Chip Stocks Look Cheap. AI Spending Is Surging. So Why Is Cameron Dawson Watching for Peak Earnings?

Space Exploration Technologies

Across the shows: 36 speakers in other episodes, 34 bullish and 16 bearish calls, past 30 days

• Dave Nadig · ETF.comBearish

  SpaceX hasn't held up and won't with lockup unlocks coming; its stock is at risk.

  The wealth of about 10% of new centimillionaires in the next three months depends on whether the stock holds up. He says SpaceX has not held up and does not expect it to with unlocks coming. Unlocks add supply to a market already having to absorb the next two big IPOs. He sees the stock as a bellwether for top-of-K wealth and spending.

  “It's all dependent on whether or not that stock actually holds up. And SpaceX hasn't held up, and I don't think it's going to with the unlocks coming.”at 17:00

Intel

Across the shows: 13 speakers in other episodes, 4 bullish and 11 bearish calls, past 30 days

• Dave Nadig · ETF.comUncertain

  Intel fell 30% on a beat; its realistic 2-3 year outlook was punished against priced-in expectations.

  Intel was down more than 30% on a quarter he calls fantastic, with about 80% data center growth. He says 4-5% of the move was priced in, so the beat was not enough. From his reading of the transcripts, management was fairly realistic about its two- to three-year prospects and said it would not grow 100% a year forever. He thinks the market punished that realism, which he sees as a warning sign for other names and an enormous wealth wipeout.

  “they were fairly realistic about what their two, three year prospects were. It's like, we're not gonna grow a 100% a year forever. Surprise.”at 22:26

Micron Technology

Across the shows: 16 speakers in other episodes, 12 bullish and 7 bearish calls, past 30 days

• Cameron Dawson · NewEdge WealthBearish

  Micron's 85% gross margin will likely normalize as in 2022-23, so earnings can fall while spending continues.

  Micron's forward gross margin is about 85%, and the street expects it to expand in 2027, hold stable in 2028 and decline about 200 basis points in 2029. Micron's gross margin last peaked at 45% in 2022 and troughed at negative 9% in 2023, so there is no historical precedent for a sustainable high plateau in gross margins. Her scenario: the US economy is fine and data center spending continues with positive top-line growth, but earnings fall because of margin normalization.

  “The last time that Micron's gross margin peaked was from '22 into '23. It peaked at 45% gross margins in 2022, and it troughed at negative 9% in 2023.”at 24:32

US stock market outlook

• Dave Nadig · ETF.com

  Heavy IPO and unlock supply plus top-of-K selling could leave equity demand unable to absorb supply.

  Markets must absorb a large amount of equity supply from SpaceX unlocks and the next two big IPOs. At a Chicago event, CIOs said this raw capital flow issue was their top concern, more than consumer demand or the health of the core economy. He notes that each year about a third of equity demand comes from defined contribution plans, which is the baseline, and the rest determines whether the market rises or falls. He sees top-of-K allocations shifting out of hyperscalers and a rotational market where everyone derisks but stays fully invested, which is problematic when the top of the K funds spending from high equity values.

  “this enormous amount of equity where all the markets are just supposed to absorb both from SpaceX unlocks and from the next two big IPOs that are coming... every year, a third of equity demand comes from monolithic defined contribution buying.”at 15:39

• Matt Zeigler · Excess Returns

  A multiyear sideways market is on the table, with rotation risk and 15-20% drawdowns likely along the way.

  If the returns from AI capex do not arrive, the market can go flat for three years, as it did after the financial crisis during an earnings recession. He expects a 15-20% drawdown at some point and advises clients to plan for it. He thinks the rotation risk is more credible and durable than people recognise, because attention is on CoreWeave, CDS and Intel. A multiyear sideways market is totally on the table, and individual names will get blown up.

  “If the returns don't come, we can also go flat, and we can go flat for three years. ... a multiyear sideways market, I think, is totally on the table here because of that rotation risk.”at 26:56

• Cameron Dawson · NewEdge Wealth

  Defensives are leading and the June speculative top looks in; expect big countertrend rallies, not a straight-line fall.

  Defensive sectors (staples, utilities, healthcare, insurance) and dividend stocks have been the main source of market resilience. Midterm narratives favouring selling tech and buying other stocks could extend that rotation. The tone change in markets suggests the June blow-off top for bubbly speculative stocks is in place. That means the decline will not be a straight line: a stock can rally 20-30% in a countertrend move and still make new lows. In 2000 the pattern was an initial flush, then dispersion, with tech bottoming well after value, so initial corrections can look like sideways charts.

  “We've seen strength out of defensives. So that's been the key source of market resilience is things like Staples, utilities, health care, insurance doing better.”at 27:41

• Dave Nadig · ETF.com

  Index may grind flat while hyperscalers fall 30% underneath, a seesaw market hiding large rotation.

  About 300 S&P 500 stocks were already more than 20% below highs. Even on a day the Nasdaq fell 3%, around 50% of names were making new 20-day highs. He expects a seesaw market that could grind along in flat 3-4% ranges, with the index hiding large moves underneath, such as hyperscalers down about 30%, as Intel is now.

  “we could grind along 3%, 4% flat markets. While under the hood, we've got, you know, hyperscalers down 30% like Intel is today.”at 29:57

• Cameron Dawson · NewEdge Wealth

  Market trades ahead of earnings: 36% growth priced to fall to 12%, with risk nearer 7%.

  The market leads the earnings cycle. In 2015-16, the industrial earnings recession showed up in 2016 but markets bottomed in February 2016 when Yellen backed off planned rate hikes. Today the market could have traded at 23 times last year because it was anticipating 36% earnings growth. It trades at 18.5 times this year because it anticipates that 36% growth falls to 12%, with a higher risk it is closer to 7%. Her point is that the multiple is pricing in a sharp earnings slowdown.

  “the market will always trade ahead of whatever the earnings recession is. ... the market could trade to 23 times last year because it was anticipating 36% earnings growth. It's trading at 18 and a half times this year because it's anticipating that 36% becomes 12% with a higher risk that it's something closer to seven.”at 31:58

US consumer

• Matt Zeigler · Excess Returns

  Consumer resilience is holding in the data with no cracks visible yet, despite client fears.

  Clients are scared, but the consumer data is still fairly resilient, and he is not seeing the cracks. He cites tracker data and the labor market as the signals he watches. He is not convinced the consumer has fallen off a cliff, though he notes labor-market weakness as a risk.

  “the resilience of the consumer is the underlying piece in client conversations that I've been having, where it's get scared all you want, but the consumer data is still fairly resilient. And we're not seeing the cracks.”at 12:01

• Cameron Dawson · NewEdge Wealth

  Spending holds on top-of-K wealth effect while bottom-K households are in clear pain that aggregate data hides.

  Real spending is outrunning real income, which is only possible if households save less. Household net worth at new highs gives consumers room to save less, and equities now exceed homes as a share of household net worth. Wealth is concentrated among boomers and high-income households, who have high cash yields, low locked-in mortgage rates, soaring equities and home prices that have not cracked. At the bottom of the K there is distinct and clear pain. Aggregate spending data hides this because high earners carry more weight, while sentiment data shows it because every vote counts equally. Household consumption forecasts are higher than before the Iran war despite an energy price shock, partly from OBBB but mainly from wealth effects. A protracted equity correction with statement shock could hit consumption.

  “But then at the low end of the K or the bottom part of the K, I mean, there is distinct and clear pain.”at 13:12

AI infrastructure & data centers

• Dave Nadig · ETF.com

  AI buildout increasingly funded by equity and bonds, with raised money sitting idle as data-center projects delay.

  Over the last six months the AI capex boom has shifted to capital markets. Companies now do secondaries and float bonds to build data centers, which was unthinkable six months ago. That money is being raised, but the projects are being delayed, so a large pool of capital is effectively sitting idle waiting for deployment. He argues this has pulled a lot of capital out of the system, and it is a real problem today.

  “we have the AI CapEx boom, which is increasingly now leaning on capital markets... That money is being raised, yet the projects are being delayed.”at 14:39

• Cameron Dawson · NewEdge Wealth

  Capital funding AI buildout remains abundant but pricier; scarcity could emerge if deals stop clearing.

  This week the Meta equity deal and Google or Microsoft bond issuances were oversubscribed, which she cites as evidence that capital is not scarce but is growing more discerning and more expensive. The open question is whether this is a first step toward scarcity: if financing deals stop clearing and liquidity disappears, the market would have to rethink how much growth can be funded. Street forecasts show free cash flow getting worse in 2027 than 2026, then rebounding in 2028-29 as operating cash flow from the investment pays off, which she sees as a stretched assumption.

  “Meta equity deal oversubscribed. Google or Microsoft bond issuance oversubscribed. And so the argument we've been making is that capital is not scarce. It's growing more discerning and more expensive.”at 17:27

AI chips & semiconductors

• Cameron Dawson · NewEdge Wealth

  Semis' 17.5x forward multiple may mean markets see 2027 as peak earnings; street still expects growth.

  The semiconductor index made a high in June and has bled lower since, now trading at 17.5x forward earnings. Semis trade countercyclically: their multiple is highest when earnings are lowest and lowest when earnings are highest. Across recent cycles, multiples have troughed around 13-14x and peaked near 30x. So a low multiple here signals the market is pricing in that twelve-month forward earnings are peaking, not that the group is cheap. Either this is a buying opportunity with prices prematurely discounting peak earnings, as in 2025 and early 2026, or the market is increasingly saying 2027 is peak earnings. The street's 2027-2029 Sox earnings estimates keep rising each year, so she sees a dissonance she thinks is unresolved.

  “remember that semiconductors trade with countercyclical multiples. You put the highest multiple in the lowest earnings, the lowest multiple in the highest earnings. So when the semiconductor multiple drops, it's not that it's cheap, it's that it's pricing in that the current earnings, on a twelve month forward basis are peaking”at 19:49

• Matt Zeigler · Excess Returns

  Semis operating margins near 47-48% are close to peak, so earnings will recede as in past cycles.

  Semis carry a forward operating margin of about 47-48%. He sees the same playbook as before: margins are somewhere near peak, and when they recede, earnings fall. The street assumptions for hyperscalers already bake in a steep fall in free cash flow, from $175 billion in 2025 to zero in 2026 and negative $90 billion in 2027, which could also be the basis for a recession in semis earnings.

  “To the Micron piece, we've seen this playbook before. We're somewhere close to peak, and when that recedes, this is what happens.”at 26:33

US rates & the Fed

• Cameron Dawson · NewEdge Wealth

  Two-year yield prices three more hikes, but in 1994 and 2018 something broke and hikes never came.

  The two-year yield sits about 75 basis points above fed funds, implying roughly three more hikes, even though Fed officials talk about one more hike and no moves in 2027. Checking history, she finds only two episodes where the two-year priced hikes the Fed did not deliver: 1994 and 2018. In both, something broke and the Fed did not hike as the two-year expected. She frames the live question as what would have to break for the two-year to be wrong and the Fed to instead deliver cuts. This is a scenario she is testing rather than a firm call.

  “And in both of those times, the Fed did not deliver the hikes that the two year thought it would because something broke.”at 6:55

US Treasuries & yields

• Dave Nadig · ETF.com

  Global investors still treat Treasuries as the safe haven, so yields fall each time markets panic.

  Higher US yields have created demand for Treasuries. He observes that yields keep falling a little each time something panics. To him this means the rest of the world still sees the Treasury market as the safety asset, so Treasuries keep drawing safe-haven flows in each selloff.

  “we keep seeing... yields come down a little bit every time something panics, which to me means the rest of the world is still looking at the treasury as the safety market.”at 15:21

AI bubble & tech valuations

• Dave Nadig · ETF.com

  Hyperscalers and AI leaders are over-earning rather than expensive; the reset has already started.

  He frames the problem as over-earning, not expensive valuations. Many S&P 500 stocks, around 300, were already more than 20% below all-time highs, so the reset looks already underway. Hyperscaler earnings are at peak levels and are likely to come down, which he links to competition and margin normalisation rather than a bubble in multiples.

  “The way we would frame it is over earning. They're not expensive. They're over earning. Isn't that kinda what's already happened?”at 29:29

Credit markets

• Matt Zeigler · Excess Returns

  After the muni selloff, bonds are now total-return drivers worth rebalancing into, alongside infrastructure and real estate.

  Every blip in the market is a rebalancing opportunity, and his firm is shamelessly rebalancing into bond allocations, listed infrastructure and real estate. Munis fell 4.4% last month, the worst month for munis since 2008. He argues that bonds now belong in capital market assumptions as total-return drivers, because their yields make them an attractive allocation after the selloff.

  “Every blip in these reports, we are rebalancing, and we are gonna be shamelessly rebalancing into the bond allocations.”at 34:19