Invest at your own risk; this is not financial advice! This is not a recommendation to buy or sell any securities discussed in the article.
AI eats the world has become a phrase that is stated very often these days. It used to be “software eats the world” in the early 2010s, and increasingly, it seems that AI is having a feast now.
In June, I published my article about investing in the age of AI with examples of companies that will benefit and also some that will lose. Make sure to have a look!
Are we in a bubble?
We are somewhere in a bubble, and fantastic technological progress at the same time. I don’t believe the doomsayers that we are close to a 2000s internet crash, but I do see some overhyped AI startups and wannabe datacenter companies.
All in all, the large players Amazon, Alphabet, Meta, Microsoft, and Nvidia are still reasonable within a normal valuation window. That being said, the current insanely high spending on CapEx will leave its mark on the income statement going forward. All those lovely GPUs want to be depreciated, and the depreciation time will be closer to 2 years than 6-8 years. Nvidia is releasing its newest generation on an annual basis now, and the improvements are staggering. That means that the previous generation will lose value even faster. Great for technoclogies breakthrough, not so great for your accountants.
On a forward P/E basis, we are still within a normal span, far away from triple-digit P/E ratios. I have excluded Amazon from this chart since Amazon had a negative P/E ratio in early 2022, which messed up the chart.
Talking about Amazon: I believe Amazon is a fantastic company and a great addition to any portfolio. Read my most recent analysis here:
Looking at the forward EV/FCF ratio shows a more mixed picture. The insanely high CapEx is leaving its mark on the FCF. The only company with a forward EV/FCF ratio below 49 is …. NVIDIA. I guess you would not have expected this one.
Amazon has been traditionally a very high CapEx spender. Think of all those warehouses and the AWS ramp-up. Now, only companies such as Alphabet are massively increasing their spending.
Alphabet CEO Sundar Pichai said last year
The risk of under-investing in AI is dramatically greater than the risk of over-investing
Meta CEO Mark Zuckerberg followed up with this quote just weeks ago
“The very worst case would be that we have just prebuilt for a couple of years”
These guys are ready to spend/burn billions of dollars on the new AI buildout.
The Cisco vs Nvidia Case
The valuation will go up, but I believe we are still in an environment where we are far away from the Dot-Com bubble. There are countless charts that somehow overlay the current rally with some random data from the past and try to make a claim.
Let’s start with this one: Your logical conclusion must be: Oh my god, we are all doomed, Nvidia is about to lose 90% of its value. Well, not so fast, my brilliant Watson.
While the charts look somewhat similar and Nvidia has been rising like there is no tomorrow, there is a slight but very important difference:
Nvidia is actually earning a lot of money, and its P/E ratio is ok, or maybe even cheap. Cisco’s P/E ratio dropped like a rock after the bubble burst. P/E or price/earnings ratio is influenced by two metrics: The numerator, stock price, and the denominator, earnings.
If the P/E ratio falls, it is either due to a falling stock price, rising earnings, or both. In Cisco’s case, the P/E ratio dropped due to a collapsing stock price.
It took Cisco 25 years to get back to its height from 2000. Ouch!
Coming back to our beloved Nvidia:
The surge in the stock price is driven by an insane increase in the fundamentals. Remember my mantra: Price follows earnings.
Compared to 2017, Nvidia’s revenue is up 19x, and the net income increased 44x. Talking about non-linear growth rates. You might ask: “Hey, 41, why didn’t you invest all your money in Nvidia?”
To be honest, I did not expect this. We humans (myself included) are bad at extrapolating exponential growth. I just could not imagine this happening. I tip my hat to long-term shareholders for their sheer conviction of holding onto a stock that has gone as ballistic as Nvidia. This could mean life-changing sums of money. Chapeau
Ending with another chart: Maybe we are not in a bubble at all. Charts can be useful or misused to transmit the message that you want to convey.
So where are we today?
Technological Shift
Luckily for you and me, I am not the only soul on this planet who thinks about this a lot. There is a fella called Ben Evans, who just published an amazing slide deck with the same title as this article: “AI eats the world”. You will find it here, and I highly recommend that you read it.
https://www.ben-evans.com/presentations
He has done such a great job, and I am using some of his slides to explain what is happening today:
AI is the next big platform shift. You still remember the first PCs, and it was pure magic. A machine that can do calculations in a fraction of the time it took a human. The web connected all of us to each other, and with the rise of the smartphone, we had all the world’s knowledge in our pockets. GenAI is now bringing smart assistants for “free” and there will be so many use cases that we cannot imagine yet.
The companies leading today (OpenAI with ChatGPT, Alphabet with Gemini, etc) might not be the ones who will lead in a couple of years. Below is an example of PC unit sales; another example is the beloved Finnish company Nokia, which once seemed indestructible. And then came the smartphone…
The internet has had a major impact on our lives. Almost 60% of all US heterosexual couples meet online these days. How did we ever find a partner in the past?
I’ve mentioned it before, and this graph visualises in a great way the ridiculous increase in CapEx by the Big 4 players.
The US Data centre construction value is about to overtake the office construction value. An interesting side note: You can see nicely the housing boom from 2005-2009 until the big crash. Retail (in grey) looks like an insect pretending to be dead. That is another effect of the internet boom.
Circular Revenu
All these investments are either driven big the big 4 or ChatGPT owner OpenAI. Unlike the Big 4, OpenAI has barely any revenue and happily signs contract after contract to buy more computing power than it can ever pay for. This led to the very interesting phenomenon of “circular revenue”.
Circular revenue sounds harmless, but it might have real consequences. Nvidia is using the power of its balance sheet and the large sums of cash to create its own demand.
Talking about AMD: You will find my AMD deep dive here
OpenAI uses its name and current high mind share to aggressively build something out of nothing. I guess the long-term plan is to become too big to fail and ask the US government to vouch for its credit in order to have a national champion at home. So far, there is no real plan or pathway for a long-term profitable business case.
ChatGPT has 800 million global weekly users. The world has 6 billion people with internet access, and China, with 1 billion internet users, banned ChatGPT, leaving 5 billion potential users. Therefore, ChatGPT has already reached 16% of all potential users, but managed to convert only a fraction into paying users. How is this ever going to be profitable?
Who is successfully using AI?
There are many startups working on new solutions. The startup accelerator Y Combinator is focusing almost exclusively on AI startups. Interesting fact: OpenAI Sam Altman was the Y Combinator CEO from 2014 to 2019.
Despite all the headlines of AI taking over every kind of job and millions of jobs in danger, the rollout takes time. Only a fraction of companies have successfully implemented AI agents.
Talking about agents: I believe that Salesforce is on a great track and will profit from the rollout of its agents. You will find my deep dive here:
Innovation is a strange beast. It is intriguing and all-encompassing but takes its time. Cloud, which has been around for quite some time, only covers 30% of all enterprise workloads these days. AI will also take its time.
What is the effect of using AI?
AI offers you “very cheap” or almost free intellectual labour. It is sometimes described as having access to an unlimited number of interns. Your typical intern is smart but still has a lot to learn about how to apply his knowledge in the real world. But he/she is very good at getting basic research and repetitive tasks done. What if you had an unlimited amount of these guys? What would you do with them?
Would you automate many tasks and drastically shrink your workforce, or would you get so much more done with the same human workforce?
With the rise of the steam engine, there were many new job profiles and things that could be done. Given the AI analogy, this did not lead to massive unemployment in the UK as a result of the steam engine.
I believe that certain sectors will be directly hit by the rise of AI. I create the main picture of my articles with Midjourney. For example, last week’s post prompt looked like this:
As a writer, these AI tools are a gift for me. I do not violate the copyright of somebody by taking their pictures, and my self-created pictures or taken photographs are often not exactly fitting for my article. AI helps me with this tremendously, and I can spend more time focusing on my writing, which benefits you, the reader, as well.
The new AI world is exciting, and companies such as Alphabet are pushing hard to move the boundaries. Alphabet’s self-driving cars (Waymo) are conquering more and more cities and are doing literally millions of miles already.
Alphabet was such an obvious buy for so many months. I hope you read all my best buys articles and acted upon them
What do we learn from the past?
This is the end of the free section. If you want to
✅ Read the full article and learn about what we can learn from the past, and my final thoughts
✅ Have access to all previous articles
✅ Profited from previous picks
✅ and want to support me for my time and effort
Then, please become a paid subscriber. To all existing paid subscribers: You rock! 🙏
