing you say it brought me back to that. I feel like a lot of the FX has been hard. I’m old enough to remember Long-Term Capital Management blowing up, but hopefully you get comfortable with Erdoğan and the Turkish central bank not doing anything funky or going awry.
Yeah. Yan, how do you think about what’s more exciting right now: buying liquid public markets in crypto, like the token markets, or the private venture side of things?
Right now, and in general, I think there are just going to be more opportunities on the liquid side. It’s purely a numbers game, and the return profiles are much shorter as well. If you’re thinking from an IRR perspective, I think the liquid situations are pretty interesting because there are these legacy tokens that have fallen by the wayside and are completely forgotten, but are actually building in the background.
You can get incredible valuations, and I’m happy to show you if you want.
Yeah, we’ll get there in a second. I do want to hear about what the book is. Santi, as you started nibbling—maybe not nibbling, because I don’t think you’ve started allocating yet—but as you started looking at crypto, do you have a thought on whether you’ll ramp back up your angel investments or just start buying tokens?
Good question. I have ramped up and bought, just on FOMO. I always have a YOLO account. I was early on WIF, and that did phenomenally well. It’s 1 of my best-returning investments, funny enough.
The market pulled me back in, and I’ve been dabbling in some names just for shits and giggles, to be honest. You’ve got to feel alive, and what’s worse is not being allocated and missing out. We’re all human; it’s very psychological.
But on the big part of the portfolio, I’m rotating from locked and/or vesting positions. I like HYPE. I don’t have a meaningful position there, but I like HYPE. I held through Ethena. I sort of put it in this DAT thesis, and I was like, “I feel like DATs were like—we all knew that they would unwind, and they trade at a premium when bull markets are all the rage. Then you just know that the discount is going to be really, really brutal.”
I remember when this DAT launched, I was like, “I know they have X amount of Ethena on the balance sheet that they won’t sell.” It’s a massive discount to the underlying value. Ethena, like every other token, went down 60%. The beauty and the curse of venture is that you’re always fully allocated because you’re vesting and locked, so I’m not jumping out of my seat. I feel like I’m fairly sized on the venture side.
Projects like Pump.fun—you mentioned that, Yan. I’m curious how you guys are looking at that or not, but it’s a project that’s cranking out a lot of cash flow and has staying power. I didn’t fully—I'll tell you, in a prior cycle I probably would have bought a bunch of Zcash, but I just can’t wrap my head around it. Trading commodities is extremely hard, and I can’t wrap my head around the pure supply and demand. I want to bet on some things.
The good thing now is that you have 3 or 4 years of a protocol like Pump.fun. You look at the cash flow, go look at the Blockworks data, and it’s like, “Okay, it was a pretty brutal market the last 12 months, and these guys have continued to crank cash.” I’d rather just be super concentrated in 5 names, which happen to be venture positions.
The other name I have, which I’ve just been an investor in and have been sitting on for a long time, is Arbitrum. My thesis for L2 investing was, “Okay, this is a levered bet on Ethereum, and they’re going to capture more of the fees.” I did MegaETH, I did Arbitrum, and I did a couple of others. For a while I was crucified. People were like, “Dude, this is a dumb trade. You shouldn’t do it. Just move on. Don’t do it.”
Now it’s starting to feel like, with Arbitrum, if you think Robinhood fees—Robinhood Chain continues to do well—would you rather have Arbitrum in your book, or would you rather have Ethereum? What do you think?
Yeah. I guess if you had to pick—which is the answer—it’s like, I want to pick both. That says it all.
Yeah. The tricky thing is, if you’re right on that bet, then there’s probably better exposure to have, right? That’s kind of where it ends up falling for me. I get the Arbitrum side, but have sequencer fees grown materially? I don’t know what—
It’s directly correlated to their take rate. I think of it as a take rate. What’s a GMV? It’s like Robinhood: they take their cut, which is 8% of Robinhood. If that happens, then it’s sitting at around $1.5 billion to $2 billion in FDV.
I just feel like I don’t necessarily like doing pair trades, but it’s like short ETH, go long ARB, kind of thing. Or if you have a huge position in Ethereum at $300 billion or so, it doesn’t take much to move that L2. Could you see Arbitrum being a $10 billion token?
Yeah, I can.
5. Ads (Token2049, Avalanche)
Welcome to Token 2049. Token 2049 is back October 7th and 8th, bringing together 25,000 attendees, 300 speakers, and 500 exhibitors for the world's largest crypto event. Token 2049 is happening alongside and in partnership with our own Digital Asset Summit Asia, so you can experience both conferences in Singapore during the same week. Across Token 2049 week, there'll be more than 1,000 side events culminating with After 2049 and Formula 1 weekend, and the speaker lineup is stacked: Shane Coplan of Polymarket, Jeff Yan of Hyperliquid, Arthur Hayes, NASDAQ CEO Adena Friedman, and many more. Join us in Singapore October 7th and 8th for Token 2049 and Digital Asset Summit Asia. Avalanche Summit is coming to New York City September 16th to 17th. Join the business and technology leaders building the next generation of financial products, enterprise systems, and consumer applications on-chain. Learn how blockchain is enabling faster settlement, more efficient markets, and new business models at Avalanche Summit NYC. Register for Avalanche Summit and use promo code BLOCKWORKS15. Ahead of the summit, you can track Avalanche tokenized equities alongside tokenized treasuries and their growing footprint on Avalanche. See the data for yourself with Blockworks Research's Avalanche dashboard. And don't forget to join us in New York City for Avalanche Summit.
6. Inside Yan’s Crypto Portfolio
Yan, maybe this is a good time to get into your book. You told me you’re holding 6 tokens right now. I don’t know how comfortable you are sharing those 6 tokens, but maybe we can start there.
Yeah, let’s start with the portfolio, and the more detail, the better here. So, what’s the biggest one? Santi?
Just send us your wallet address. I’m terrible.
Yeah, exactly. Yan told me he’s going to be transparent. He’s a man of the people.
So, yeah, I probably won’t break down the sizing of them, but they’re all—they move around so fast, and I’m not necessarily rebalancing too much right now. The obvious ones were HYPE, Zcash, and Lighter. What else did I send you? I have some smaller stuff as well. Yeah, Venice.
I’m just a fan of the revenue-story ones, and revenue stories that aren’t highly dependent on crypto activity, because I think they can be a bit more resilient. I think the TAM story also becomes a lot stronger.
Venice was 1 of those, right? The idea of private inference is that you’ll have these periodic events that remind people why private inference is valuable. The revenue story there made sense. I think there was some hiccup when the raise happened, and it shined a light on the issues of having equity and tokens simultaneously, and on who’s higher up on the pecking order of value accrual.
And so I think that gave the token pause, but it's clear that they are planning on continuing to direct value to that token in additional ways. I think the business itself—the margins—will continue to improve as they bring on more of their own hardware and as the revenue story gets bigger. You have economies of scale on some of these business lines, which will allow them to direct more value to the token, because I do think it's one of those where exiting the equity might be a little trickier. I'm less optimistic on that.
And so I think the token actually ends up being the long-term value-capturing mechanism, and all the investors are kind of exposed to it. You have this easy way to just position to that.
I like Grass a lot. Grass is actually one of the tokens I was alluding to, in terms of one where they've continued to build but have fallen by the wayside. They're doing the hard part of actually delivering a business that generates a bunch of revenue, but they're not really doing much on the market-awareness side. People kind of assume the worst, and they don't really communicate much. So I think with a correction there, you can have a pretty strong repricing.
We're venture investors in them, and I'm also an investor personally. That business is expected to do $70 million in topline revenue, and based on what they've suggested, it's $40 million in operating profit.
The business itself is basically data scraping right now. Data aggregation and scraping off the web is difficult because a lot of sites block data-center IPs, so you need to have a residential IP network. That's what they started with: You download this kind of plugin, and all it does is use your bandwidth to scrape the internet. It doesn't actually see what you're seeing or any of your keystrokes; it's just using excess bandwidth to do ad hoc scrapes. You want to have a geographically distributed base because you can touch all these areas.
They've built up this massive pool of data, and their customers are all the frontier labs. Frontier labs need to continuously buy new data for each model because they don't store as much of it themselves, and also because the data is constantly updating. So you have this really profitable underlying business, and then on top of that, there are additional products that you can build.
One that I think they'll probably be releasing this year is called LCR. It stands for Live Context Retrieval. The ability to scrape a bunch of live data right now is really difficult. You can't have agents accessing it; even if you're using current agents, they're blocked because of Cloudflare. Basically, any Cloudflare-blocked website can't have agents accessing it, and so you can access it with this.
There will be a subscription product that launches, and the token is trading at a $330 million FDV. Investor unlocks are done in October, and roughly 30% of the supply is sitting in the foundation.
So you have this really attractive setup, and it's one of those that actually did the equity-token distribution correctly, in the sense that there is no equity. There's just a foundation, and there is no other value-accruing mechanism. So you effectively have this business; if it was a pure-equity, non-crypto business, it would probably be trading in private markets, raising at a $1 billion to $2 billion valuation, I think, just based on the growth rate.
But it's a token, and the token was necessary. To be fair, it helped them build out this residential IP network of around 8 million nodes, so it was a means to an end. They raised at a $500 million FDV in the last round, which was a couple of years ago, and realistically, they wouldn't have been able to raise at that valuation if it wasn't peak crypto mania. The token is being punished now, but it was an absolute necessity to get them to where they are. So I think there is an eventual repricing that comes once awareness comes to the market about the token.
7. What Returns Justify Crypto Risk?
Yeah. Maybe just zooming out: what's your expected rate of return in crypto now, and how has it changed? You've been around for quite a bit of time. Has that evolved over time?
Yeah, I think just taking less risk has forced it.
10, but you're taking a shitload of risk. You're still—if you think about broad asset classes, no matter what, whether it's cash, Bitcoin, or a top-100 coin by market cap, it's still as high-beta as you could possibly think in a world where we're going to be—
I was just listening to the All-In podcast today, and they're discussing whether it's 1996–97 or 1999, which tells you we're closer to a recession than we are to not being in a recession. Whether that happens in 2 or 3 years, if your venture book is going to be a 5- to 10-year ride, maybe you don't care as much if you enter at the right balance.
But for liquid stuff, do you want to be sitting in tokens where liquidity might dry up? What kind of return do you need to get in your underwriting to justify being in the highest-beta of asset classes, full stop?
Well, maybe starting from there, we're all hooked. [laughter] But I feel like back in 2017 and 2020, you're like, “This is a 10x. I'm shooting for at least a 10x.”
At least personally, I'm like, “I'm not doing a venture investment in crypto. I don't think I can make at least a 10x.” Ethereum at $300 billion—I can't see a 10x cleanly, and it carries way more. The asymmetry is very skewed on the downside.
Maybe Zcash goes to $10,000; I don't know, man. But on a relative basis, there's a very credible—there's way more probability of a lot of these things really repricing down versus up.
8. Stop Trying To Be Clever?
Yeah, to answer your question, that's just always been in my head. Maybe it's just too much scar tissue, but I definitely want to have cash when there's a very brutal unwinding of the economy and whatnot. I found it increasingly hard because the market's more competitive in crypto. You have much more dispersion, which is a healthy thing, right?
To your point, if you're allocated to 3 or 4 things over the last 5 years, you not only underperformed Bitcoin, you just lost a shitload of money. And that's really bad. So I don't know. What do you need to believe, maybe concretely, like a price target? If you're in HYPE and HYPE's at $80, what do you need to believe for that to be a $40 billion asset?
Yeah. Is that rhetorical, or—
No, no, no. It's not rhetorical. I'm just thinking, in the range of scenarios, it's possible that we'll see HYPE maybe at $500 billion. I'm not saying I don't want to be cynical. There is a very credible chance of this. I'm just trying to understand: Where's the volume coming from? Where's the demand coming from? What do you need to see for HYPE? It comes to the U.S.; it gets adopted in XYZ.
I feel like now you definitely need to be thinking about much more tangibility. Where are the flows going to come from? How does HYPE get there? If you're allocating to HYPE at $79, you need to probably clear a 2x to 3x. What do you need to believe for that to happen, other than “my buddy and my dumb cousin are going to buy it before I sell it”?
Right? Yeah. I mean, to cover both points, I think the comfort level comes more from sitting in productive rather than reflexive assets. The productive side is the earnings durability. How susceptible is it to the reflexivity of the reflexive assets, right?
If Zcash goes down because the whole market's coming down, that means trading fees are probably going to be coming down. So, yeah, some of these are dependent on that, right?
HYPE, I think, has seen its multiple expand a bit because of the dominance in non-crypto trading assets. On the other hand, they charge much lower fees, but I think over time those fees can step up a little bit. Each minor increase produces quite a bit in revenue for them right now.
That adds a bit of comfort to HYPE, and it also expands the TAM. It becomes the everything exchange, and so it becomes a more comfortable hold.
I think similar things can be said about Grass or Venice and some of these others. They're crypto assets with AI demand, or AI revenue, and that's the intersection of the two: You have a discounted token for a handful of reasons, but the business itself is doing really well and is probably less susceptible to recession.
Although the question is: What causes the recession? Is it, “All right, AI revenues didn't come in, and now we're bloated in AI-funded debt”? That's what would happen then.
And I'd argue that's almost bullish for crypto because before that even happens, yields will start coming up, and the Fed and Treasury will have to do something about that. We've seen how crypto responds to that.
9. Where Can Crypto Venture Still Win?
Yeah, there are elements of the precursors to the recession that I think will be bullish for crypto. Ultimately, when it all goes to [__], everything comes down. But I think in the buildup, the narrative will be that basically what's been happening now—where the debasement trade is in vogue and capital flies into harder assets.
When you think about the venture book, talk to us a little bit about the quality of the founders and the evolution of that over the years. Just for context, folks listening, you guys were early on some of the big winners. Like Axie was a big position of yours. Are you still excited about gaming? Are there categories that you feel more strongly about, that you think are non-consensus and you're willing to take a 5- or 10-year bet on a strong team?
It's a good question. On the gaming front, I think gaming in general is a difficult vertical to invest in, just based on the life cycle of games and how they don't exist as businesses for too long a period of time. When you NPV some of these, it becomes a harder long-term value story.
I thought the idea with crypto was that you ideally enable functionality and trading activity that wasn't really previously possible. We've tried a few more. Gunzilla was a big swing for us that didn't work out well.
Why is that, by the way? You're talking about being a bagholder too; I'm curious.
Nice.
Because we talked about it on the pod. It felt like it was a great game: the skins, the way they were iterating on it. It felt like fast-casual, but a much more engaging game. What happened there?
I don't want to necessarily speak ill of the founder or anything like that. I think it was an issue of management and spend, to some degree. I attribute it to that, because I agree the game itself was really high quality, and that's what got us really excited. It was AAA-caliber, in line with basically your Call of Duty, PUBG, and Fortnite.
So, yeah, that was an unfortunate one. I think that, for us, was really going to be the final big swing where we tried to get an AAA title. We've taken a decent amount of swings at smaller attempts, micro-titles that could benefit from a token, and the idea was to barbell the approach a bit and see if we could have an AAA title as well.
It's going to be hard. I don't know that gaming will come back meaningfully. I think you'll have small iterations of small versions of games do well, but I think broadly, as a category, it's going to be a bit tough. For us, honestly, the amount of verticals that we think are really investable is pretty limited. It's basically just financialized apps.
We looked at variations of prediction markets that we thought were really interesting, like Impact Markets. I don't know if you saw those, but the idea is basically you have 2 markets that exist simultaneously. If you're going into an election, it's like a Trump BTC and a Kamala BTC, and the 2 markets trade with the assumption that the price is reflecting what would happen if that were the reality. It's a compelling way to express bets.
We looked at those for a while. We were pretty nerdy about the idea, but ultimately didn't end up investing. I think it's going to be tricky to get broader adoption, and there's also just this issue of moats. In the end, whoever has distribution can just take over, and that's been a big issue for us in general when thinking about where these—
Yeah, what do you think is going through the mind of people who are putting money into prediction markets like Kalshi? What is Kalshi raising at—$80 billion? $40 billion? No, $80 billion, I saw. I don't remember.
I think your point around distribution is important. When you think about the moat of these open, permissionless protocols—well, Kalshi isn't open or permissionless, but you know what I mean—I think we discussed on the pod that Robinhood is a huge share of the volume. At some point, Robinhood is going to turn that on. They have a good product team; they can launch their own prediction market. At that point, what is the real moat of something like Kalshi? And who's investing at, like, $40 billion? What kind of assumptions are you making when you're buying it at $40 billion?
So, my take is that it's basically regulatory arbitrage on sports betting. In particular, because it's peer-to-peer, you have better pricing than you would with somebody who's betting purely against the house, because there's not that plus-110 or minus-110 type vig. So I think you have tighter spreads, and basically you have kind of a regulatory arbitrage, and they do have a decent amount of capture.
I have other friends that work in sports betting, traditional sports-betting companies, and they're just basically saying Kalshi—even though they're breaking state laws—is where they think it can backfire for them. The issue is any traditional sportsbook pays a decent amount of tax on every dollar of revenue, and that's why the states allow them to do it. Kalshi doesn't, because it's labeled trading revenue, not sports-betting volume. There's a lot of money that they don't have to pay to these states. That's the push and pull where I wonder if, at some point, the rules are going to come home to roost. I don't know how you can continue on that front, but I think that's the bet they're making.
I had a buddy who was investing in the Polymarket round. I kind of forgot what it was, and it's been a 2x from here. I was like, “I don't know, man. That seems pretty rich,” and I'm the one who got in his face.
Yeah, it's tricky.
Maybe on that point, we all have one of those rounds that we passed on that comes back and haunts us, especially on the venture side. There are ones that really torture you. What has been one in recent memory where you're like, “Gosh, I wish I had done that”? Maybe it's FOMO, or maybe I don't know.
Yeah, we had a look at Pump.fun. It was either a $25,000 or a $50,000 ticket, and obviously right now that's still worth a good amount. But it was basically that first round, and we just thought, “We can't really do this as a venture deal.”
I had the same thing, Yan. I passed at the $20 million level.
Yeah, that one.
Yeah.
Brutal. Brutal. You know, that happened to us with Axie, actually, and that would have been an astronomical return. It was the same construct. So I learned from that mistake, and when I saw Pump.fun, I was like, “Yeah, no, I'm not making this mistake again.”
What do you think about the quality of founders right now in crypto?
Honestly, we've been seeing way fewer venture deals in general. I guess if you're talking about the quantity or quality of founders, it's just going to come down as the denominator comes down. There haven't really been that many new startups, and this is one of those situations where I think AI is just the talent vacuum. In 2019, 2020, and 2021, a lot of the smartest people were trying to build in crypto, and then after everything that's taken place, that talent has pivoted to AI based on what's most interesting.
10. Aerodrome’s Catch-Up Trade
I don't see there being that many new talented founders, but I think you do have some legacy ones that you can invest in. That's why I think the liquid side ends up being potentially a bit more compelling, at least when markets are doing well.
Yan, what are the 6 tokens in the portfolio? I think we got 5. We've got Venice, Grass, HYPE, LIT, Zcash, and I think there's 1 more.
Yeah, I think AERO is probably one of the more interesting tokens right now, based on what's coming. UNI has traded really well. They did the right move of moving to Unichain, and they turned on value accrual, so the 2 combined have led to really strong performance.
AERO was supposed to basically do this migration where they would move to mainnet and also be able to move to other chains. That was supposed to happen in mid-July, and then it got delayed, so the price obviously suffered as a result. But right now, the release is slated for basically the 2nd half of September, and they've announced that they're moving to ETH and also Arbitrum. It's unconfirmed whether they move to Robinhood as well, so I think there's just this massive opportunity for them to really step up in terms of the amount of TVL, volume, and fees that they'll accrue in a very short period of time.
I think people kind of use it as a catch-up trade. The other issue around AERO was always this relationship between revenue and emissions. Basically, the idea was emissions would go to LP pools and revenue would go to stakers, right? The criticism was that you're offsetting revenue with emissions, and ultimately the value accrual isn't necessarily real.
What was happening now is twofold. One, they’re going to be making emissions dynamic, so it will be one-to-one with the trading volume and trading revenue that happens. And then, in addition, they will be—so you have positive reflexivity there: as price goes up, emissions come down. They’re also adding multiple new revenue lines that they estimate will contribute about 40% in additional revenue, and those won’t have an emissions offset.
Those revenues will go to token holders or buybacks. Right now, you do get pretty decent yield just by staking the token. But I think you basically have this big opportunity. If you look at OI relative to FDV, there’s not a lot of positioning in it. I’ve chatted with desks and teams, and there aren’t a lot of folks in it.
I think it’s one of those situations where you have this obvious event that’ll happen, you get clarity on what the token should be trading like, and you have invalidation. You have an event, so the thesis can play out, and you can either have confirmation or invalidation. I think it naturally fits this value-accruing token. There isn’t much positioning in it, and there’s a reason for the fundamentals to materially improve at a very rapid pace. So I think that one is a really interesting one.
Yeah, we also—I like AERO as well. When you are expressing these 6 trades, Yan, are you putting on leverage, or are these just spot positions?
I usually do leverage early on, and then I’ll unwind that into spot. I get the torque early, when it’s—
—you know, moved the least.
And then I just don’t really want to hold leverage after it’s moved considerably.
Like today, we’re recording this on September 11. Are you sitting in these 6 things in spot, or are you leveraged up right now?
Some spot, some leverage.
Got it. Cool. We’re early. I believe. Yeah.
I mean, for me, I think Bitcoin probably goes to the high 90s, maybe. I don’t really want to be too long after that unless we get some new information around a more accommodative Fed or something like that, or a really good inflation print that allows them to start cutting.
11. Why Delphi Backs Emerging Managers
Yeah. Yan, why are you guys— I was talking to Anil the other day, and it sounds like you guys are now starting to allocate to a bunch of emerging managers. I’ve seen Jose’s podcast that he’s been doing with the emerging managers. Why is this a thesis that you guys are getting excited about?
We started doing a mix of crypto and AI probably in 2023. That’s when we started down that path, right after the ChatGPT moment. We started to explore, and at first it became the intersection of the two because that’s just naturally what we knew best. Shortly afterward, we also started looking outside of that and into robotics as well.
More recently, over the past 12 to 18 months, we’ve been investing in defense tech. As soon as the war kicked off, with all the drone warfare both in Ukraine and in the Middle East, we’ve been trying to invest on both sides of that. Naturally, it’s a new area we’re investing in, so you have this adverse-selection-bias fear: Why am I seeing this deal? Who, with much better reach, is passing up on it?
In crypto, I had an idea of why I was seeing a deal and why I might potentially be first to it. In this market, where we’re new, it’s a much harder thing to get comfortable with. In the interest of building up our network, building up our understanding of the space, speaking to as many smart people as possible, and understanding how they’re thinking about the market, we also wanted to do this fund-of-funds.
Basically, it’s all first- and some second-time fund managers. We’ve chatted with probably 500 or so now, and we’ve invested in 7 or 8.
You’re talking with 500 emerging managers, and you’ve invested in 7 or 8.
Yep.
What are the key standout characteristics of these managers? Is it people who have hit 1 or 2 huge winners? Is it hustlers? Is it expertise in a really niche topic?
It can be any of those. The key is a repeatable edge, I guess—why they’re right to win, a repeatable edge, and something that they can continue to do well with. That’s what we want to understand.
Taste is a big one. Prior to this, we hadn’t really invested in any funds, so we had an idea of what a good founder looked like, but not as much of what a good investor looked like. Those are 2 very different things, and I think that’s been a fun exercise.
There needs to be a hunger, something novel about their approach. They need to have taste. They need to have strong theses or principles that allow them to be early to trends. A lot of character references are involved as well, but really, it’s about chatting with them enough, understanding what makes them unique, and understanding how they think about the world.
We try to do it geographically distributed as well. We have some in China, some in San Francisco, some in Europe, 1 in South America, and 1 in India. We’re trying to be as geographically distributed as possible and learn from them. I think it’s a really great source of information about where they’re seeing growth, what they’re finding interesting, and what they’re not. It’s just a great learning process for us.
Yeah. To double-click on that, what are the things in your investment process or discussions in investment committee that have evolved or changed the most? If you were to sit on the investment committee of Delphi 8 years ago, 5 years ago, and today, and take snapshots, would you say, “Holy shit, we’ve really totally changed the conversation”? Are you way more focused on X or Y and less focused on these other things? Has there been a material change in your underwriting process and discussions at committee?
Yeah. If I had to pick 1, it would definitely be the degree of founder diligence that we’re doing. In crypto, especially early on, it was very obvious what needed to exist, so you’d invest more on the merit of the idea and its defensibility versus the founder. Obviously, the founder component came into play, but I think the weight that we assigned to it was lower than what we’d assign to it now in these areas.
That happens for 2 reasons. One is the nature of crypto, where we overweighted the idea versus the founder. The second is that, as you move into a lot of these other verticals that are very specialized, there’s just no way you’re really going to be able to diligence the idea, its defensibility, and its moat.
Over time, you can get better at understanding moats and where the market might be, but early on, especially with how wide a surface area there is to invest in on the tech front, the only repeatable thing that you can scale is the founder side. It’s also been a conscious strategy on that side.
Yeah, that’s interesting. I’ll say some of my best investments were definitely very uncomfortable, but I didn’t fully appreciate the quality of the founder. That founder evolved and grew a ton from the moment I put money in—from the first check—to where they are now.
Honestly, I don’t know how to measure that. A lot of times, people just look back and say, “Of course, he was tenacious. He was just a beast, and he would never give up.” But it’s hard, because there are people walking through your office and you’re definitely miscalculating.
There’s 1 thing I could tell you: there are 99 things that you can do, primarily pet peeves, that will tick me off, and I won’t invest. Going to crypto conferences is the best way to really assess a founder in the wild. A lot of them do dumb shit, and you’re like, “Okay, yeah, you’re definitely not worthy of—you’re not going to build a big company.”
But sometimes my gut is really off. I think in venture, your gut needs to be pretty good to measure people, but sometimes it’s really off in the sense of, “Holy shit, if I had really gone with my gut, I would not have made that investment,” and that ended up being one of the best investments. It felt really uncomfortable.
So how do you guys think about that? Maybe this is the value of having 5 people, or however many of you have on the committee. Is there a lot of disagreement about the quality of the founder, or do you have a really established framework or process to drill down into it?
It’s a great question. I don’t think they’re super-strong disagreements. I think it’s probably more a matter of the magnitude of quality rather than just broad disagreement.
The other thing is that we end up relying a bit on each other. We can’t all spend 2 hours plus chatting with the founder. Sometimes it’ll be 2 of us, sometimes 1, sometimes 3, based on the amount of time that we have to invest. Obviously, we want to be respectful of their time.
To some degree, we’re sharing all the notes and everything, and we have a framework that we use to create various criteria.
Not massively. I think we’re beginning to get aligned on what we think a great founder is. It’s funny—not to say that this is a massive one—but the extent to which you see childhood trauma being one of the actually valuable factors: the chip on their shoulder and this undeniable desire to win at all costs. I think probably one of the best crypto founders we’ve invested in is Guy from Ethena, whom you obviously know very well. I think that guy will stop at nothing to win, and anytime I talk to him, I’m incredibly bullish on what he’s building. He’s got this innate ability to really get you and the team excited.
It’s like a “fuck everyone else” energy—not, “I will do this and I don’t care what you think. I’m just going to do it.” He’s tenacious.
No, no, those are 2 separate, tangential points. I wasn’t alluding to the trauma being something he’s part of. But I think it’s a spikiness, and it’s different from the good investor gene. On the good investor side, it’s taste and the ability to network: Who are you speaking to, and why are you going to be seeing these deals early?
On the founder side, it’s just, were you really successful at something early on? It doesn’t have to be school; it can be gaming or sports. Were you just the best? It’s very hard to be the best, or one of the best in the world, at anything, right? You have to respect that, and that’s definitely something we chat through. We basically start with, “What’s your earliest memory?” and walk you through your entire life.
Does it bother you that you look back over the last 8 years—or even the last 5 years—and think, “I could have bought secondary in SpaceX and vastly outperformed, or bought Nvidia and very likely vastly outperformed, on an absolute basis and certainly on a risk-adjusted basis, our venture”? There’s the idea of the venture power-law distribution: the top 1% of VCs make the vast majority of the returns.
I’ve certainly thought deeply about that. What’s the return profile of my book? I benchmark myself against things like, “What’s my DPI?” I’ve built a lot of that with Claude and have just been trying to measure it. Sometimes I wonder, “Maybe I should just buy Ramp.”
Does that ever compute with you guys?
The tricky thing is that I don’t know that I would have been in SpaceX, for one, and I wouldn’t have had the money to do it, right?
Nor could you get access. You couldn’t get access to SpaceX.
But you could have gotten access if you worked through it. You could have gotten access 5 years ago.
I think AI has repriced public-market comps massively. Now you have 8 or 10 trillion-dollar companies. However many years ago, there were so few of those, and when you’re thinking about the upside in SpaceX, it probably didn’t seem as obvious that it would be this high.
With Nvidia, you probably have to do a fork between November 2022—whenever ChatGPT launched—or October, because the value of GPUs completely shifted, right? If you anchor to that, yeah, I’m sure you’d still massively underperform. But kudos to the people who held those things, too.
But you’re investing in tokens, right? They’re way more volatile, and the volatility will just spook you—not you guys, just generally—and make you more inclined to sell.
Less so on the volatility, but more so on the unrealized gain portion, where you’re just sitting on it.
And I think about that constantly.
Yeah, totally. I do look back at singular moments, and I think people try to be clever in how they make money. It feels like you want to be non-consensus, and that’s the best validation. If you bought Hyperliquid at a dollar, you look like a genius and you’re insufferable. Or Zcash—it feels great to be non-consensus early, and then eventually be right.
But you could have sat there and listened to Stan Druckenmiller talk at the Sohn Conference and say, “Go long Nvidia.” That was the ChatGPT moment. I remember that a lot of people I talked to at that point were saying, “No, it can’t be this easy. You need to find the next Nvidia.” You always need to find the next Zcash, the next Hyperliquid.
He said the same thing about copper literally a year and a half ago, and that was a clean 2x—2.5x, or whatever it was. Nvidia has now been over a 2x at any size you can possibly think of. I think about that constantly.
Anytime I’m talking to a founder, anytime I’m going to pull a dollar away from that, I’m like, “I’ve sort of stopped trying to be clever and just trying to make money.” Do you want to be right, or do you want to make money?
I think a lot of the stuff in crypto is hanging on to narratives or this idea that it needs to work. But I’m just not sure we’re going to see a trillion-dollar business. It’s extremely hard for me to believe that we’re going to see a trillion-dollar business that isn’t a commodity in crypto. I don’t have any issues seeing that in AI or in other industries.
Honestly, that’s predicated on needing to see a whole lot more user onboarding and stickiness to products, or believing that the ISOs of the world are just going to hand us a huge favor. It’s hard to underwrite that, to be honest. I don’t want to depend on these singular entities coming on. I don’t want to need to believe that the token can do a 5x if there were 10 other Robinhoods. It’s just hard to underwrite that.
Yeah, no, I fully agree. The majority of my venture book is not in crypto.
It’s shocking how quickly things in this world can change. Think about the things that were happening in 2021 and how completely absurd those things seem now. Even when we did a podcast in 2023, we talked about what happened in 2021 and said, “Who in their right mind would even do those things?” But they were so normalized.
There’s a chance that in 2 years SpaceX is the most valuable company in the world and Nvidia is another 2x from here. There’s also a chance that we say, “Can you believe that people were piling into SpaceX at that price?” I think it’s so easy to say that today.
Totally. No, I definitely hear that.
Momentum can just change on a dime.
100%. You’ve had, over the last 3 months, the highest momentum-unwind spikes since 2008.
Like, what is Ramp raising at right now? Ramp is raising at $60 billion. There’s a chance that Ramp is a $100 billion company in 2 years. There’s also a chance that we say, “Can you believe they fooled the world into valuing an accounting B2B SaaS company at $60 billion?”
By the way, I love Ramp. We use Ramp at Blockworks. I’m not saying that about Ramp; I do love Ramp.
Our team hates when I talk about Ramp because I talk about it so much, but there is a chance that that’s the case in 2 years.
100%. Look at Bending Spoons. They acquired Airtable at around $1 billion, just clearing the preference. Then Miro just got acquired.
Miro was valued at $19 billion—$17 billion?
100%. If you were invested in Miro’s last round, that was a $17-billion-and-change valuation. ICONIQ led that round, so you wanted to get into it, and then they acquired it for $1.3 billion.
Speaking of this, Yan, did you guys do this? There were some very discounted secondary deals about a year ago in crypto, with a lot of the Series B and Series C companies.
No, we don’t really do too many later-stage things. It’s just hard.
What would that have been like? There were many custodians and prime brokers—people who had raised at, let’s say, $7 billion or $8 billion, or somewhere between $5 billion and $10 billion—whose secondaries you could buy for between $500 million and $1.5 billion.
No.
Yeah. Even Kraken had wild secondaries.
Kraken secondaries you could get between $3 billion and $7 billion, and now they just got an investment at around $22 billion.
That’s right. Was it $22 billion?
$21 billion or $22 billion, yeah.
$20 billion.
It was a clean 2x on the last round a year and a half ago.
Yeah, we did that round.
12. Is Market Froth Taking Over?
Well, I hear you, man, but what I’m trying to say is that there’s a lot of frothiness in AI. The way that round came together screams to me that there’s a lot of froth in the private markets, because there’s a lot of FOMO trying to chase the next SpaceX. There have been phenomenal outcomes, and Anthropic is going to be massive. I heard that the value generation there will be 4 times larger than all the IPOs that have come out of San Francisco-based companies combined. They were talking about the real estate in San Francisco, and it’s just insane.
Insane, right? But I don’t disagree with you. Both can be simultaneously true: there’s a lot of frothiness in the private markets in AI, and there are also unsustainable valuations in crypto.
Most tokens have traded below their last private round. If you look at the pre-TGE valuations and where those tokens are trading today, I think 99% of them are down. There are clear gems in there that you could have bought at a $10 million fully diluted valuation.
I just wonder what that sector would be. So Yan, thanks for coming on, man.
Appreciate it. Thank you for having me. It was a blast.
You’ve got a nice background here, making us jealous. Oh man, I want to fly out there.
I'm not in the Caribbean or tropical island.
Yeah, Yan. Good to see you, man.
Good to see you guys, too.
Cheers, folks.