Nvidia

As seen by Bulls, Bears, and Millennials

Before we begin today’s essay, we here at Mytwocent$ wish to apologize to our readers. The demographics of our subscriber base indicates there are very few millennials. As such, we recommend you get your hands on a “Millennial to English” dictionary, just in case any translation is required in today’s blog. With that out of the way ….

This past Saturday, Bag spent the whole day (14hrs) playing in a poker tournament. In tournament play, seating is random. Drawing a seat next to someone you don’t like can make for a long day. The opposite is true as well. Drawing a seat around people you find interesting makes it an ideal setting for a nuanced conversation between multiple parties, thanks to the downtime between hands. Bag was blessed to have drawn a seat at a table with a fellow greybeard (Mike), whom he has known for decades. Mike, like Bag, always enjoys talking stocks. Sandwiched between us were two 20-something millennials with the same first name, although one spelled it Bryan, and the other Brian.

Bag starts: “Mike, you make any moves in the markets lately?”

Mike Responds: “Yeah, you are not going to believe it - but I bought some Nvidia.”

Bag chuckled and said: “You are right, I don’t believe it. How does a guy who only buys value stocks rationalize paying 40x earnings?”

Bryan chimed in: “There is uber-value in Nvidia, even at a 40 multiple.”

Bag said to Bryan: ”You do realize that paying 40x earnings means every dollar invested returns only 2.5 cents … where exactly is the value in that?”

Bryan shot back: “Value is relative. It’s a vibe, bro.”

As with most millennials, Bag didn’t know where to begin to communicate with Bryan, so he turned to Mike and said: What say you, Mike?”

Mike said: “Look at the growth rate in their revenues and income, and then tell me it’s not worth paying 40x.”

Bag responded: “Yeah, they are growing like a weed, no doubt. But paying 40x is only justifiable if you expect that growth to continue. What makes you believe that’s the case?”

Mike replied: “They have huge margins, a near monopoly on the chips they produce, and a backlog of orders - I think the growth will continue.”

Bag replied: “I’ve seen this movie before. Those are the EXACT same arguments Wall Street was making for Intel & Cisco in the late 1990s.”

Bryan jumped in: “Okay, Boomer …. stop being so extra … it’s cringe.”

Mike shot back: “The 1990s were a bubble. Today it’s still early in AI buildout, ain’t no bubble here.”

Bag responded: ”Agreed, there is no bubble in AI - yet …. But there is one in Capex spending, and Nvidia has been the primary beneficiary. For the last 4-5 years the hyper-scalers (Microsoft, Google, Meta, etc) are spending like the Kardashians planning a wedding. Did you know MSFT expects to spend 120 billion in Capex this year, which is up 6-fold since 2022? During that same time period, MSFT revenues are up only 67%. Think about that for a sec, revenues up 67%, Capex spending up 600% - does that sound sustainable to you?”

Mike replied: “MSFT is not Nvidia’s only customer; they have plenty of others.”

Just then, the Poker tournament went on a 15-minute break. As Bag always does, he went for a walk during break to stretch his legs, and worked Grok the whole time to come up with the numbers below. When the break was over, the conversation continued.

Bag said: “Ok, so MSFT is not their only customer, but the story is similar at both Google & Meta. In 2022, Google reported its Capex at 25B; Meta’s was 19B. They are guiding Capex in 2026 to be 175B and 125B, respectively. In the same time frame, their gross revenues are up 67% for Meta, and 60% for Google. Just to be clear, revenues up 60-67%, spending up 600% - 700%, …. I ask again, how is that sustainable?

Mike replied: “Like I said, they have many other customers”

Bag said: “The hyper-scalers are the biggest players in the industry, and therefore, are an excellent proxy as to what the industry as a whole is doing. It’s no coincidence that Nvidia’s revenues are up 7.5x in the last four years, which is remarkably close to tracking the 6x-7x increase in spending by the hyper-scalers. What do you suppose happens to Nvidia’s revenues, income, and ultimately stock price when the hyper-scalers turn off the Capex spigot?”

Mike said: “It’s still not clear to me there is a direct connection between the growth in Capex spending, and Nvidia’s revenues. Even if there was a connection, those companies you mentioned have the strongest balance sheets in the S&P. If they need or want to continue to spend, they can, and they will.”

Bryan jumped in: “Those companies have monster bags yo.”

Bag said: “What?”

Bryan replied: “Their Gucci, they know how to make it rain, bro.”

Bag replied: “If I understand what you are saying, Bryan - those are indeed some of the most profitable companies on the planet. However, they are borrowing tens of billions because they don’t earn enough to cover the rate at which they are spending. Meta now has 60B in debt on their balance sheet, most of it added in the last year. Google is very similar; they now have 50B in debt, again, most of it added in the last year.”

In a somewhat snide manner, Bryan said: “Sounds like girl-math to me.”

Bag responded:”I’m not sure what girl-math is - but i do know what real math is. And the real math says, not only can the hyperscalers not grow their Capex from here, they can’t even maintain these levels without wrecking their balance sheets by adding debt.”

Mike said:”I’m not so sure Nvidia’s future is tied to the Capex spending of the big boys, it might have an effect - but it won’t crater the stock.”

Bag replied: “Sounds optimistic, no?”

Bryan added: “Ain’t no hopium here.”

Bag asked: “Do you know who the largest retailer of computers was in the late 90’s?”

Bryan snapped: “Little throwback Thursday … we talking about the Stone Age again?”

Mike said: “Not sure, was it Hewlett-Packard?”

Bag replied:”No, I think they were printers. It was Dell, and there wasn’t really a close second. From 96-99, their Capex quadrupled. When you think about corporations & Capex spending in the late 90’s, the biggest beneficiaries had to be Intel & Cisco. Everyone wanted Intel chips, because they were the best; and they also wanted the computers to communicate with one another, so they needed the cards, ports, and routers Cisco was selling. I’m not saying Dell was responsible for all the revenue growth at Intel & Cisco, but they were the biggest player at the time, and thus a good proxy for what the industry as a whole was doing. In the same time period Dell’s capex increased 400%, Cisco’s Revenues went from 4B to 18B - an eerily similar increase of 450%.”

Mike responded: “Let me guess, Dell quit spending in early 2000 - and so the tech bubble popped?”

Bag said: “Actually, they spent like drunken sailors all through 2000.”

Mike replied: “Sounds like there is no connection then.”

Bryan looked at Bag glaringly and said: “Yeah, bro … where’s the receipts?”

Bag replied: “What Dell did was announce in the spring of 2000 that Capex cuts would be coming in 2001. That was all it took to crush the price of the Intels & Ciscos. They were both 70-80% off within a year. If the hyper-scalers announce today they will turn off the spigot in the near future, isn’t it reasonable to assume Nvidia may pay the price?”

Mike replied: “I don’t think that announcement will be coming anytime soon. As we agreed, the AI buildout is in the early stages.”

Bryan looked at Mike and said: “Word.”

Mike added: “There is some merit to what you are saying, I just think because the AI buildout will continue, the reckoning is years away. Until then, I’m going to enjoy the ride.”

Bryan nodded in agreement and said: “Slay.”

Bag then looked over at Brian with an “I” and said: “You have been quiet all day, surely you have an opinion on this subject.”

Brian said: ”F*ck those haters on Wall Street - buy Bitcoin.”

Bag was blessed to appear on the Rigatoni Capital Podcast, discussing in more detail the outlook for Nvidia as well as Bitcoin & Gold. For those interested - Link is here!