Is The AI Trade Over? | Jack & Max on Panic in AI Credit, SpaceX Earnings, and Semis

Anthropic

Across the shows: 27 speakers in other episodes, 17 bullish and 36 bearish calls, past 30 days

• Max Wiethe · Monetary Matters NetworkBearish

  Anthropic's Q3 revenue growth slowed sharply from the first half and missed expectations; that is not good.

  Max accepts that Anthropic's revenue growth slowed tremendously from the first half of 2026 to Q3. It was not what the company's investors or analysts expected. He says this is not spin and that it is not good, and he will not explain it away. He treats it as a real slowdown rather than the narrative that Anthropic's problems are caused by Oracle's data-center delays or by a lack of money. This is his main reason for caution on the AI complex.

  “their revenue growth slowed... It slowed tremendously from the first half of this year to what we saw in q three. Like, there's no disputing that. It's not good.”at 21:09

• Max Wiethe · Monetary Matters NetworkBearish

  Anthropic's enterprise adoption has flattened in recent months after rapid growth through September 2026.

  In the Ramp AI Index, Anthropic's share of businesses paying for it rose from 17.8% in December 2025 to September 2026. Over the last few months that growth flattened, and business adoption slowed. OpenAI has sat near 40% since the start of the year. Most businesses spending on AI use both. Max says there are certainly concerns, even though the overall index still rises.

  “Anthropic really ramped up from 17.8 in December 2025 to September 2026, and that has flattened. It has flattened over the last few months with the growth in terms of business adoption really slowing.”at 31:20

Amazon.com

Across the shows: 10 speakers in other episodes, 6 bullish and 3 bearish calls, past 30 days

• Max Wiethe · Monetary Matters NetworkBullish

  Amazon's sold-out compute means the market is tight everywhere, so CoreWeave's fall doesn't add up.

  Amazon said it is sold out on compute for years, and Amazon stock rallied while CoreWeave fell. Max finds this inconsistent. If Amazon is sold out, the compute market is tight, and customers who cannot get capacity there will go to other providers. That should support neoclouds like CoreWeave, not hurt them. Jack adds that AWS customers may be stickier, but Max still sees the reaction as odd. He is not certain the sold-out comment was made directly by Jassy.

  “Amazon is a buy today because they're saying they're sold out. Well, if they're sold out, the market is tight everywhere.”at 11:29

NVIDIA

Across the shows: 20 speakers in other episodes, 16 bullish and 7 bearish calls, past 30 days

• Max Wiethe · Monetary Matters NetworkNeutral

  Nvidia's muted reaction to big compute news shows the market no longer rewards AI capex headlines.

  In earlier phases of the AI bull market, a compute announcement of SpaceX's size would have pushed Nvidia up 15% to 20%. This time Nvidia was up only about 2% after hours on SpaceX's announced 2027 compute. Max uses this to show that the market's reaction to good AI capex news has weakened, even though the news was positive. He does not give a directional call on Nvidia itself.

  “if a company of the size of SpaceX came out and said, we're gonna bring on this much compute, like, NVIDIA would be up 15%, 20%, something like that, and you're just not getting that.”at 16:53

Oracle

Across the shows: 12 speakers in other episodes, 0 bullish and 11 bearish calls, past 30 days

• Max Wiethe · Monetary Matters NetworkNeutral

  Oracle's force majeure over a data center was about permitting, not a lack of money.

  Recent coverage suggested Oracle declared force majeure on a data center because it lacked the money to build. Max says the story was entirely about permitting and the physical constraints on building data centers. He thinks people misread the article. The point is that narratives about AI buildout delays need to be checked against the actual reasons for the delay.

  “Oracle declared force majeure because they couldn't build this data center... And it was entirely about permitting, right, the constraints on getting these things physically built.”at 20:49

xAI

• Max Wiethe · Monetary Matters NetworkBullish

  xAI is becoming a real contender, with its business share roughly doubling since December.

  In the Ramp index, xAI now accounts for 4.1% of businesses with AI-model spending, up from 2.1% in December. Max notes that base effects inflate the change, but says there are certainly other horses in the race now. It is one of the most discussed developments in enterprise AI adoption. He is positive about xAI's momentum, but the evidence is a small, base-effect-influenced share.

  “X AI now represents 4.1... Has 4.1% share of businesses with spending on AI models... there are certainly other horses in the race now.”at 31:55

CoreWeave

Across the shows: 24 speakers in other episodes, 6 bullish and 18 bearish calls, past 30 days

• Max Wiethe · Monetary Matters NetworkUncertain

  CoreWeave's fate depends on whether compute prices stay high; the spending trend is not clear-cut bearish.

  Max says the SpaceX call's implied compute price suggests CoreWeave is fine. The real question is whether compute prices can stay high or whether cheaper models will make customers overpay. He does not have a definitive answer. The trend in spending, which he has been tracking, does not look as clear-cut as skeptics say. So the outcome is uncertain and depends on pricing that is not yet visible.

  “the implication from what the cost of compute is from the SpaceX call would indicate that that core weave is just fine... I don't think it's as clear cut as a lot of the skeptics would say.”at 33:00

Market sentiment, positioning & volatility

• Max Wiethe · Monetary Matters Network

  US fundamental long-short funds at zeroth percentile of net equity exposure; AI trade not crowded.

  Max cites positioning data, about ten days old, showing US-focused fundamental long-short hedge funds at the zeroth percentile of net equity exposure. Globally, the readings ranged from zero to ninth percentile depending on geography. The aggregate hides sector detail, since funds could be long AI and semis while short health care. Still, he says the overall picture is not people taking a lot of risk, so the everyone-is-all-in narrative doesn't match the data.

  “The fundamental US focused fundamental long short hedge funds were at the zeroth percentile in terms of their net equity exposure ... just two weeks ago, about about ten days ago.”at 50:19

• Max Wiethe · Monetary Matters Network

  Long-semis/short-software pair trade looks washed out; software fell as much as semis in the flush.

  On the day of the CoreWeave bond blowout, the software ETF IGV fell as much as the semiconductor indexes SOX and SMH. That showed the common long-semis, short-software pair trade had been washed out, since software was dropping alongside chips. Max doubts any risk manager at a legitimate hedge fund would let traders offset AI longs with software shorts after the July deleveraging.

  “IGV was down as much as the SOX and SMH were. Right? Like, if it was. Yeah. Software was down. So the long semiconductors, and we're short software against it trade. That indicated to me that, like, that trade has been washed out.”at 51:27

US Treasuries & yields

• Max Wiethe · Monetary Matters Network

  A slowdown in AI would lower yields through slower growth and lower corporate bond supply.

  Earlier yields rose because strong growth, driven largely by AI, pushed them up, and because corporate issuers building compute competed for capital in fixed income. If the AI build-out ends, growth falls and bond supply falls. That is a dual hit that pushes yields lower. The recent bond rally, despite a hot inflation print, looks like a growth scare tied to the AI slowdown. He says the bond market is pricing a growth slowdown.

  “if that's all coming to an end, it's kind of a dual ding of, well, growth is gonna be lower, but also supply of bonds is gonna be lower.”at 7:35

AI infrastructure & data centers

• Max Wiethe · Monetary Matters Network

  Big tech will likely reaffirm 2027 AI spending forecasts in this earnings season.

  Max says investors have relative certainty about 2027 spending. He expects the upcoming earnings season to bring reaffirmations of the spending forecasts companies have already made. There may be small raises, like SpaceX's compute increase. This gives near-term confirmation of 2027 build-out plans, but the stock reaction will depend on clarity about 2028 and 2029.

  “we know what people are spending... I think we're gonna get that this earning season. People are going to come out and reaffirm the forecast that they've had.”at 16:28

AI bubble & tech valuations

• Max Wiethe · Monetary Matters Network

  Without clarity on 2028-29 build-out, AI stocks could trade sideways or down for a long time.

  Max says he does not believe the AI trade is over, and that is not his base case. Investors have relative certainty about 2027 spending, which earnings will reaffirm. But the stock market is looking past 2027 for clarity on how long the build-out lasts into 2028 and 2029. Without that clarity, the stocks could go sideways or decline for a long time despite no real fundamental slowdown. He notes that Nvidia's reaction to SpaceX's compute news was muted compared with earlier phases of the AI bull market.

  “it's not really about 2027. That's not the question. And so if we aren't gonna get clarity on 2028, 2029, like how long this build out is really gonna go, you could just trade sideways or even worse, trade down for a long time despite no actual fundamental slowdown.”at 17:21

International equities

• Max Wiethe · Monetary Matters Network

  Korean equities topped in June and are perhaps in a bubble that has now deflated.

  Max points to deleveraging around the situational awareness trade and the top of the Korean equity market in early June. He says it is safe to call Korea perhaps a bubble at this stage. A recent Korean IPO fell 30% on its debut despite being oversubscribed about 389 times. Foreign money that went into Korea has been flowing out since the unwind.

  “the topping of the Korean equity market, I think it's safe to call it perhaps a bubble at this stage considering that Korean equity markets basically topped in the beginning of June.”at 24:31

IPOs, M&A & deal activity

• Max Wiethe · Monetary Matters Network

  The AI IPO rush is insider-driven and has backfired, and oversubscription doesn't guarantee gains.

  Max argues that VCs and early holders had an incentive to rush AI IPOs such as SpaceX and Anthropic, after a strong year for venture returns. He says they were talking their book to lock in wins. The rush has backfired for VCs, and the Korean IPO that fell 30% despite being oversubscribed about 389 times shows that oversubscription does not mean the stock will trade up after listing. He says the Anthropic IPO's oversubscription would not guarantee gains either.

  “there was a lot of incentive to try and, like, let's just get all of these IPOs done, and that's why we got those crazy numbers. And and it has backfired.”at 25:20

Frontier AI labs & model race

• Max Wiethe · Monetary Matters Network

  Most everyday AI work doesn't need superintelligence-level frontier models, questioning the race's payoff.

  Max says the assumption was that everything would be done with frontier models solving harder problems. But most work does not require a superintelligence to get done. Reading emails or sorting podcast transcripts does not need it, and most people don't need to be at the cutting edge. He takes this as a reason to doubt the value of the frontier-model push.

  “well, actually, maybe most of the work doesn't require, like, a 250 IQ superintelligence to get things done.”at 40:04

Credit markets

• Max Wiethe · Monetary Matters Network

  High-yield yields near teens and spreads widening; attractive on absolute return if credit risk is underwritten.

  Higher risk-free rates and high-yield spreads widening for the first time in a while have pushed yields toward the teens. Max says that makes high yield attractive on an absolute-return basis for investors willing to underwrite the credit risk. He says these levels have not been seen in years.

  “Like, if you, you know, take the time to underwrite, you know, the credit risk and you feel comfortable with it, like, we are approaching levels that, like, we truly have not seen in years.”at 43:00

Earnings & corporate profits

• Max Wiethe · Monetary Matters Network

  S&P 500 beat rate hit 86.3% in Q2 2026, well above the 77% forty-quarter average.

  Max checked historical S&P 500 earnings beat rates while testing Jack's claim that analysts are sandbagging estimates. The forty-quarter average beat rate is 77%. In Q2 2026 it was 86.3%. Even the weakest reading, Q1 2020 at 65%, was still a beat rate above that level in a bad quarter. Max also says companies can misrepresent the health of their businesses, which is separate from the beat-rate data.

  “The forty quarter average of beat rate is seventy seven percent. And right now... In q two twenty twenty six, it was eighty six point three percent.”at 55:57