Tech Sell-Off, AI Panic and Market Rotation: Power Lunch

Apple

Across the shows: 13 speakers in other episodes, 3 bullish and 11 bearish calls, past 30 days

• Mark Newman · BernsteinBullish

  Apple is a defensive haven in tech because its low AI capex leaves little downside if the AI bubble bursts.

  Mark Newman argues Apple is seen as the defensive safe haven within tech because it is not spending the roughly $100 billion a year in capex that other companies are. Its capex was about $13 billion last year and is around $11 billion this year. Its free cash flow machine, continuous buybacks and growing earnings mean little risk if the AI bubble bursts. He expects the trade to flip back as sentiment on AI improves, and Apple to lag then.

  “I think it should be, because they are not spending the 100,000,000,000 or so in CapEx a year that some of these other companies are. So they are somewhat protected from this AI trade.”at 27:43

• Mark Newman · BernsteinBullish

  Apple has a strong moat as the consumer gateway to AI because the phone remains the main AI interface.

  Mark Newman argues Apple has a near-perfect moat even though it spends little on AI. He says Apple is the consumer gateway to AI because people want something easy to talk to while walking around, and that will mostly be on the phone. He says wearables will not have enough processing or battery power to replace it yet, so AI will need to tether to a device like the phone.

  “I think they do. I mean, I think they are, despite not spending anything in AI, they are the consumer gateway to AI. How you to use AI? It's mostly going be on your phone.”at 29:48

Intel

Across the shows: 10 speakers in other episodes, 3 bullish and 8 bearish calls, past 30 days

• Victoria Greene · G Squared Private WealthBullish

  Intel's 30% drop is overdone despite messy results and uncertain PC trough, so she is buying.

  Victoria Greene says Intel's outlook was not great and its results were messy, with concern about when the PC demand trough will end, which she puts around the first half of 2027 or Q4. She notes that extraordinary items from a fab fire in Israel also affected results. She still thinks a 30% drop in one day is overdone, especially since the stock is still up about 75% this year. She says she made tactical buys the previous day and added more Intel on the day of the drop.

  “I did do some tactical buys and now yesterday and absolutely picked up a little bit more [Intel] today.”at 35:50

VanEck Semiconductor ETF

Across the shows: 5 speakers in other episodes, 1 bullish and 4 bearish calls, past 30 days

• Craig Johnson · Piper SandlerBearish

  Semiconductors are in a confirmed downtrend with lower lows and highs; more pain ahead.

  The charts of SMH, the SOX, Intel and Nvidia show lower lows and lower highs, which meets his basic definition of a downtrend. Semiconductors had grown to about 20% of the investable market and technology to about 40%, both historical extremes. He says extremes like this tend to resolve painfully, and 'this is not done', so he expects more downside in chip stocks.

  “You're making lower lows and lower highs that use your basic definition of a downtrend.”at 38:08

US stock market outlook

• Tom Lee · Fundstrat Global Advisors

  Stocks could fall until the November 3 midterms because institutions lack dip-buying conviction, though today is a good buying opportunity.

  Tom Lee argues institutional investors do not have conviction to buy dips until after the midterms, so between now and November 3 it takes little for stocks to fall sharply. He cites that the equal-weight market is still rising, so there is plenty of money in the market. He also notes that many people are hesitant because of high-stakes midterm worries, such as a Republican Senate. He still sees today as a good buying opportunity, but expects the market to be bumpy until the midterms.

  “reality is our institutional investors just don't have conviction until after midterm. So I think between now and November 3, it doesn't take a lot for stocks to get... To fall a lot because I just don't know if there's as much dip buying conviction till after November 3.”at 4:49

• Steve Sosnick · Interactive Brokers

  Investors are rotating from risky to less risky assets rather than fleeing the market, with defensive sectors holding up.

  Steve Sosnick says investors are not exiting the market but reallocating from risky to less risky positions. Bond yields are falling, and defensive sectors were doing okay. He says the selling is still very much concentrated in tech. He attributes it to several factors, including institutional positioning and concerns about returns on AI spending.

  “What we're seeing is people not fleeing the market so much as reallocating from risky to less risky.”at 7:23

• Craig Johnson · Piper Sandler

  Tech concentration is at an extreme and will unwind painfully; rotation out is underway.

  Semiconductors reached about 20% and technology about 40% of the investable market, both historical extremes. Market gains have been highly concentrated in tech. Money is rotating out of tech into energy as a hedge against conflict, into health care, and possibly into financials. His model portfolio has cut technology to 18% against 40% for the market and has been rotated into energy and health care. He says extremes like this resolve themselves but tend to be painful, and this is not done.

  “Semiconductor stocks had grown to 20% of the total investable market. Technology is 40%. Those are historical extremes.”at 37:59

• Victoria Greene · G Squared Private Wealth

  Tech sell-off is a buyable dislocation; she expects payoff over 30 to 90 days.

  She sees the sell-off as an angsty, panic-driven dislocation: a bear market in two days with a Bitcoin hack, risk-off flows, and AI fears. Historically, buying dislocations during transformative technology periods has paid off, as with DeepSeek, the SaaS Apocalypse, and beaten-down cyber and services names. She does not expect a Monday rebound and warns against leverage or inverse ETFs, but she thinks the trade will pay off over the next 30 to 90 days.

  “I want to be a buyer of this because I think over the next thirty, sixty, ninety days, I think this will pay off for me.”at 40:40

Market sentiment, positioning & volatility

• Steve Sosnick · Interactive Brokers

  Crowded institutional positioning in tech left no buyers, so stocks fall even on good numbers.

  Steve Sosnick argues that institutions are now fully overweight popular tech names like Alphabet and Microsoft, and everyone already knows those stories. With nowhere left to go and no one to buy, stocks sell off even when the numbers are good. When the numbers are bad, overweight institutions rush out, and there is little support. He says this positioning was a big part of the recent tech selloff.

  “The positioning got so crowded, and I think that was a big part of it... That's why a lot of these stocks do poorly even when the numbers are good because there's no one to come in and buy them.”at 10:22

• Tom Lee · Fundstrat Global Advisors

  Sell-offs are driven more by thin liquidity and forced hedge-fund de-risking than by crowded trades, which is a overused explanation.

  Tom Lee says a liquidity vacuum explains many drops better than crowded trades, and calls 'crowded' a very overused word. He notes that many hedge funds have already lost a lot and could be forced to go flat, so indiscriminate selling creates price opportunity. He points out that Intel's 30% drop is a change in present value that reflects a shock and thin liquidity, not proof of crowding. He also says VIX above 21 signals some panic and indiscriminate selling, while history suggests such days are not certain repricings.

  “I would say to me, I think that liquidity explains a lot of drops more than crowded trades. Think that's a very overused word.”at 12:14

US Treasuries & yields

• Tom Lee · Fundstrat Global Advisors

  The 10-year yield has fallen back to levels supportive of risk assets, backing the view that the selloff is a buying opportunity.

  Tom Lee says history shows that days like this do not reliably reprice the market. He points to the 10-year Treasury yield, which has dropped back to levels he considers supportive of actual risk assets. He uses that backdrop to argue that today is a good buying opportunity.

  “history has shown days like this are not about the market repricing with certainty. In fact, look at the ten year yield. That's kinda back to levels that are very supportive of actual risk assets.”at 10:01

AI bubble & tech valuations

• Tom Lee · Fundstrat Global Advisors

  Investors are not overweight or overly bullish on AI; many are under-allocated to tech, which drives earnings.

  Tom Lee argues that tech is the most important driver of earnings, so investors need to own it. He says he does not think people are overweight tech or that consensus is overly bullish on AI. His conversations with clients show mixed views on AI, and he thinks many investors are under-allocated. He uses this to reject the idea that stocks are falling because they are crowded.

  “People need to own tech because that is probably the most important driver for earnings... I would say I don't think people are overweight and consensus overly bullish on AI.”at 12:29

Private credit

• Ken Kensell · Churchill Asset Management

  Private credit underlying credit quality is benign, but expect manager dispersion as weaker firms are exposed.

  Ken Kensell says things in private credit and lending have been pretty benign, and underlying credit quality has normalized a bit. He does expect dispersion among managers, with the best institutionally validated firms that stuck to strategy doing well and weaker firms being exposed as the tide goes out. He explains that problems arose at retail-focused managers that had to deploy capital immediately from retail inflows. His firm, Churchill, is 94% institutional, which gives it more ability to manage flows and avoid that pressure.

  “Look, I think if you look at what's going on in our world, in the credit world and the lending world, that the reality is that things have been pretty benign. I mean, underlying credit quality, it's normalized a bit.”at 15:17

Shipping, trucking & logistics

• Vincent Tizzio · AXIS Capital

  Marine war insurance rates for Hormuz transits keep rising, with 30% a floor and cover still available.

  Vincent Tizzio says marine war underwriters price the vessel's value, and rates are escalating from the last conversation. A 30% increase is a minimum floor, and estimates range from 5% to 12% of vessel cost. Cover is still available, though selectively. He argues the risk is highly targeted: about 2,000 vessels crossed the Strait in Q3, roughly 45 to 50 were hit, and the US Navy cites a 97% success rate. Costs are being passed on to goods and services, and he says the most acute exposure is the few hours of transit, which is carefully underwritten.

  “There's two ways to price the risk... Most Marine war underwriters are pricing the vessel... Rates are escalating. 30% is a minimum floor.”at 21:04

AI: overall outlook

• Victoria Greene · G Squared Private Wealth

  AI is not dead; the broad tech sell-off is panic and the AI trade still holds.

  The sell-off hit Sandisk, Seagate, Nvidia and Broadcom, and she says 'everybody hates AI' and data centers. Her argument is that the trade is not dead because the companies are still posting strong revenue growth. Only Microsoft had really beaten expectations that week, but the growth trajectory still holds. She points to earlier panics such as DeepSeek and the SaaS Apocalypse as evidence that the market overreacts to AI scares.

  “I think you've really got to be looking at this and saying, look, AI is not dead. All of these companies are producing great revenue growth.”at 36:49

AI infrastructure & data centers

• Craig Johnson · Piper Sandler

  AI is not dead but is shifting from data center build-out to utilization.

  He agrees AI is not going away, but argues the investment cycle is moving from the build-out phase, where capex on data centers and chips drove the trade, to a phase centered on using AI. That change is part of why he expects technology to keep having more pain and downside. He treats the transition as a reason to reduce technology exposure, not as a reason to buy the build-out names.

  “I agree AI is not dead. It's not going away. It's just transitioning into now a utilization of it rather than a build out from my perspective.”at 38:20