Bitcoin Breaks $95k, Crypto’s Valuation Problem, & The Path To Real On-Chain Users

Summary

• Santi’s core call: ex-Bitcoin crypto at $1.5T is unjustifiable and this cycle proves it was priced in. Every headline imaginable arrived — Larry Fink evangelizing, JPMorgan launching a deposit token on Base — and prices still went down, meaning “there was zero margin of safety.” Nvidia trades at 30-40x earnings while “Ethereum is trading at 200 times sales. Make it make sense.” His verdict on ETH as a $400B asset generating $1-2B of non-recurring fees: “I just cannot make a case to buy Ethereum. I’m not getting paid enough. I’d rather punt AI.”
• The single most important metric is active on-chain users, and it isn’t growing. Crypto ex-BTC carries ~$1.5T of value on ~40M active users while Santi says OpenAI has 800,000 users and likely IPOs at ~$1T — a comparison he describes as “20 times more users.” Santi’s bar: “if we wake up one day and we have actually 500 million users on chain… I can actually start to get behind the valuation.” Until then he’d “rather candidly buy OpenAI at a trillion than hold any position in crypto.”
• The shorting paradigm may finally have arrived. Jonah revisits Don Wilson’s old line — “tokens don’t file for bankruptcy” — that let vaporware hold multi-billion valuations for years, and argues the dream that supported them is dying as insiders and bagholders capitulate: “maybe we’re just supposed to suck it up and short stuff.” EOS getting delisted from Coinbase after a $4–5B raise “kind of starts breaking the spell.”
• Bitcoin has graduated; everything else is a tech stock. Santi sees a world where BTC dips to 80K/75K while the rest of crypto draws down far more severely — 80-90% drawdowns are normal, not different this time. Both hosts stay long BTC: Jonah is adding below 90K on the affordability-crisis-means-stimulus thesis (“I want to be holding Bitcoin when that debasement occurs”), Avi expects a bounce because “every single thing is going down” — gold, NASDAQ, uranium — which reads as deleveraging, not a top.
• The investable thesis is killer apps, not infrastructure — crypto is replaying Cisco in 2000. Over $100B went into infra over the last decade; blockspace is now a commodity and “Ethereum acts like a federal government, wants a valuation as a federal government, but only collects state or city tax” while L2s take the rest. What rallied out of the 2000 bust were apps — Google, Amazon — and Jonah’s summation: “if you’re investing in infrastructure, you’re investing in Cisco in 2000.”
• The long-side setup under debate: quality cash-flow assets decoupling in the washout. Jonah sketches Hyperliquid (~$37) squeezing hard if foundation buybacks from real trading fees meet an illiquid market — “you just have to stay solvent to be able to play for that.” Santi challenges the framework but admits this is “where I’ve been wrong the most”: he’d double or triple a position in something cash-flowing that falls 20-30%, but can’t build a framework for memes.
• Santi’s own answer is private equity with a chain attached: Inversion buys real businesses (think Western Union at 4x P ratio — trust, brand, distribution), cuts costs with stablecoin rails, and directs the settlement flow to its own chain — “the company we acquire is that whale.” The industry’s real bottleneck isn’t tech, it’s go-to-market: “if it were so easy to convince users to come and use this technology, then I wouldn’t be wanting to acquire these businesses.”

Deep dive

1. Every narrative arrived, prices fell anyway — that means it was priced in

• The occasion: Bitcoin at 94K (above 100K last recording), the timeline anxious, and guest Santi (Inversion Capital) on the back of a blog post arguing “you just cannot justify the valuation of most projects in the space.” His framing of crypto’s disease: “a classic Silicon Valley meme of never show revenue, always sell the dream.”
• The load-bearing logic: the market got every headline it could have dreamed of — the regulator not discounting it, Jamie Dimon no longer fighting it, JPMorgan launching a deposit token on public infrastructure (Base) — and prices went down. “The market’s not stupid… it was very much priced in. There was zero margin of safety.”
• Crypto is “no longer the main character, right? It’s AI.” So ex-BTC assets now compete as technology plays — and the comparison is brutal: Nvidia at 30-40x earnings versus “Ethereum trading at 200 times sales. Make it make sense.” On ETH specifically: a $400B asset “cranking out one to two billion of fees that are not recurring, that are going to dry up in a bear market… I just cannot make a case to buy Ethereum. I’m not getting paid enough. I’d rather punt AI.”

2. “Tokens don’t file for bankruptcy” — but the shorting era may finally be here

• Jonah’s formative story, worth keeping whole: he once asked his old boss Don Wilson of DRW why obvious vaporware — teams that “packed up shop, buying villas in Sardinia” — held multi-billion valuations. Wilson’s answer: “Hey Jonah, tokens don’t file for bankruptcy.” Jonah took it as gospel, but argues the regime is changing: what supported those valuations was the dream of future revenues, and now “participants, insiders, crypto natives, bagholders are throwing in the towel… maybe we’re just supposed to suck it up and short stuff.”
• Avi’s evidence the spell is breaking: EOS got delisted from Coinbase — once “the hyped next-generation blockchain that was supposed to eat Ethereum,” it raised $4–5B, failed massively, and became “a massive value extraction event” (the buyer — likely Brendan Blumer of Block.one — paid ~$160M, the highest purchase price in Italy). “You can raise 4–5 billion, be faster in theory than Ethereum, and it doesn’t matter.”
• Santi’s important nuance — he is not claiming this time is different: “It’s like every other cycle… you should expect to see 80%, 90% drawdowns on these things, from majors.” The reason is structural: “there’s no real economic activity beyond speculation anchoring the valuation of these things,” and in a macro bear “you need stimulus checks to really get people to come back and punt your altcoin of choice.”

3. The elephant in the room: 40 million users vs OpenAI’s 800,000

• Santi’s north-star metric, via a recent a16z report: active users on chain — “and that number hasn’t really gone up.” The per-user math he wants everyone to run: OpenAI has ~800,000 users (with Santi saying it has “20 times more users” than crypto) and likely IPOs at a trillion; crypto ex-Bitcoin is $1.5T on ~40M users. “If we wake up one day and we have actually 500 million users on chain — okay, I can actually start to get behind the valuation.” Until then: “I’d rather candidly buy OpenAI at a trillion and go long that than hold any position in crypto.”
• The casino mechanics behind the churn: “the longer you exist in the casino, the higher likelihood that you’re going to be served a drink, you’re going to get drunk, or you’re going to get liquidated” — and the vig is 1-5% per trade plus extractive MEV. A 24/7/365 casino is “a real industry, same as Vegas is real in Macau” — his question is only whether that’s worth $1.5T. “I don’t think that’s worth 1.5 trillion, because you churn through that constantly.”
• Jonah adds a competitive threat to crypto’s one moat: leverage. Crypto was the only place a $1,000 account could go 100x long on BitMEX; now Hyperliquid is introducing stocks, and within 2-3 years levered stock trading everywhere is “another punch in the throat for all the stuff in crypto that isn’t real.” Meanwhile the stock market rewards actual thinking — rare-earth plays, small-cap drone stocks. Avi’s line: “maybe penny stocks are the new crypto.”

4. Bitcoin graduated; everything else is a tech stock

• Santi’s taxonomy: “Bitcoin is like a commodity… I would characterize everything other than Bitcoin as a tech stock” — competing for attention and flows against AI, praying liquidity keeps flying into Vegas. Asked what an ex-BTC crash from $1.5T to ~$700 million does to Bitcoin, he sees decoupling: “Bitcoin may go down to 80K or 75K, but the drawdown on the rest of crypto is much more severe — and there’s a world where Bitcoin rallies and tokens just don’t move. You’re already seeing that this cycle.”
• The BTC case survives his own bear framework: less than 10% of gold’s market cap, the digital-gold thesis intact, vol structurally falling as ETF and institutional flows deepen. “Something like Bitcoin is the perfect meme… there’s only one real digital gold” — and, memorably, “institutions can’t really pronounce Solana or Ethereum. They definitely now understand Bitcoin.”
• His anti-thesis on everything else is about stacked assumptions: “I want to make sure Jensen Huang doesn’t puke, that Powell has a good day, that regulatory clarity shows up, that there’s no government shutdown, and then people continue to want to punt this… you stack all those together, you’re likely going to be wrong. It’s not a clean bet.” Priced-to-perfection assets die on “a mouse fart from Powell.”

5. Ethereum is a federal government collecting city taxes

• The episode’s best structural metaphor: L2s have gutted the L1’s take. “Ethereum acts like a federal government, wants a valuation as a federal government, but only collects state or city tax — most of the tax is being collected by L2s. The revenue generation of Ethereum the L1 no longer supports a credible $380-400 billion valuation. That’s kind of cooked.”
• Adoption headlines don’t equal value capture, and Santi walks the chain of disbelief: JPMorgan launched “on Ethereum” — but on Base. Base has no token. So buy Coinbase stock? “Just because we have more adoption doesn’t mean the value capture is going to be there.” His discipline: category-killer SaaS with contracted, recurring revenue gets 10-15x ARR; ETH at 200-400x non-recurring sales requires “way more Larry Finks and Jamie Dimons” than exist. On Tom Lee’s ETH thesis: “walk into a committee and give me the reason why I should buy an asset trading at 200, 300, 400 times price-to-sales when revenue is not recurring.”
• Blockspace itself is the stranded asset: “no longer in a state of affairs where you’re paying $1,000 for a transaction” — it’s a commodity, like the fiber glut of the internet era, waiting for consumer applications to eat the excess. And the moat can be forked: “Stripe launching Tempo is kind of a warning shot… this could be co-opted.”
• Avi, unrepentant: “I made a name for myself by being the biggest Ethereum hater that’s ever existed — two and a half years ago I said Ethereum was going to be the Ripple of this cycle” (the show even got a jokey cease-and-desist from the Bankless hosts). His read on DATs: they’re reintroducing 2021-style tribal bag-holding — “Tom Lee needs it to be true.”

6. The long-side setup under debate — and where Santi admits he midcurves

• Jonah’s setup: as the casino ejects the leveraged and the rage-quitters, quality could decouple. If Hyperliquid’s foundation is using fees from real economic activity to buy back tokens into an illiquid market, “I could see that thing squeezing pretty hard to the upside” — the dark-fiber-era pattern where most tokens die but “the Amazons and Googles of crypto” get sold hard and then re-rate. The catch, at ~$37: “you just have to stay solvent to be able to play for that and not have it go to $5 on you.”
• Santi’s pushback is about run-rating: he comes from enterprise software, where multi-year contracted revenue is underwritable. Pull up the Blockworks dashboard — Solana did ~$150M of total revenue (not earnings) over 90 days, Ethereum similar — and analysts run-rate the peak. “All I’m saying is: how do you think that revenue stream behaves in a down market? Because we’re not in a down market yet.” Crypto is “a hyper-macro asset class” — the whole relative-value game (ETH is a fraction of BTC, SOL a fifth of ETH, and Zcash a fraction of BTC) depends on central banks refilling the casino.
• Then a rare, flagged change of register — Avi asks whether Syrup, Hyperliquid and reasonably-priced DeFi can rise while Cardano-tier stuff bleeds out, and Santi concedes: “this is where I’ve been wrong the most” — he underwrote DeFi correctly in the early 2020s and the entire sector was still worth less than Dogecoin. “Maybe someone can say that guy totally doesn’t understand how flows in crypto work — total midcurve.” His resolution: stick to cash flows anyway. “If it falls 20-30%, I’ll actually put way more on the line and double and triple the position. I struggle with creating a reasonable framework for memes — even though they could be worth 10 trillion, God knows.”

7. It’s Cisco in 2000: the only investable thing from here is killer apps

• The through-line both hosts converge on: value accrual is shifting from infrastructure to applications. Avi’s history: people bought infra because “there were no applications to invest in” — it was catch-all exposure to the idea of crypto. But the email-protocol argument was always backwards: “the underlying protocol shouldn’t make a ton of money… it should be the things that live on top.”
• Santi’s version: the last 10 years were over $100B invested in infrastructure, and being early to Solana/Ethereum/Filecoin paid; “I don’t think that strategy works over the next 10-15 years.” The question now is user aggregation: “who’s going to capture more value — Robinhood, or your new startup stablecoin issuer?” Like the law firm that uses AI to be 10x more profitable per employee, the winners deploy the tech, not build it.
• Jonah’s summation of the whole episode: “The paradigm has shifted. Ignoring Bitcoin, the only thing investable here on out is killer apps. If you’re investing in infrastructure, you’re investing in Cisco in 2000 — AT&T, local broadband providers, the stuff that didn’t 1000x.” Cisco never recovered its all-time high; what came out of the bust were apps — Google, Amazon. Avi’s corollary: he’s stopped caring about MegaETH and Monad (“I’m just not excited about these things anymore — I’m excited about what’s being built on top of them”), and respects Plasma — down in a straight line as it is — for at least “trying to build a company, not a platform.”
• The MegaETH exchange stays honest: Avi got flamed for saying “we don’t need another L2 — the issue is not throughput, the issue is activity,” and credits Brad from the Mega team for a good-faith rebuttal. Santi, an early Mega investor, sympathizes with teams pushing throughput but holds the line: “blockspace is no longer scarce. How are you going to fill that? You want to fill it with quality demand, not pure speculative activity.”

8. Inversion’s playbook: buy distribution outright, pipe the flows to your own chain

• Why launch yet another chain? “For a very specific reason: we’ll go acquire businesses and plug in this technology to make them more efficient… and drive all that activity and settlement to our chain, because we don’t want to leak that value.” The target user “doesn’t even realize crypto’s working on the back end — they’ll get a loan, the rate is lower; they’ll get a stablecoin that for all intents and purposes is a dollar, earning yield like a money market account.” Structure: a private equity fund acquiring real businesses, Inversion Labs building the chain with a “likely Palantir model,” and an Inversion token settling the activity.
• The specimen case — Western Union (“too big for us right now, but”): trades at 4x P ratio because the market is “drunk on techno-optimism and leaves certain businesses for dead.” It has scale, distribution, and immigrant trust; its cost structure is dominated by sourcing local-currency liquidity while charging 2-4% and hiding FX spread. If the recipient keeps a digital wallet instead of converting, that cost collapses. “Cutting cost is the name of the game here, gents” — explicitly not leverage-and-fire financial engineering. A likely Howard Marks reference anchors it: price is the main determinant of returns; margin of safety is the game.
• Pressed by Jonah on how token holders actually get paid versus Maple or Plasma doing the same thesis without the PE angle: “Think of our go-to-market — the company that we acquire is that whale” (the Hyperliquid analogy: one whale trading generates the fees). Unlike RWA protocols doing BD, “we’re putting equity on the line to buy the business and we control it, so we can direct that flow to the chain” — each incremental acquisition brings a durable, sticky stream of settlement fees you can inspect on-chain and underwrite “apples to apples” against Ethereum or Solana.
• The confession underneath the strategy, and the episode’s most honest passage: crypto’s biggest problem is go-to-market. “People are lazy, man. They don’t want to switch bank accounts.” And the skeptic’s question has no good answer yet: “If stablecoins are truly amazing and everyone wants dollars, you let me know when we have a billion users using this stuff.” His Uber parable seals it: phenomenal service, terrible unit economics from customer-acquisition costs — “don’t confuse great service with a good investment… everyone’s a genius when rates are zero.”

9. Macro: stimulus over socialism keeps Bitcoin-to-a-million alive

• Jonah’s zoom-out: affordability is the crisis of our time — it’s what recent elections communicated and inflation is global. Governments have two outs: socialism (“bearish for everything”) or stimulus, and the current Western political class leans stimulus — “look who’s in charge. The guy wants his face on Mount Rushmore. He doesn’t want to be the next Herbert Hoover.” So: short-term inflows buoying all assets, plus a 25-year upswing in crypto adoption now that the tech “went from being illegal to legal” six months ago. He’s adding BTC below 90K, doesn’t expect forced OG selling, and believes the million-dollar thesis: “I want to be holding Bitcoin when that debasement occurs.”
• Avi’s tactical read: the selloff is indiscriminate — gold, NASDAQ, AI, uranium all down at once — “and what that tends to mean is general repositioning and deleveraging. Once that’s over, you get at least some sort of bounce.” He doesn’t think the stock-market bubble is over (“I wouldn’t even necessarily call it a bubble”) and expects a good end of year — but in crypto, “I’m out, ex-BTC. I’m back to being an ETH hater.”
• Santi signs off rotating “a lot of stuff into Bitcoin” and leaves the discipline that titled his blog post: rule number one is never lose money; rule two, don’t forget rule one. The post’s title — “Thank you, dear LPs: we outperformed Ethereum, but the fund’s still down 80%” — and the closer: “How many times has that goddamn letter been written?” Jonah: “Too many.”