AI and The Magnificent Seven
The first half of 2023 has been pretty good for the market with a handful of the large tech companies leading the charge. I recommend watching my video quarterly market review, but I have to make one correction. In the video I say Magnificent Seven companies are responsible for 90% of the S&P 500’s return for the first half of 2023 when they are actually responsible for 73%.
You can also read the market review below with a section on innovation and the history of titanium, which was left on the cutting room floor for the video.
Going into 2023, many people felt a U.S. recession and market decline was inevitable. However, despite interest rate increases, a flash in the pan regional banking crisis, and debt ceiling standoff, the inevitable market decline has failed to materialize. What gives? Well, to understand the past six months you need to have a working knowledge of videogame intros, artificial intelligence, and The Magnificent Seven.
Videogames and A.I.
If you’ve played videogames at any time over the past twenty-five years, you may have noticed that the NVIDIA intro is a common developer credit you’ll see before your game starts. This is because NVIDIA designed the GPUs (graphics processing units) that are the backbone of the gaming industry. NVIDIA’s GPUs are what make the difference between a video game from 2001 and one made in 2021.
Are videogames responsible for NVIDIA’s 190% return in 2023 and increase from a $400 billion company to being worth more than $1 trillion? Not quite, but this is where the story gets interesting. The same GPUs responsible for improved videogame graphics are the best tools to train neural networks on the datasets necessary for large language models (LLMs).
LLMs are a form of Artificial Intelligence (A.I.) that are trained on vast amounts of text data to generate human-like text responses. They took the world by storm in November of last year with the release of ChatGPT 3. GPT 3 (and later GPT 4) was the first widely available LLM that was useful for tasks such as coding, marketing, and writing. The hype over A.I. hasn’t died down with developers working to discover new applications every day. NVIDIA, being the supplier of the GPUs necessary to train LLMs, has skyrocketed in value to become one of the top ten companies in the U.S.
The Magnificent Seven
You may have noticed the S&P 500 has had an exceptional year with a 16.89% year to date return. Which is welcome news for investors who had their portfolios bruised in 2022. What this number masks though is the narrowness of where that return is coming from. Enter the Magnificent Seven, who some believe will be the greatest beneficiaries of the A.I. revolution. 73% of the gains from the S&P 500 can be attributed to Apple, Microsoft, NVIDIA, Amazon, Meta (Facebook), Tesla and Alphabet (Google). While “Magnificent Seven” companies returned on average 89% in 2023, the remaining stocks in the index averaged roughly 7.5%, with just under 40% of the companies posting negative returns. The below chart highlights the disparity of returns between the largest companies and the rest of the market. At the end of May, the five largest stocks had been responsible for 78% of the market’s return. Historically, the five largest companies have made up 3% of the market’s overall return. The concentration of returns amongst these outliers is not the breadth we’d like to see and deviates from the long-term average.
The Takeaways
This leaves us with lessons we can learn from this quarter. First, despite the concentration of returns this year, I still believe they provide a strong case for diversification. For matters of practicality, we just don’t know when great returns are going to arrive or where from. The 89% average return from these seven companies sounds wonderful in 2023 but would you have wanted to own them in 2022 when their average return was -46%? I think not.
On the philosophical front, as NVIDIA’s rise shows, we still have no idea where innovation will come from and when. Few people appreciated the role GPUs (remember, primarily used for videogames) played in developing A.I., fewer foresaw the potential productivity of LLMs, and no one had November of 2022 for the date of their arrival. This randomness of innovation reminds me of the history of titanium.
Titanium is known in the aerospace industry for its durability and lightweight compared to stainless steel, but in 1952 it also became a useful tool for the medical industry. Prior to 1952 it was believed the body would reject any foreign object. However, that year while the Swedish researcher Per-Ingvar Brånemark was studying blood flow in the healing of rabbit bones, he saw the titanium camera casing used to take pictures of the healing process had bonded with the rabbit bone. This was revolutionary and Brånemark’s later research showed that titanium could be implanted within the body with no illeffects. His research also discovered by accident that bone can conduct sound. The first bone conduction hearing aids, artificial hips, and modern dental implants can be traced to a titanium camera casing used in rabbit experiments. Innovation is random.
The final lesson is one we repeat time and time again: stick with your strategy and stay in the market. There is always a news story that tests our resolve to stay invested whether it be Silicon Valley Bank back in March or the debt ceiling standoff in May. However, trying to time these events consistently is impossible and a fool’s errand. Investment returns do not happen at a steady march. They occur at their own unpredictable pace whether marching slowly forward, backsliding, or with random sprints like we saw in June. Therefore, missing the best week can significantly impact your investment return.