Oct 20, 2026 · 59 min
AI financing strain tests the boom’s staying power
Is The AI Trade Over? | Jack & Max on Panic in AI Credit, SpaceX Earnings, and Semis
The episode separates a genuine collapse in AI demand from a financing, valuation, and positioning shakeout inside a still-expanding market.
- 1CoreWeave’s borrowing and exposure to compute prices make credit conditions central to the AI trade’s durability.
- 2Anthropic’s decelerating growth raises competing explanations, from high costs and open-source models to stronger offerings from OpenAI.
- 3Weakness in semiconductors has not broken the broader market, partly because hedge funds may be rotating rather than abandoning risk.
Don't miss
Jack discovers live on air that SpaceX and Tesla have formed a special committee to discuss a merger.
The brief
Jack Farley and Max Wheatley open on SpaceX’s bullish outlook and the market’s split verdict on whether AI enthusiasm has peaked or merely changed its branding.
The discussion moves from slowing AI-company growth to physical infrastructure limits, arguing that permitting, compute costs, and model competition complicate simple claims of collapsing demand.
CoreWeave becomes the credit-market test case: heavy borrowing and high compute prices could transmit an AI slowdown through convertible, high-yield, and private-credit markets.
Market resilience complicates the bearish case, as semiconductor selling coexists with low hedge-fund net exposure and possible sector rotation rather than universal retreat.
The standout moment arrives when Jack discovers that SpaceX and Tesla have formed a special committee to discuss a merger, adding corporate drama to the market debate.
The episode closes by weighing dot-com-era optimism against unusually strong recent earnings-beat rates, leaving the AI question focused on financing rather than demand alone.
What was said on this episode
25 statements · 9 positive · 11 negative · 1 mixed · 4 neutral
CoreWeave’s unsecured 2032 bonds imply roughly 60% cumulative default probability.
“Those bonds, The unsecured bonds 2032 is pricing in a roughly 60% of default cumulatively from now until 2032”
Listen at 3:16
A large share of expected semiconductor profit growth is concentrated in memory.
“a very large percentage of the profit growth, the expected profit growth in AI, excuse me, in semiconductors is specifically in memory.”
Listen at 3:54
“Superintelligence” is a better label than artificial intelligence.
“I think that President Trump is right that that's a better name.”
Listen at 5:42
SpaceX superintelligence revenue rose 195% quarter-over-quarter and 830% year-over-year.
“SI revenue was up 195%. quarter over quarter, 830% year over year.”
Listen at 9:40
SpaceX receives $1.2-$1.3 billion monthly through 2029 under its AI arrangement.
“they basically are receiving 1.2, 1.3 billion per month through 2029.”
Listen at 9:57
SpaceX’s implied AI compute price is roughly $5.30 per GPU-hour.
“So $5.30 per GPU hour is a rough estimate.”
Listen at 10:20
An AI-trade collapse could drive the S&P 500 down 30% or more.
“If you told me, oh, the AI trade is going to collapse, I think the S&P, a decline of 20% is conservative for the S&P. I think it probably declines. 30% or even more.”
Listen at 12:14
The rotation into non-AI stocks is likely a short-covering or underweight rally.
“I caution that the rally will be anything other than a short covering rally or really an underweight rally.”
Listen at 15:09
AI-related stocks could stagnate or decline despite no fundamental slowdown.
“you could just trade sideways or even worse, trade down for a long time, despite no actual fundamental slowdown.”
Listen at 17:30
Oracle’s data-center constraint was physical permitting, not financing.
“The constraints on getting these things physically built.”
Listen at 20:57
Anthropic’s revenue growth slowed substantially in Q3.
“Anthropics Q3, their revenue growth slowed, like that is true. It slowed tremendously from the first half of this year to what we saw in Q3.”
Listen at 21:07
Anthropic revenue increased from $4.7 billion in Q1 to $16.8 billion in Q3.
“The revenue for the first quarter was $4.7 billion. Revenue for the second quarter for Anthropic, $11.6 billion. And revenue for the third quarter was $16.8 billion.”
Listen at 22:12
Anthropic revenue grew 45% quarter-over-quarter from Q2 to Q3.
“The rate of change from Q3 to Q2 was 45%.”
Listen at 22:27
Most of Anthropic’s reported net loss reflects an accounting change.
“the vast majority of that net loss is an accounting change”
Listen at 28:40
AI spending trends do not clearly support the strongest skeptical interpretation.
“I don't think it's as clear cut as a lot of the skeptics would say.”
Listen at 33:36
CoreWeave bonds at 15% yield imply 50%-60% cumulative default probability by 2032.
“at 15 percent it's implying a, like a cumulative default probability of, of, you know, 50 to 60% by, by 2032.”
Listen at 43:39
Investors usually panic at the wrong time during market stress.
“the time that when people want to panic is often, almost always, not the time. to actually panic.”
Listen at 44:52
The short-term AI trade appears more likely to rise than fall.
“the short-term trade seems to be higher.”
Listen at 45:21
Disruptive technologies generally eventually develop bubbles.
“over the long term... almost all disruptive technologies have ended up in a bubble”
Listen at 45:47
Retail investors are unusually underweight semiconductor stocks relative to three-year history.
“the retail community is among the most underweight semiconductor stocks that they've been in the past three years.”
Listen at 49:24
U.S.-focused fundamental long-short hedge funds had zero-percentile net equity exposure.
“The fundamental US focused fundamental long short hedge funds were at the zero with percentile in terms of their net equity exposure just two weeks ago”
Listen at 50:36
Lam Research shares will likely perform well on Thursday.
“I think it's going to be a good day for LAM on Thursday. That's my bet.”
Listen at 55:06
The S&P 500’s average earnings-beat rate over 40 quarters was 77%.
“The 40 quarter average of beat rate is 77%.”
Listen at 55:59
The S&P 500 earnings-beat rate was 86.3% in Q2 2026.
“in Q2, 2026, it was 86.3%.”
Listen at 56:08
Musk may ultimately consolidate his companies into one parent company.
“Musk does have a history of consolidating his empire into one top co. It wouldn't surprise me if this is the ultimate consolidation here.”
Listen at 57:11
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.
Weekly email
State of the Pod, every Monday
The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.