Tech Is the Cheapest Since 2022. Walmart Is 46 Times Earnings. Where Is the Bubble?

Walmart

Across the shows: 19 speakers in other episodes, 17 bullish and 3 bearish calls, past 30 days

• Robert Hagstrom · EquityCompass Investment ManagementBearish

  Walmart at 46 times earnings is expensive for 5% growth and low returns on capital.

  Hagstrom compares Walmart with Nvidia on the same valuation basis. Walmart grows about 5% with much lower returns on capital and trades at 46 times forward earnings, while Nvidia trades at 19 times. He says Walmart, Procter & Gamble, Colgate and Costco are trading very rich relative to fundamental value because investors are buying low-volatility defensive stocks during the selloff.

  “Or I can go buy Walmart today that's growing at 5% with much lower returns on capital for 46 times [forward] earnings.”at 5:20

• Robert Hagstrom · EquityCompass Investment ManagementBearish

  Walmart is very expensive; its valuation looks too high for its reinvestment-driven cash flow profile.

  Hagstrom says he is pushing back on Walmart because he thinks it is very expensive. He cites Michael Mauboussin's point that Walmart had negative free cash flow in its first eighteen years yet earned above its cost of capital. Sam Walton kept reinvesting all the capital, so earnings and free cash flow yield looked weak. Hagstrom draws a parallel with the hyperscalers, which also reinvest heavily. His valuation concern rests on the gap between the high multiple and the reinvestment-driven cash flow.

  “I'm obviously beating on Walmart because I think it's very expensive. Walmart, the first eighteen years, had a negative free cash flow.”at 57:48

NVIDIA

Across the shows: 104 speakers in other episodes, 86 bullish and 58 bearish calls, past 30 days

• Robert Hagstrom · EquityCompass Investment ManagementBullish

  Nvidia trades at about 19 times earnings, cheap for 10% growth, and is not in a bubble.

  Hagstrom sets a conservative case: Nvidia's returns on capital are not above the cost of capital and revenue grows 10% a year for ten years. Even on those assumptions the stock trades at about 19 times earnings, which he contrasts with Walmart at 46 times. He adds that he does not think Nvidia is in a bubble, while the newer neocloud companies are the bubble candidates. His fund bought Nvidia in October, and AI's rise came faster than he expected, which he calls luck.

  “let's just say they're gonna grow revenues at 10% for the next ten years. That stock today is trading at 19 times.”at 4:58

GE Vernova

Across the shows: 8 speakers in other episodes, 12 bullish and 1 bearish calls, past 30 days

• Robert Hagstrom · EquityCompass Investment ManagementBullish

  GE Vernova benefits as data centers must build their own power plants, with 60% of revenue outside the US.

  Hagstrom argues that data centers cannot easily tap the grid without raising household electric bills, so they will need to bring their own power. He says this benefits companies like GE Vernova, which build power plants. He notes that about 60% of GE Vernova's revenue comes from outside the US, where it builds plants for data centers. He presents this as a path around the power constraint that holds back data center buildout.

  “So that just lays into things like GE Vernova, is you gotta build these things. You gotta build them with their own power pack. ... 60% of GE [Vernova]'s revenues are outside the... They're building power plants outside The United States for data centers.”at 1:04:03

US stock market outlook

• Robert Hagstrom · EquityCompass Investment Management

  Investors will rotate out of defensive staples into tech, and the reversal could come within a day.

  Hagstrom argues that the market has moved defensive: investors are buying stocks that have not fallen, which pushes staples and low-volatility funds such as SPLV to extremes. He says the set-up favours selling consumer staples and leaning into tech. He admits the timing is the million-dollar question, but he expects the change to be abrupt. Once it begins, he expects it to take less than 24 hours, catching investors who stayed defensive off guard.

  “it's gonna set up some very interesting dynamics where the reverse trade is you should be selling your consumer staples in here, and you should be leaning in to tech.”at 6:21

• Chris Mayer · Woodlock House Family Capital

  The current tech selloff is an ordinary drawdown, like Amazon's 90% fall in 2000.

  Mayer points to past bubbles and drawdowns, including Amazon falling more than 90% in 2000 and other corrections, to argue that the current tech selloff is not unusual. The drop feels awful while it happens, but he expects that in ten years it will look obvious. His view is that drawdowns of this kind are normal in the market cycle.

  “Mean Amazon went down 90 some percent in 2000. Think about that where it is now or, you know, one of any number of other corrections and drawdowns. So what we're seeing is not really anything unusual.”at 7:21

• Chris Mayer · Woodlock House Family Capital

  Stock prices should eventually follow free cash flow per share growth, despite volatility.

  Mayer describes free cash flow per share and returns as the North Star, and says stock prices should follow their growth over time. He admits that the current period makes him question his approach and his own mistakes. He still says the general approach holds, and that volatility along the way does not change the destination. The point connects the current selloff to Ben Graham's weighing-machine idea.

  “I still think the North Star is the stock prices actually follow that free cash flow per share and returns.”at 13:39

• Robert Hagstrom · EquityCompass Investment Management

  His firm's portfolio trades near 75 cents on intrinsic value, the cheapest it has ever been.

  Hagstrom says his firm weights the intrinsic value of its holdings and compares it with price. The portfolio was about 85 cents on the dollar during COVID, about 80 cents in September 2022, and now about 75 cents. Morningstar's estimate of intrinsic value in 2022 matched his own, which supports the method. He says the portfolio is cheaper now than ever, so he tells clients not to sell and to consider buying.

  “we're worth today about 75¢ on the dollar. We're cheaper today than we've ever been.”at 48:36

AI: overall outlook

• Chris Mayer · Woodlock House Family Capital

  AI will be used far more heavily in ten years, though the path there is uncertain.

  Mayer says he is highly confident AI in some form will still be around in ten years and used much more heavily than today. He adds that the trajectory from here is what makes the outcome hard to call. He uses this long-term backdrop to argue that short-term swings in AI stocks matter less, and that some very good businesses in the space will emerge.

  “I feel highly confident saying that AI technology in some form is going to be here in ten years from now, and it's going to be used much more heavily than it is now.”at 7:01

• Robert Hagstrom · EquityCompass Investment Management

  If AI is universally adopted, it is a big opportunity; a few winners should do extremely well over three to five years.

  Hagstrom sets out two scenarios. If AI is a scam or a bubble and is not used in five or ten years, he will be wrong and write cookbooks. If AI works and is adopted worldwide, it is a large opportunity. Not everyone will capture it, but a few companies will do extremely well over the next three to five years. He is betting on the adoption scenario.

  “if you believe that AI is gonna work and that it's going to be universally adopted by pretty much everybody around the world... Then you gotta go, boy, this is a big opportunity.”at 54:34

• Chris Mayer · Woodlock House Family Capital

  AI gains may accrue to businesses that use the tools more than to the infrastructure producers.

  Mayer draws on the automobile era, when many manufacturers existed and most failed, to argue that picking winners is hard. He thinks the obvious AI infrastructure plays may not be where the best returns lie. Businesses that use AI tools to improve themselves may benefit more, the way the internet improved some businesses that were not producers of it. He expects the opportunity to be broader than people now assume.

  “it may not be that the opportunity is the obvious stuff that we think of, like the AI infrastructure stuff per se. It might be some other business around that's going to vastly benefit from using the tools, you know, not necessarily be a producer of the thing.”at 55:25

• Robert Hagstrom · EquityCompass Investment Management

  AI companies have handled investor communication badly; they must show returns on capital to restore confidence.

  Hagstrom argues the AI industry has done the worst public relations job of any industry he has seen. Companies rushed to release frontier models and be first, so they forgot to explain why AI matters to owners of their businesses. He says they need to explain returns on capital and show that AI is profitable, not just claim they will have the most data centers. He calls the current narrative about as negative as anything in technology since October 2000, and says there is a big lack of faith in AI that companies must address.

  “I would play unabashedly that I think the AI industry has done the worst job in public relations I've seen of any industry that has.”at 56:37

AI infrastructure & data centers

• Robert Hagstrom · EquityCompass Investment Management

  Hyperscalers already earn above cost of capital but reinvest so heavily that free cash flow turns negative.

  Hagstrom compares hyperscalers such as Amazon and Google to Walmart's early years. He says they already earn above their cost of capital on their investments, but they put every dollar back into the business as fast as they can. As a result their cash flow yields turn negative, the same pattern Walmart showed. He therefore thinks the market is misreading the AI spending. The companies need to show that the spending is sustainable and profitable, which he believes they have not yet done.

  “if you look at the hyperscalers, particularly Amazon, Google, they're already earning above the cost of capital... But they basically are just throwing it back in there as fast as they can, so their cash flow yields turn negative just like Walmart did.”at 58:08

• Chris Mayer · Woodlock House Family Capital

  Data centers are a hard sell to communities, though the buildout will likely proceed anyway.

  Mayer says data centers sit there, use power and water, make noise, and employ few people. Earlier, companies like Amazon brought visible jobs, so communities could see some benefit. He thinks this makes data centers a tough sell to local residents. He still expects the industry to work it out, since the technology is out of the bottle and will not be reversed.

  “They just sit there. They make [noise]... And there's no people... So I think it's a tough sell in that way, but I think they'll figure it out.”at 1:04:49

• Robert Hagstrom · EquityCompass Investment Management

  Data centers bring large tax and contractor economic benefits to host states and municipalities.

  Hagstrom says a data center may employ only about 20 people, but building one brings large inflows to local governments and state treasuries. Contractors and ancillary jobs add further economic activity. He argues this produces a positive economic return for municipalities and states, calling it a fact. He concedes the real question is how much is too much and how the compute cost is spread out.

  “I would argue that there actually is a lot of economic inputs to the state and local municipalities that come from a data center going up. ... a positive economic return for municipalities and states. Fact. It's a fact.”at 1:05:12

• Robert Hagstrom · EquityCompass Investment Management

  Compute power, not GPUs, is the AI bottleneck for years; supply-demand imbalance drives excess returns.

  Hagstrom says many data centers are delayed or unfinished, and the main reason is that they cannot get power. He says GPUs can be made and memory will be solved, so the bottleneck will be compute and how to power these facilities. He calls this a supply-demand imbalance, and says that is where excess returns are made. He expects the bottleneck to last for many years.

  “If you really wanna think where the bottleneck is in AI, it's compute. It's not GPUs. ... The big bottleneck for many, many years will be compute. How do we power these things?”at 1:05:52

Market sentiment, positioning & volatility

• Robert Hagstrom · EquityCompass Investment Management

  When investors trade the same way, the market loses diversity and volatility rises.

  Hagstrom says the market is no longer as diversified as it was, with fundamental, quantitative and pod-shop investors in very different sizes. He cites hedge fund assets of about $5 trillion, pod shops trading heavily each day, and notional options value above daily stock trading. When the market loses diversity and moves to one side, volatility increases. He says the structure explains why the market behaves this way, even if it is still hard to live through.

  “the market loses its diversity, and when it loses its diversity and goes to one side, the volatility goes up.”at 2:54

• Robert Hagstrom · EquityCompass Investment Management

  The AI-stock selloff will reverse quickly and sharply as shorts cover and momentum buyers return.

  Hagstrom says AI and electrification stocks have broken through their 50-, 100- and 200-day moving averages, so there is no technical support. He expects selling exhaustion, then short sellers worried about a squeeze to cover, which improves sentiment, changes the narrative and brings new buyers. He says speed and magnitude have risen, so the way down has been fast and hard and the rebound will likely be just as fast and powerful. He also notes that a recovery could take about six months to pay off, though he expects momentum traders to move quickly.

  “I strongly believe that the bounce back will be equally as fast and equally as powerful on the way back.”at 50:50

• Chris Mayer · Woodlock House Family Capital

  Short-term stock prices float on confidence and narrative and can detach from financials for long periods.

  Mayer says stock prices, at least in the short term, move a lot on confidence and narrative. This can push a price far from what the financials support for quite a while. He ties this to the current selloff, where confidence and story have driven the moves. He treats this as a reason not to react to price moves alone.

  “the stock prices are at least in short term seem to float a lot on confidence and narrative... So you can get these situations where the price has really become unwarrred financials for quite a while.”at 59:56

AI bubble & tech valuations

• Chris Mayer · Woodlock House Family Capital

  AI infrastructure stocks may be the next turn after a 35% monthly drop, with buyable names.

  Mayer says AI infrastructure and semiconductor stocks fell about 35% in a month. He compares this with software stocks last year, which were thrown out with the bathwater, cut back, and are now green. He suggests the infrastructure names may be next, and he thinks some of them must be worth buying. The argument relies on a past example rather than on fundamentals.

  “now maybe this is the turn for the infrastructure stock. So certainly there's there's gotta be something in there that's that's worth it.”at 4:20

• Robert Hagstrom · EquityCompass Investment Management

  AI infrastructure stocks are as cheap as they have been since 2022.

  Hagstrom reviews the AI infrastructure names he follows and finds them as cheap as he has seen them. He says the only comparable level is 2022. He ties the view to fundamentals: he wants sales and earnings to frame the value, even if the growth or return assumptions are lowered. He notes that this is a valuation judgment, not a call on timing.

  “What I see in that space that I'm looking at is these things are as cheap as I've ever... You'd have to go back to to 2022 to start it.”at 7:42

• Robert Hagstrom · EquityCompass Investment Management

  Neocloud AI infrastructure companies without earnings are the bubble candidates among AI stocks.

  Hagstrom separates the established AI infrastructure players (AWS, Google, Azure) from newer neocloud operators. The neoclouds raised capital in public markets to build new data centers outside the big clouds and have no earnings yet. He says that if the story turns, they will be hit hardest, so they are the bubble candidates. He does not see Amazon, Google or Nvidia as bubbles.

  “The new guys, if they're gonna get their head handed to you because they don't have any earnings yet. So those are the ones that are more bubbleicious in my mind.”at 8:36

Earnings & corporate profits

• Chris Mayer · Woodlock House Family Capital

  Mega-cap companies rarely benefit from buybacks at drawdown prices versus reinvestment returns.

  Mayer says that once a company reaches a trillion-dollar market cap, the normal checks on undervaluation, such as private buyers or small-cap buybacks, become theoretical. A mega-cap could in theory buy back a lot of stock during a drawdown. He thinks this is probably not the best use of capital, because the incremental returns inside the business are higher, even at the lower share price.

  “Unless, of course, they can always decide to buy back a bunch of stock, but maybe that's probably not the best use even at this drawdown compared to the incremental returns they get in their businesses.”at 16:44

• Chris Mayer · Woodlock House Family Capital

  Buybacks during deep drawdowns probably work out well, but only where buying stock is the best use of cash.

  Mayer says some companies with big drawdowns are using excess cash to buy back stock, and he expects that to work out well for them. He stresses that this is a company-by-company question. Buybacks are not right for every company, even one whose shares have fallen, because its highest incremental return may come from something else with that cash.

  “some companies are big drawdowns and they're taking some excess cash to buy back stock. That's probably gonna work out really well. But again, it's a company by company story”at 1:00:22

Private equity

• Robert Hagstrom · EquityCompass Investment Management

  Private equity should take private cheap, high-return businesses but is not doing so.

  Hagstrom argues that some companies are so cheap that a buyer could borrow at about 7% and earn 15% to 20% returns on capital. He says private equity has plenty of capital and should buy businesses of $50 billion to $100 billion, and that management teams should consider buying back or taking their companies private. He finds it puzzling that this is not happening, which he sees as the mechanism that should correct the mispricing.

  “by having management go out and say, well, I can borrow at 7%, and, you know, I'm... My returns on capital are 15 to 20%. I can make a go of this, take it private, but you just don't see that.”at 15:48