After the close today, Jensen Huang at Nvidia will reveal third-quarter financial results. At least twice every year, we like to use Nvidia’s quarter earnings report to write about our investor origin story. Our business at Markman Capital is about teaching investors how to build real wealth. This process begins by understanding the importance of compounding.
We started putting investors into Nvidia shares back in 2015, more than a decade before ChatGPT. Shares were deeply unpopular. The consensus opinion, even back then, was that Nvidia was an expensive, niche business with limited upside because the sale of its graphics processing units (GPUs) depended largely on high-end video game players.
Huang told investors they were missing the point. GPUs were becoming ground zero for artificial intelligence (AI), a new business vertical that would transform high-performance computing. He said that business would be worth trillions. Skeptics laughed, but operators of hyperscale data centers were intrigued. Executives at Google, Amazon.com, and others began buying and testing GPUs
The hyperscalers soon realized that GPUs were foundational for manufacturing AI.
They took Nvidia GPUs, added electricity, and created intelligence. The possible use cases for this output were/are unlimited. Internally, Google began experimenting with mathematical and material science problems. Executives knew immediately they could use AI to materially change their business.
The crux of the Nvidia business story, the part that people still don’t get, is that Nvidia is making the machines that make AI. Its customers are now building next-generation AI factories to distribute intelligence at scale. They understand AI is going to become the foundation for how every business operates in the future. It’s a market worth trillions of dollars.
As a point of reference, Nvidia’s sales to data centers in fiscal 2015 were worth $317 million. When Huang reports Q3 results this evening, that business is expected to be about $43 billion, and growing quickly.
Data centers is a $170 billion business less than a decade later.
Skeptics see Nvidia as the most vulnerable business in a bubble. They say that eventually, the hyperscalers will stop investing in Nvidia gear because the payoff doesn’t make investment sense. Again, this is a fundamental misunderstanding of the product. AI is not ChatGPT, nor is it self-driving car technology. These are applications that sit atop AI. There are many other applications, some that have not even been imagined.
But I’m getting off topic. We started putting our members into Nvidia shares at less than 50 cents per share, adjusted for stock splits. Shares currently trade in the $187 range. The return on investment is 37,100%. Many of our members have become millionaires many times over from relatively small initial investments. The secret is compounding.
Buffet, Bezos, Gates, Musk, and countless other ultra-wealthy investors owe their fortunes to the magic of compounding. While most investors are constantly looking for what comes next, buying and selling, the super-rich simply hold onto their shares. If Nvidia shares rise by only 50 cents between now and year-end, Markman members will earn a 100% return on their initial investment.
Nvidia traders and bears will try to impress you with their acumen and stories of doom. Their success is fleeting at best. Trading involves paying taxes, and that assumes successful timing. The maximum unleveraged gain on any Nvidia short position is 100%. You don’t have to be a mathematician to quickly learn that buying and holding is more efficient and safer.