The AI Daily Brief: Artificial Intelligence News and Analysis - Nvidia Becomes World's Biggest Company: Bubble or Destiny?

Episode Date: June 19, 2024

Nvidia has ridden the AI wave all the way to the top of the public markets, exceeding the market cap of Apple and Microsoft to become the world's biggest company for the first time. NLW discusses what it says about the state of AI in public markets.

Today on the AI Daily Brief, Nvidia is now the world's biggest company. Is it a bubble or simply destiny? The AI Daily Brief is a daily podcast about the most important news and discussions in AI. To join the conversation, follow the Discord link in our show notes. Hello, friends, a couple quick announcements before we get into today's show.

Today is June Teeth, a holiday in the USA. And so instead of a full show, we are doing just a main episode, no headlines this morning, focused, of course, on Nvidia becoming the biggest company in the world. Two quick additional things, though, before we get into that. First of all, this podcast is now available as a video on Spotify. Secondly, I mentioned yesterday that Super Intelligent, our platform for learning how to use AI in a fast, fun, practical way, is getting ready to launch a new team's experience, and we are looking

for partners to be early testers of that. If you're interested in learning more, check out be super.aI slash partner. Send us a note, and we'll talk about what that new enterprise offering is going to include, how you can get your team up and running on Super. And by the way, we like teams of 10 as much as we like teams of 10,000. So if you are interested, please reach out. With that stuff out of the way, though, let's dive into the episode. Welcome back to the AI Daily Brief. Today, we are discussing Nvidia. On Tuesday, June 18th, Nvidia for the first time overtook Apple and Microsoft to become the largest company in the world. In some ways, this felt inevitable. It's the track that

Nvidia has been on. However, there is still something remarkable about it, and it has a huge amount of chatter everywhere from the AI space to conventional markets. Since the launch of chat GPT in late 2022, Nvidia has had an absurd 860% gain, reaching a market cap of more than 3.3 trillion. To say the media coverage is breathless would be an understatement. Bloomberg headlines are framing this as a 591,078% rally since the 1999 IPO. Dan Ives of Wedbush securities wrote in a research note, Nvidia's GPU chips are in essence the new gold or oil in the tech sector as enterprises and consumers quickly head down this path with the fourth industrial revolution well underway. The question, of course, is whether this is the mother of all bubbles, or whether there is something

really legitimate here, something that we should be paying attention to, or perhaps whether it's both. Here's two different statistics that tell different sides of this story. Microsoft and LinkedIn recently put out their 2024 WorkTrend Index annual report, something regular listeners will know I have referred to quite a bit. One of the big banner statistics is that 75% of the global knowledge workers are now using AI. As an aside, 78% of those are effectively smuggling it into work, which says a lot about how workers themselves are thinking about AI and the lag to their organizations actually figuring out their AI strategies. But the point is that there is a significant amount of AI usage. And that seeing that much enterprise level adoption, even if it

is bottoms up enterprise adoption of a technology, suggests for just how significant this is going to be in the future. On the flip side, another statistic that you might have heard a number of times is the disparity between how much has been spent on Nvidia chips and how much aggregate revenue generative AI startups have brought in. In an article from the end of April in the Wall Street Journal, Sonia Huang from Sequoia said everybody is assuming if you build it, they will come. AI is a field of dreams. The amount of money it takes to build this stuff has

vastly exceeded the amount of money coming out so far. The specific statistic that they pointed to was the idea that 50 billion had been spent on Nvidia chips to train LLMs, but the generative AI startups had only brought in $3 billion in revenue. The reason that this matters is that it speaks to the difficulty of valuing NVIDIA right now. First, let's talk about revenue. Invita brought in a massive $60 billion in revenue during 2023, and is currently growing that at more than a 250% pace. In other words, this isn't just about pricing in some amorphous future. It's also about pricing in one of the most staggering revenue growth stories in history.

On the flip side, by traditional metrics, the stock is getting to fairly extreme levels. Nvidia is currently trading at a 77 price-to-earnings ratio. That's well above the 10-year average of around 30 for the median S&P 500 stock. But at the same time, it's nowhere near as extreme as previous frothy stock market darlings. Tesla hit its peak PE ratio in December 2020 at more than 1,200. For a more apples-to-apples comparison during the dot-com bubble, Cisco was considered the infrastructure company underpinning the internet buildout, and that stock hit a peak-pea ratio of 200 in 1999.

Also, even though this isn't how startups are valued, if you do apply these metrics to startups like OpenAI, it looks a little less frothy than you might assume. The information recently reported that OpenAI's annualized revenue had doubled to $3.4 billion, which based on their latest valuation of $86 billion in a recent tender offer of employee shares, that's only about 25 times forward revenue. Let's move away, though, for a second, around the question of whether Nvidia is valued appropriately, overvalued or even undervalued by the markets, and instead look at its place in the overall market dynamics.

Something that is pretty clear is that NVIDIA's rapid ascent has totally distorted the stock market. Both the S&P 500 and NASDAQ indices are at all-time highs on the back of NVIDIA's strength. And that is quite literal. In many ways, NVIDIA is putting the entire market on its back. Nvidia's rise has accounted for more than a third of the gains in the S&P 500 so far this year.

Heading into this year, the U.S. stock market was already at historically high levels of concentration in a small handful of top companies. Currently, the top seven companies in the S&P 500, the so-called Magnificent Seven, account for 28% in market cap. This number exceeds the level of concentration at the height of the dot-com bubble. The performance of the Magnificent 7 is also masking significant weakness below the surface. The S&P 500 is up 15% so far this year,

while an equal-weighted version of the index which measures the average stock performance is down 0.5% for the year. Only 48 stocks in the index are currently above their 50-day moving average, and this phenomenon has been getting acutely worse over the past month. According to Bloomberg data going back to 2002, this is the worst decline in market breadth during a rally. There is no other four-week span where the index has gone up so much with so many stocks within the market going down. The question is whether this is a divergence that is bound to close or if the leading stocks can continue bounding ahead. For some recent precedent, the second worst span in recent history for market depth was in late May last year,

and over the following few months, the gap between mega-cap stocks and the rest of the index didn't snap closed but simply moderate it a little. This is a huge point of conversation on FinTwit. For example, TXMC trades writes, This isn't the 2000s, but it has all the hallmarks of being a mega-cap tech bubble, and odds strongly favor it resolving in a similar fashion. Timing is impossible, though. Several times I've said I think this is a bubble, only to be told the volume of unprofitable.coms in 2000s isn't here today.

This is a concentration bubble. A clown car of names in the top 20 are absorbing all flows, while the majority of stocks from ranks 21 to 5,000, languish. This is something that we've been watching for more than a year now, really ever since the beginning of the tightening cycle, which happened to coincide, by the way, with the launch of ChatGPT. Throughout geopolitical chaos, banking crises, somehow AI optimism continued to keep the market afloat. Mike Green has one of the better explanations for this growing trend of mega-cap outperformance with his work on passive investing. The super condensed version of the thesis is that most U.S. stock investing is done via index-linked ETFs,

which is a mainstay of retirement funds in particular. When these ETFs receive inflows, they mechanically buy stocks in proportion to their market caps. These products don't care then whether a stock is overvalued or undervalued. They simply buy at market price. That means that more capital is continually chasing after the largest companies in the index. This effect compounds as the biggest stocks go higher and become a more dominant part of the index. Green concludes this thesis by asking what happens if those perpetual inflows turn into outflows. This could come about if more people are drawing down on their retirement than are contributing to it, either through demographics or a spike in unemployment. Green's point is that the reflexive

loop that has pushed mega-cap stocks higher is also likely to be reflexive on the downside. Fascinatingly, with all that framework laid out, some are starting to compare NVIDIA to Bitcoin because of the similarities in their returns and volatility. Over a 10-year time horizon, NVIDIA stock has now become basically the only other asset to come close to Bitcoin's returns, and over shorter time periods, NVIDIA has outperformed Bitcoin. An important note, however, is that Bitcoin is of course not present in most retirement accounts, whereas a standard 6040 portfolio now has a 4% exposure to Nvidia. The S&P index funds that form a core of most retirement savings now have a 7% allocation to

Nvidia and another 30% across mega-cap stocks. These funds would obviously rebalance if big cap tech stumbled so losses would be muted. However, it's difficult to imagine that a rise in the rest of the stock market could make up for a hard landing for Nvidia, meaning that in some ways, a heavily financialized U.S. economy is betting heavily that the Nvidia bubble never burst. But let's come back to these two statistics. On the one hand, AI being used by 75% of global knowledge workers versus 50 billion spent on Nvidia chips versus $3 billion in revenue.

It may surprise you, not at all, but I think that the more misleading of these statistics is absolutely the revenue versus amount of money spent on Nvidia chips. We are at the dawn of a fundamentally new computing paradigm. It is already changing how we interact with computers. It is changing jobs. It is changing what jobs are done. We are barely scratching the surface on the impact of this technology.

Viewing it strictly in terms of how much generative AI startups have been able to figure out how to monetize inside of about 18 months is a ludicrously short-sighted way of looking at things. And Nvidia is fairly arguably the most significant company in that transformation. So how do you put a price on being the foundational layer for an entirely new era of computing, an entirely new era of work? It's really difficult. Is 77 times forward earnings? which again is less than half of what the bubble was looking like for certain key stocks in the dot-com

era inappropriate? I'm not sure. One thing that is for sure, my bet is, of course, that the AI market not only increases dramatically, but absorbs and eats huge swaths of other markets over the coming decade. Whether that means that Nvidia is in for a price correction or not is a totally separate question, of course. One thing that is notable is that for as well as they're doing, Nvidia's CEO Jensen Huang is not resting back on his laurels. Writes the information NVIDIA's Jensen Huang is on top of the world, so why is he worried? Subheader, NVIDIA's CEO, mindful of the downfall of one-time hardware giants like Cisco, is aggressively pushing his company into software and cloud services, putting it in competition with its biggest

customers. The article is worth reading in its entirety, but part of Jensen's concern is, quote, whether NVIDIA's biggest customers were going to run out of data center space to install its artificial intelligence chips. Point being, if you think that the picture of how the industry is organized right now is likely to stay static, I think that that's absolutely crazy. So really fascinating moment in time. If it's not clear, I'll say it again, that none of this is financial advice. I think one of the most interesting things about right now is that questions that are usually simply in the domain of venture capitalists who are inherently forward-looking are now splashed all over the public markets as well. When Nvidia, a public company, the biggest

public company in the world, is at the center of this massive technology transformation, it creates some real weirdness on how to think about valuation. I'm sure a story that we'll continue to return to. For now, though, that is going to do it for today's AI Daily Brief. Until next time, peace.