CNBC Fast Money · · 48 min

Stocks Drop After Post-Election Rally… And The Future Of The AI Trade 11/5/26

With Brian Kelly, Jim Paulsen, Oliver Renick, Stephen Yalof

Summary

CNBC Fast Money discussed a sharp post-election stock sell-off and whether the AI trade is breaking down. Jim Paulson argued tech's long bull run is overextended and new-era tech could fall about 25% from its highs, while the broader economy is weak outside AI spending, with real wages declining and growth the real problem. Courtney Garcia and Brian Kelly saw the build-out as delayed rather than cancelled, with Kelly favouring AI users like Accenture and Palantir over sellers of AI, Circle on tokenized finance, and a possible government backstop for the build-out. Tim Seymour liked Altria, gold miners, and Alibaba, while Karen Finerman bought Nvidia and Bank of America on the dip and was comfortable buying more if volatility spiked.

Key takeaways

  • Brian Kelly (bullish on Accenture): Accenture is a buyer's pick: an AI user whose work gets better as intelligence gets cheaper.
  • Courtney Garcia on AI infrastructure & data centers: Some of next year's data center capacity may slip, but AI demand is delayed, not gone.
  • Jim Paulsen on AI bubble & tech valuations: Tech bull market is complacent and overlong; tech likely falls about 25% and goes out of favor in 2027.
  • Tim Seymour (bullish on Altria Group): Altria is a defensive holding with strong recent numbers and a free cash flow story from past investments.
  • Karen Finerman (bullish on NVIDIA): Bought Nvidia on the sell-off; its balance sheet and valuation made it easy to buy.

What was said, by company and theme

33 views from 6 speakers. Every quote is checked against the episode's transcript (how).

Circle Internet Group

Across the shows: 2 speakers in other episodes, 1 bullish and 0 bearish calls, past 30 days

  • Brian Kelly · BKCM LLCBullish

    Finance moving onto blockchain rails over three to five years makes stablecoin players like Circle the trade.

    Brian's three-to-five-year thesis is that finance is going to come on chain and replumb the entire financial system, probably in a regulated way rather than on Bitcoin rails. Companies like Circle will provide the substrate, which is stablecoins. He sees this replumbing as the trade to own, with crypto-adjacent companies as the vehicle.

    “the reason to own crypto now and crypto adjacent and some of the companies is finance is going to come on chain... there are companies like a circle or something like that that are going to have the substrate, which is, the stable coins”at 30:00
  • Brian Kelly · BKCM LLCBullish

    Circle is building an ecosystem for the agentic economy, making it a massive trade worth accumulating near current levels.

    Beyond the replumbing of the financial system, Circle is building an ecosystem for the agentic economy, where AI agents transact. Kelly runs an operation filled with AI agents (about eight, with sub-agents planned), so he sees the agentic economy as a real trend. He calls it a massive trade and has it on his buy list, wants to see the stock base first, and is not buying tomorrow.

    “Circle is building out an ecosystem for the agentic economy. And for a guy who runs an operation filled with AI agents, I think this is just going be a massive trade.”at 36:42

Accenture

  • Brian Kelly · BKCM LLCBullish

    Accenture is a buyer's pick: an AI user whose work gets better as intelligence gets cheaper.

    Brian wants to be a buyer of intelligence rather than a seller of it. Companies that sell AI are what fell today, but companies like Accenture are using intelligence, which is getting cheaper, and transforming their businesses. He would look at those names for buying after the sell-off.

    “if I can use it so I can be somebody like an Accenture... These are all people that are using the intelligence. It's getting cheaper and they're transforming companies.”at 5:03

Palantir Technologies

Across the shows: 14 speakers in other episodes, 16 bullish and 2 bearish calls, past 30 days

  • Brian Kelly · BKCM LLCBullish

    Palantir, down sharply today, is still a stock he likes as an AI user.

    Palantir fell quite a bit today, but Brian still likes it. He puts it in the same group as Accenture: companies that are using intelligence, which is getting cheaper, and transforming their businesses, rather than companies selling the intelligence that sold off.

    “I know Palantir was down quite a bit today, but even a Palantir is something that I like. These are all people that are using the intelligence.”at 5:03

Alibaba Group Holding

Across the shows: 6 speakers in other episodes, 5 bullish and 1 bearish calls, past 30 days

  • Brian Kelly · BKCM LLCBullish

    Alibaba is a great trade: cheap valuation, a clear level to trade from, and China's no-new-stock stance.

    Brian sees Alibaba as a great trade because there is a clear level to measure against, such as last week's low or the June low. Its business model differs from US players: it gives away its model for free and charges customers to rent server space. Valuations are low, and China has said it won't issue more stock and wants its market to go up, which he calls a macro tailwind.

    “No, I think it's a great trade here in Alibaba because you've got something to shoot against... They're giving away the model for free and you've to rent their server space.”at 26:36

Bitcoin

Across the shows: 28 speakers in other episodes, 20 bullish and 13 bearish calls, past 30 days

  • Brian Kelly · BKCM LLCBullish

    Bitcoin's drop just tracked the Nasdaq, not new sellers; crypto held up well in the downturn.

    Bitcoin's beta to the Nasdaq is about one, so the roughly 5% drop matched the Nasdaq's 5% fall. He reads that as market-related selling, not new sellers or concern about what Democrats will do. He is encouraged by how crypto and crypto stocks traded broadly in this downturn.

    “Bitcoin today, it's beta the Nasdaq's about one to one, right? So down 5%, Nasdaq down 5%... this was just market related, not new sellers coming in”at 30:00

Altria Group

  • Tim Seymour · Seymour Asset ManagementBullish

    Altria is a defensive holding with strong recent numbers and a free cash flow story from past investments.

    Tim has owned Altria for years. He says it just had really good numbers, and there is a free cash flow story supported by investments the company made years ago. He cites it as a defensive pick in a selloff.

    “a name like Altria is a name I've owned for years. And it's a name that just had really good numbers. And there's actually a free cash flow story there”at 33:49

NVIDIA

Across the shows: 74 speakers in other episodes, 69 bullish and 28 bearish calls, past 30 days

  • Karen Finerman · Metropolitan Capital AdvisorsBullish

    Bought Nvidia on the sell-off; its balance sheet and valuation made it easy to buy.

    Karen bought Nvidia after today's drop, saying the balance sheet and the valuation made it easy to buy. She expects she will have a chance to buy more at a better price, and suggests other names might bounce back more.

    “I bought some Nvidia... I think, you know, other ones might bounce back more, but I mean, the balance sheet here, the valuation you talked about. So, you know, that to me, it was easy to buy”at 34:55

Bank of America

Across the shows: 8 speakers in other episodes, 2 bullish and 7 bearish calls, past 30 days

  • Karen Finerman · Metropolitan Capital AdvisorsBullish

    Bought Bank of America at 11 times earnings; less exposed than JPMorgan, Morgan Stanley and Goldman.

    Karen bought a little Bank of America, which trades at about 11 times earnings. She thinks it has less exposure than JPMorgan, Morgan Stanley and Goldman Sachs, which is why she bought it amid the bank sell-off tied to fears about the deal and trading calendar.

    “I also bought a little Bank of America... it's 11 times earnings now for Bank of America. And that, I think, has less exposure than a JPMorgan, Morgan Stanley, and a Goldman Sachs.”at 35:15

AI infrastructure & data centers

  • Courtney Garcia · Payne Capital Management

    Some of next year's data center capacity may slip, but AI demand is delayed, not gone.

    Courtney argues the issue is timing. Expectations for capacity coming online next year look too high, with about 30% possibly questionable, but these projects are not being cancelled. If AI demand is being pushed down the line, which she thinks most people believe, there will be buying opportunities, and valuations are becoming attractive.

    “Is AI demand just gone? Or is this getting pushed down the line? And if you believe it's getting pushed down the line, which I think most people do, I think there's going be buying opportunities here.”at 7:47
  • Brian Kelly · BKCM LLC

    Demand for intelligence compute will stay robust, so installed compute stays valuable despite the sell-off.

    Brian says hyperscalers rallied yesterday because installed compute would become more valuable, and he still sees that as a trade. Chips are piling up because power and cooling can't be obtained, and some capacity won't come online, but he thinks demand for intelligence compute is going to be robust. He separates the market situation from demand for the underlying product.

    “I still think that's a trade... I think the demand for intelligence compute is going to be robust.”at 8:17
  • Karen Finerman · Metropolitan Capital Advisors

    Would pay a higher multiple for hyperscalers with balance sheets that can survive; she holds and wants more.

    Karen would rather pay up at a higher multiple for a balance sheet that can survive the build-out than own cheaper companies that can't. She has had exposure to hyperscalers and wants to continue with that. She expects the street to view lower capex positively, though she doesn't think that is better.

    “I would rather pay up at a higher multiple but have a balance sheet that can survive. So hyperscalers. I definitely have had exposure to that. I want to continue with that.”at 9:27
  • Brian Kelly · BKCM LLC

    Government likely to backstop the AI build-out within about six months, since it is a national security priority.

    Brian argues the AI build-out is too important to national security, especially against China, for it to fail. He expects some form of bailout if the sell-off gets really bad, such as the US government buying AI debt or taking shares, as companies already do with individuals. He frames it as a possible trade that could play out within about six months.

    “We cannot get behind China... you cannot get behind China. So why wouldn't you give a bailout? ... The US government perhaps maybe buys AI debt. You get shares”at 13:07
  • Courtney Garcia · Payne Capital Management

    Highly leveraged neoclouds are most vulnerable, since delayed data centers jeopardize the debt they rely on.

    Courtney says the build-out is too big to fail, so it will happen, but the timing matters. Companies with high leverage, such as neoclouds, are more susceptible because they depend on data centers getting built to finance their debt. She thinks that is what investors should be more concerned about than the overall build-out.

    “So the companies who have high leverage on their balance sheet so think of your Neo Clouds, those are the ones who are much more susceptible. Because if their data centers aren't going to get built in the near term, they're dependent on that in order to finance their debt.”at 13:38
  • Karen Finerman · Metropolitan Capital Advisors

    Data center approvals have become a political bargaining chip, with states now asking what they must do to win projects.

    Karen says data centers went from getting giant tax breaks to politicians now asking what the builders have to do to be allowed in. She calls it the new deal, and says it creates a win for politicians who can say they fought for their state. She expects this to happen.

    “you get a giant tax break... now, all right, what do we have to do to get you to allow us? And that creates a win for politicians... This is the new deal. I think that will happen.”at 14:31
  • Brian Kelly · BKCM LLC

    Building a data center in 2027 will cost more than today, with more community spending.

    Brian says data centers are going to be more expensive: more has to go to communities, and there is more demand for green power, so it will cost more to build a data center in 2027 than today. He takes the positive view that more spending means more GDP and more economic growth.

    “it's going to cost more to build a data center in 'twenty seven than it is today. But if you want to take the positive of that, more spend is more GDP is more economic growth.”at 15:00
  • Courtney Garcia · Payne Capital Management

    Hyperscaler capex will come down from the roughly $1 trillion expected next year, but won't go away.

    Courtney is still under the impression that there will be a pullback in capex. The five major hyperscalers were expected to spend about $1 trillion next year, which she thinks likely comes down, but she doesn't think it goes away. She sees the capex cut as a pullback rather than the end of the build-out.

    “I still under the impression here that we do see a pullback on CapEx, but I don't think it's going away... the five major hyperscalers were expected to spend $1,000,000,000,000 next year in CapEx. That comes down likely, but I don't think it goes away.”at 35:45

US stock market outlook

  • Tim Seymour · Seymour Asset Management

    S&P at 17.5x 2027 earnings; market looks interesting ex-Mag Seven, with the rest of the market more insulated.

    Tim argues the key question is valuation: the S&P trades at about 17.5 times forward 2027 earnings. Remove the Magnificent Seven and a couple of other names and the market looks quite interesting, with the rest of the market more insulated from AI fears. He thinks the index can still go lower, but below the surface many stocks have traded as if in recession for six to nine months, which he sees as opportunities.

    “the S and P that's now trading, you know, 17.5 times forward 2027 full year... And again, you could remove the MAG seven and the market looks really quite interesting as we go even to the rest of the market”at 3:12
  • Jim Paulsen · Paulsen Perspectives (Substack)

    S&P probably avoids a bear market but may take a 15% correction before this is done.

    Jim thinks it is early in the sell-off, with buy-the-dip rallies along the way but a continued downward trend. He expects tech to trade down about 25% from its highs. The S&P overall avoids a bear market but may have a 15% correction before it is done.

    “I think the S and P overall avoids a bear market, market, but maybe it has a 15% correction before we're done.”at 18:33
  • Karen Finerman · Metropolitan Capital Advisors

    A VIX spike around 32 is a buying opportunity, so she will add to positions.

    Finerman expects a spike in the VIX and says it will be painful but is time to buy more. Melissa Lee notes that 28 is still low and that panic usually comes around 40. Finerman says 32 is a good place to buy, and that she will buy and keep some dry powder for next week. She says she will go into the weekend long, as she always does.

    “A spike in the VIX. That will be painful, but time to buy more.”at 45:05

US rates & the Fed

  • Brian Kelly · BKCM LLC

    Fed unlikely to hike through 2027 if data center spending slows; falling gold and dollar signal lower rates.

    Gold up and the dollar down signal that the market expects lower rates and that the Fed may be done. If data center spending starts to slow, the Fed will not be on a hiking cycle and will not hike two or three times through 2027. Brian says that is what the market is reflecting.

    “Maybe the Fed is done here. They're certainly not going to be on a hiking cycle if data center spend starts to slow down. They're not going to be, you know, hiking two, three times through '27.”at 12:28
  • Jim Paulsen · Paulsen Perspectives (Substack)

    Fed's next move is probably easing, not tightening, with easing through most of next year.

    Jim thinks the Fed's next move is probably to ease rather than tighten, though it depends on tomorrow's jobs report and whether the sell-off calms before the next meeting. A tightening move would be the wrong one because policy has three-to-four-month lags that suggest a pretty big economic slowdown. He expects easing through most of next year, which the old-economy parts of the market need.

    “I think the Fed's next move is probably to ease rather than tighten... I think the Fed will be easing through most of next year.”at 19:02

US Treasuries & yields

  • Tim Seymour · Seymour Asset Management

    10-year yield falling to 4.85% is good for equities, but rates may not stay there.

    The 10-year has fallen to about 4.85%, which Tim says is very good for the equity market, though he isn't sure it holds. He thinks part of the rally is flight to quality rather than good reasons. Rates peaked on October 2, and he argues the 10-year at 5.35% was overdone and not clearly deserved, with less cyclicality in the 10-year.

    “the fact that interest rates have come all the way down to $4.85 on the ten year is very good for the equity market. And I'm not sure we stay there.”at 11:27

AI bubble & tech valuations

  • Jim Paulsen · Paulsen Perspectives (Substack)

    Tech bull market is complacent and overlong; tech likely falls about 25% and goes out of favor in 2027.

    Jim argues the main problem is complacency. The tech bull market, arguably 15 to 18 years long, is double the length of previous tech bull markets back to 1926, and investors have stayed overweight innovative parts of the market. Unlike the dot-com bust driven by irrational exuberance, this one will take longer to wash out. He expects new-era tech to trade down about 25% from its highs and go out of favor in 2027, as policy accommodation lets an old-economy rally run most of that year.

    “This tech bull market on a secular basis... you could argue it's been fifteen to eighteen years long. That's double the length of previous tech bull markets going back to 1926.”at 16:31

US growth & recession risk

  • Jim Paulsen · Paulsen Perspectives (Substack)

    The real problem is growth, not inflation: 86% of the economy grows about 1% with real wages falling.

    Jim says the economy is split: 14% of real GDP is a new-era booming part, while 86% has barely grown 1% in real annualized terms over the last two years. Higher rates, higher oil prices, higher inflation, sluggish real money growth, a flattening 2s10s curve, a relatively strong dollar and declining real wages mean the real problem is growth, not inflation. If AI spending slows, the rest of the economy is already slow, so recession fears follow.

    “we've got one small piece of the economy, 14% new era of real GDP that's booming. And we've got 86% of the economy that's barely been growing 1% in real terms annualized over the last two years.”at 17:29

AI: overall outlook

  • Courtney Garcia · Payne Capital Management

    AI trade isn't over, though investors should own the rest of the market for insulation.

    Courtney argues the AI trade is not over, but investors should own the rest of the market because a downturn like this needs insulation. She notes the S&P is only down about 15% even though a 25% to 30% downside is possible for tech, which indicates other areas will hold up better or be positive. She is not willing to call the build-out done, since markets die in euphoria, not skepticism.

    “this is exactly why I don't think the AI trade is over, but you absolutely want to own the rest of the markets because if there's a downturn like this, you need to be insulated from that.”at 21:35

International equities

  • Tim Seymour · Seymour Asset Management

    Can stay in China tech; rest of world still outperforms as US tech sells off.

    Tim says China tech has been a terrible trade, underperforming the S&P by about 54% since last October, but Alibaba has outperformed by about 15% over the last four or five sessions. If the US is tapping the brakes on AI spending, that leaves an opportunity for China's largest tech companies, which have been self-funding capex and taking a less capex-heavy approach to open-ended models. He thinks you can stay in the trade, though in the short run only.

    “So I think you can stay in the trade. I mean, do think that the rest of the world still does outperform here because there are a lot of things that are U. S. Centric”at 25:36

Stablecoins & crypto policy

  • Brian Kelly · BKCM LLC

    Clarity Act's collapse won't flatten crypto; Democrats may embrace it and banks are warming to stablecoins.

    Even with the Clarity Act effectively dead, Brian doesn't think crypto becomes a flat trade. He argues Democrats have an incentive to embrace crypto, since some in the party believe they lost because of it. Banks, including Jamie Dimon, are talking about stablecoins as useful, and crypto is being adopted, which is a headwind for now.

    “the Democrats have an incentive to embrace crypto because they lost because of crypto... you're already seeing banks, even a Jamie Dimon... talked about stablecoins being something that might actually be useful”at 31:02

AI chips & semiconductors

  • Courtney Garcia · Payne Capital Management

    Semis are selling off faster than earnings estimates, making them worth a look, though not unattractive.

    Courtney notes semiconductor prices are falling faster than their earnings expectations, so at this discount they are probably worth taking a look at. She is not piling in, and earnings could come down after this, but she wouldn't say the group is unattractive.

    “the prices are coming down faster than their earnings expectations... at this sort of discount, I mean, yes, it's probably worth at least taking a look at here. I'm not piling in by any means, but I wouldn't say they're unattractive.”at 33:28

Gold & precious metals

  • Tim Seymour · Seymour Asset Management

    Keeps liking gold; it was oversold after yesterday's rally and should keep working.

    Tim continues to like gold. GDX had a great couple of days, and he sold some upside calls there because yesterday's momentum was so extraordinary. He thinks gold is going to keep working because it was oversold, and he doesn't think you have to buy only gold and staples.

    “Clearly, I continue to like gold. I think the GDX has had a great couple of days... gold is going to continue to work and it was oversold.”at 34:20

Manufacturing & industrial policy

  • Courtney Garcia · Payne Capital Management

    Industrials like Caterpillar, Deere and GE Vernova are a buy on the dip; infrastructure spending doesn't depend on data centers.

    Industrials fell about 4% on news of a possible pullback in AI capex, with holdings including Caterpillar, Deere and GE Vernova. Garcia argues the pullback in data centers is not the whole story, because infrastructure build-out needs to happen regardless of data centers. She sees the selloff as an opportunity to nibble and to keep owning the group, and repeats in her final trade that it should be owned for the long run.

    “I actually think that's something you want to nibble on here because whether it's a pushback in data centers or infrastructure in general needs to happen regardless of data centers, I think that's a play you want to continue to be in here”at 36:14

Consumer & retail

  • Stephen Yalof · Tanger Inc.

    Retail weakness is a brand problem, not a category problem; newness-driven brands like Michael Kors are recovering.

    Asked which category is struggling, Yalof says accessories were rough last quarter but the picture has changed. Coach has been bucking the trend, and Michael Kors and Vera Bradley had a tough year but are now bringing in newness and posting positive comps. He concludes that weakness is more a brand story than a category story, since brands that connect with consumers with new product are winning.

    “I'd say it's really more of a brand story than it is a category story.”at 40:38

US jobs & labor market

  • Courtney Garcia · Payne Capital Management

    Tomorrow's jobs report must be soft enough to keep the Fed supportive, but not so soft it fuels recession fears.

    After a sharp market drop that has people talking about recession, Garcia says the jobs report is important. She wants the data soft enough to keep the Fed supportive, which means easing. She also wants it not so soft that it deepens the recession fears already surfacing. She also wants to see the VIX level out.

    “I think that jobs report is going to be important. You want to make sure that it is soft enough to keep the Fed supportive, but not so soft that it starts to exacerbate some of the recession fears you're starting to see people talk about after a day like today.”at 44:26

Across the shows

What other episodes Chorus covered said about the same companies in the 30 days up to this one.

Topics

  • 10-year yield
  • AI adoption by enterprises
  • AI compute demand
  • AI trade durability
  • Bitcoin price action
  • China tech stocks
  • Fed rate path
  • S&P 500 correction
  • agentic economy payments
  • brand vs category performance
  • chip stock buying
  • crypto market-structure bill
  • data center build cost
  • data center capex timing
  • data center local politics
  • defensive staples
  • gold outlook
  • government support for AI build-out
  • growth weakness and real wages
  • holiday retail sales outlook
  • hyperscaler balance sheets
  • hyperscaler capex
  • index valuation
  • industrials infrastructure spending
  • jobs report and Fed support
  • large bank valuations
  • neocloud leverage risk
  • rate hikes
  • semiconductor valuations
  • stablecoins and tokenized finance
  • tech bull market complacency
  • value-seeking consumer demand
  • volatility-driven buying

More from CNBC Fast Money

All CNBC Fast Money episodes

Ask your AI about this episode

Chorus gives Claude, ChatGPT and other assistants every view from this episode and every other show Chorus follows. Ask who agrees, who changed their mind, and what else was said about the same companies.

Setup guide

A weekly email with the calls that mattered from the most important voices in investing. Unsubscribe any time. Privacy Policy

Chorus reports what speakers said. It isn't investment advice. How Chorus works · About Chorus