SpaceX's IPO prospectus puts the size of the market it can address at $28 trillion. US GDP is about $32 trillion.
Most fund managers who worry about an AI bubble say so quietly, if at all. Zeke Ashton says almost none of them are quiet about it privately — they just don't say it on television.
"Every active manager that I talk to even those who are publicly much more bullish they will tell you privately that they are looking for ways to find some sort of diversification to that trade."
Ashton ran the hedge fund Centaur Capital for 17 years, including 2008 and 2009, when he protected capital through the crash and then made it all back the following year while many S&P 500 investors took five years to get back to even. He now runs Ashton Capital.
I listened to the full episode so you can skip it. 60 minutes of audio, 15 minutes of reading.
Here are the 10 lessons that matter.
👤 Guest: Zeke Ashton, founder of Ashton Capital, who ran the hedge fund Centaur Capital for 17 years and managed a mutual fund alongside it
🎙️ Hosts: Tobias Carlisle, founder of Acquirers Funds and author of The Acquirer's Multiple; and Jake Taylor, Chief Executive Officer of Farnam Street Investments and author of The Rebel Allocator
📰 Published: 10 September 2026 on YouTube (The Acquirers Podcast)
🔴 YouTube | 🟣 Apple Podcasts | ⏱️ 60 min | ✅ Time saved: 47 min
Key Takeaways
SpaceX's own IPO prospectus puts its addressable market at $28 trillion, against $32 trillion of US GDP
Ashton expects Anthropic's prospectus to claim $30 trillion, against roughly $130 trillion of global GDP
Every bullish AI manager he knows is privately hedging the trade
Meta and Google no longer generate free cash flow, because AI capex ate it
Apple is the one mega-cap that stayed out of the capex arms race entirely
If Anthropic or OpenAI can't honor their compute commitments, he doesn't know what that means for the companies that built around them
He bought a basket of stock exchanges — CME, ICE and OTC Markets — as a hedge that benefits whichever way volatility breaks
CME owns a 27% stake in S&P Global's index business, which he calls "a call option on the S&P 500"
Cap-weighted indices are a momentum trade, and passive flows keep feeding it
The S&P 500's top 10 stocks are 40% of its value
A Venetian financial trick from 1262 still describes how bonds can quietly confiscate wealth
US Treasury bondholders from 1941 to 1980 lost two-thirds of their real principal to inflation without a single default
200 purely leveraged ETFs have launched this year, and zero-day options are normalizing hyper-risk
One small ETF lost roughly 80% in a month from rolling zero-day options every day
He caps gross exposure at 200% and treats options as a way to take non-recourse leverage
His closing rule after two bear markets: "there are no participation trophies"
1. What 2000 actually felt like
Ashton spent the dot-com top watching it from inside the business, running the Motley Fool's investing operation before starting Centaur Capital in 2002. He described a market that "felt very much like this one," with Cisco trading at 170 times earnings on the strength of doubling internet traffic — "very similar to how Nvidia feels today"
His central lesson from risk-management work before that: "I never forgot that half my job is being a risk manager and the other half is being a risk taker"
The Nasdaq's collapse is hard to convey to anyone who didn't live through it. "We all watched the Nasdaq decline by 80% over an 18-month period. And it's very hard to tell somebody who did not live through that period how that feels"
His own best stretch as a manager was the 2008 crash, not a bull run. He protected capital in 2008 and made it back by late 2009 in both the hedge fund and the mutual fund he ran alongside it, while many S&P 500 investors took roughly five years to recover
2. TAM Looks Stranger Now
The addressable-market numbers being floated now dwarf anything from the dot-com era, in Ashton's view. "SpaceX, ... state their TAM is $28 trillion. US GDP is about $32 trillion." He expects Anthropic's own IPO filing to claim a $30 trillion market against a global economy of roughly $130 trillion
"So, when Nvidia's market cap is 5 trillion out of entire global GDP of 130 trillion, it feels like it's a lot"
He doesn't think Nvidia looks expensive on a trailing earnings basis — the risk is elsewhere. The real question, in his view, is whether Nvidia can sustain today's revenue level if AI never produces a killer application capable of generating a real return on all the capital being spent on it
Hyperscaler capex is heading toward a striking round number: "$1 trillion in a given year looks like that's where we're heading. That's just incredible to me"
His actual worry isn't whether AI blows up — it's the scale if it does. He said a disappointment would likely be bigger relative to the capital markets than both the dot-com bust and the 2007–2008 housing crisis
3. Capital-Light Turns Heavy
Meta and Google's cash-generating engines have stopped throwing off free cash, in Ashton's account, because of AI spending. "There's no free cash flow at Meta anymore. There's not any free cash flow I think at Google anymore." Microsoft still generates real free cash flow, and Apple, he noted, has stayed almost entirely out of the AI capex race
The question that decides these stocks, in his framing, is whether the spending is temporary or permanent. "If this is a capex trap where they have to continue to spend money to stay at the leading edge of LLMs or whatever they're trying to do, then I think the valuations have to come down, right?" He drew a distinction between Alphabet, which he sees as defending an existential search franchise, and Meta, where AI spending feels optional to him
The customer-concentration risk in this cycle has no real precedent for him. Cisco had plenty of customer concentration by type but not by single counterparty. "Whereas if you look at if you look at the capex requirements that are coming mostly from Anthropic and OpenAI if something bad were to happen at one of those two companies and they were not able to honor the agreements they've made I don't know what that means"
He specifically worries about OpenAI: "It seems like they still need a couple hundred billion dollars of capital to bridge them to whatever their ultimate profitability is going to be"
4. His Hardest Market to Short
2025 was one of his toughest years for shorting, because the market kept shrugging off real shocks. He pointed to the Iran conflict and April 2025's tariff turmoil as events that could each have triggered a serious correction, and said only the bounceback's speed, not its cause, surprised him
He no longer shorts individual stocks; he buys in-the-money puts instead, so losses beyond a certain point are non-recourse and he avoids the risk of a runaway short squeeze like the one that hit anyone caught short GameStop
He still carries index-put hedges even though they haven't paid off yet. "It's a very difficult market to short." With the VIX near a yearly low — "I think that the VIX hit a yearly low today or yesterday" — he said, "And so I'm willing to sort of take a little tax on the portfolio gains in order to protect from what I feel is, a very vulnerable market if the right situation were to occur"
5. Passive Flows Prop It Up
He thinks cap-weighted index investing has become the market's dominant, self-reinforcing trade. "Cap weighted indices are basically a momentum trade and always have been," he said, and noted, "Like the S&P 500, the top 10 companies is 40% of the value" — a figure he said makes "diversifying" into the Nasdaq 100 largely redundant given the overlap
US households now have record exposure to stocks. He cited a chart showing American households hold the highest share of their wealth in US equities they ever have, which he said is mostly exposure to the same handful of mega-cap names twice over
The danger, in his view, is behavioral rather than structural. He compared today's belief — "and that's my concern now is that people's belief now is hey risk doesn't matter as long as I'm prudent and just continue to put money in the S&P 500 every month" — to the pre-2008 belief that US housing prices never fall nationwide, a belief that changed behavior more than any underlying fact did
6. Betting on the Exchanges
His answer to a market he can't easily short or safely diversify away from was to buy exchanges. "We own CME, we own ICE, which is Intercontinental Exchange, and then I own as well the OTC markets, which is the small exchange" — a business he said benefited once trading volume ticked up during a rough stretch for AI-loser stocks in June and July
CME carries a hidden asset most investors overlook. "CME also owns kind of this hidden asset. They own a 27% stake in the S&P index business of S&P Global," which he called a royalty and a call option on continued S&P 500 inflows in one position
The valuations, on his numbers, stayed reasonable through the run. "Intercontinental Exchange was trading at and I think it still is trading at less than 20 times free cash flow," on businesses he described as roughly "60% profit margin"
7. Diversified Either Way
The obvious objection — that exchanges are just another way to bet on the top of the cycle — doesn't hold up once the three are combined, he said. CME's exposure extends well beyond equities into oil, metals and interest rates, so he expects volume to stay strong somewhere even if any single market goes quiet
The exchanges don't only profit from speculation — they profit from hedging too. "The CME and ICE, they don't just benefit from people taking risk they benefit from people hedging risk," which he expects to matter more as investors look for protection in a market he sees as unusually correlated across assets
Inflation itself is a tailwind for the trade, in his view, a point Carlisle drew out as a "secondary or derivative" way to play inflation without owning the underlying commodities directly
8. Eels, Venice, Confiscation
Taylor opened with the migration of European eels into fishing traps called lavorieri, then connected it to Bill Bernstein's "four horsemen of deep risk" — the ways an investor permanently loses capital. "So we have inflation, deflation, devastation and confiscation." He argued confiscation has a third form beyond taxation and outright seizure: "It's building a trap on the route that your money has to travel"
His historical example was Venice's 13th-century forced war loans, the prestiti. The state fixed a 5% yield, never set a repayment date, and let holders trade their claims — "And in 1262, Venice swept all of them into one big fund and called it the monte vecchio, which means mountain of debt." When war with Genoa threatened the republic in 1379, those bonds — normally trading near par — "traded down to 18"
His American parallel was the war bond, and the toll it took was not a default but inflation. Wartime price controls held T-bill yields near three-eighths of a percent, and when they lifted, inflation hit 17% by June 1947. Over 1941 to 1980, Bernstein's math says, "So, a real dollar in long Treasuries ended up being worth 34 cents over that time period" — with nobody defaulting or confiscating anything outright. "You just swam into the wrong weir basically"
Taylor gave three warning signs of a modern financial weir in one sentence: "An incentive or institutional pressure to enter meaningful friction preventing you from leaving and a return that's set or constrained by somebody who benefits from keeping it below a market clearing rate" — citing annuities, savings bonds and some stable-value funds as candidates today
9. Leverage Goes Underground
New products are quietly building far more leverage into markets than headline margin debt shows, in Ashton's account. "There's been something like 200 ETFs launched this year that are purely leveraged," many of them single-stock products, and he estimated zero-day options now account for roughly two-thirds of a given day's total option volume
"A very little known story from July that came out was there was a very small ETF that lost something like 80% in the month of July or maybe it was the month of June purely because they were mostly rolling over zero day options every day," despite marketing itself on risk management
Perpetual futures are a newer entrant, approved by the CFTC, that he views as just another leverage channel. He explained the mechanics simply: "they don't have an expiry date," and gains or losses settle between counterparties every day rather than at some future delivery date. He values Intercontinental Exchange and CME partly because their internal clearing operations protect against a counterparty failing to meet a margin call — protection perpetual futures largely lack
He thinks options activity, not passive flows alone, explains why individual stocks move so much more than the index does. He cited a statistic that daily volatility in individual names is running about four times normal, even on days when the S&P 500 itself "is just sitting there at 15" on the VIX
10. How He Manages Risk
He never lets gross exposure — longs plus shorts — run much past 200%. "If you're long 125% and you're short 75% that's 200 gross," he explained, and said every point of exposure beyond that sharply raises the odds of an unanticipated blowup
He treats stocks that trade like "AI losers" as a factor in their own right, whatever industry they're actually in, and caps that basket at 20% of the portfolio — a rule that reduced volatility but also capped his upside when the AI-loser trade unwound sharply in July
Options serve as a deliberate risk-management tool, not just a way to speculate. Buying calls or puts instead of the underlying stock gives him "non-recourse leverage" — losses stop at the premium paid — and he still layers covered calls onto favored long positions when premiums look rich
His discipline for shorting is to avoid being "the photo negative" of his own long book — each short needs its own specific thesis for near-term weakness rather than simply betting against whatever he doesn't already own
Bonus Insights
Ashton's read on the most speculative market he's seen isn't this one — it was December 2021. He defined speculation as knowingly overpaying for something on the bet that someone else will pay more, and said 2021's meme stocks, NFTs and new crypto assets fit that definition more purely than today's AI trade, where revenue is real
A friend's macro thesis, which Ashton finds compelling but unproven, is that bubbles are caused by easy money rather than by any specific new technology. His friend's counterexample is 2021 itself, when "there wasn't a new innovative technology" behind the mania. "His view is no. It's always because there's easy money and there's too much money"
Once a trend gets visible winners, social media accelerates the herd, in his account. "It did not matter what it was. As long as it could be somewhat financialized in the sense that there was a bid and ask and people could kind of see which direction the market was going, they were going to play it." He cited Reddit's "loss porn and win porn." "And you can see that somebody's account value went from $50,000 to 13 million and back to 2 million." "And then they'll post their losses and sometimes you'll see an account that went from $7 million down to $500,000 in three weeks"
He listed recent examples of the same pattern rotating through new assets: "I mean, we've already seen, silver go from $40 to $120 and now back to 60," a 2025 gold spike, and quantum-computing stocks whose combined market value he still finds hard to justify against how far off real applications remain
His closing philosophy is that survival beats chasing returns, even though it's cost him relative performance. After 15 years of roughly 17% annual S&P 500 gains, "And what I've learned is there are no participation trophies," and a 50% index drawdown over two years remains entirely normal history whenever a run like this one ends
Products, Companies & Tools Mentioned
CME Group, Intercontinental Exchange and OTC Markets Group (The basket of exchanges Ashton's fund owns; CME also holds a 27% stake in S&P Global's index business)
S&P Global (Owner of the S&P 500 index business in which CME holds its 27% stake)
SpaceX (Its IPO prospectus states a $28 trillion addressable market, against roughly $32 trillion of US GDP)
Anthropic and OpenAI (The two AI labs Ashton says carry the capex commitments and customer-concentration risk he worries about most)
Nvidia (Ashton's example of a stock that isn't expensive on trailing earnings but depends on sustaining an unprecedented revenue run)
Meta, Google and Microsoft (Meta and Google have stopped generating free cash flow on his account; Microsoft still does)
Apple (The one mega-cap Ashton says has stayed almost entirely out of the AI capex race)
Ashton Capital (Zeke Ashton's current fund)
Books & Resources Mentioned
Deep Risk: How History Informs Portfolio Design – William Bernstein (The source of the "four horsemen of deep risk" framework — inflation, deflation, devastation and confiscation — that Jake Taylor built his eel-and-Venice story around)
If this was worth your time, send it to someone who has to have a view on this.
Get the latest market chatter as it happens: