Nvidia reaching for the Sun
In early 2000, investor hype pushed the price of Sun Microsystems to an absurd extreme. It was destined to disappoint its shareholders, as its CEO noted. Likewise, can artificial intelligence (AI) help Nvidia grow enough to justify its current valuations?
In the late 1990s, Sun Microsystems was one of the world’s winningest ‘New Economy’ companies. It was a leading producer of high-performance workstations and servers (computers that serve other computers over a network), and developed the Java programming language, business software and storage solutions for companies.
It was a star of the emerging internet, and its shares raced up for years. By March 2000, the firm was worth $150bn. Its revenues at the time were about 15bn per year, so its price-to-sales ratio at its height was about 10.
Two years later, after the Nasdaq crash when Sun had lost 90% of its value, its CEO Scott McNealy addressed investors as follows:
“At 10 times revenues, to give you a 10-year payback, I have to pay you 100% of revenues for 10 straight years in dividends. That assumes I can get that by my shareholders. That assumes I have zero cost of goods sold, which is very hard for a computer company. That assumes zero expenses, which is really hard with 39,000 employees. That assumes I pay no taxes, which is very hard. And that assumes you pay no taxes on your dividends, which is kind of illegal. And that assumes with zero R&D for the next 10 years, I can maintain the current revenue run rate.
“Now, having done that, would any of you like to buy my stock at $64? Do you realise how ridiculous those basic assumptions are? You don’t need any transparency. You don’t need any footnotes. What were you thinking?” (our emphasis)1
A firm’s sales are often an excellent indicator of how the business is doing. Sales numbers tend to be clear and simple, and can’t easily be fudged by accounting shenanigans – which often confuse earnings measures.
The price-to-sales ratio is a measure of how much those sales are worth to investors. When it’s high, those sales are very expensive. In Sun’s case, McNealy said it would have needed to grow phenomenally to justify its early-2000 valuations. It did grow strongly over the following decade, as the internet spread across the world, but not nearly enough. In 2010, it was bought by Oracle Corporation for $7.4bn. Investors who bought Sun shares in the late 90s and held on lost almost all their money.
How did Sun become so absurdly expensive? Because of two features of human behaviour.
First, investors tend to be seduced by growth. They see a company growing fast, they make money from it for a while, and they extrapolate its growth far into the future. A hype cycle builds up. But few companies continue growing fast for a long time, especially large ones. Competition intensifies, roadblocks appear, and the firms’ numbers end up slowing.
Second, investors are social animals and like being invested in fashionable stuff on the way up. When a bandwagon comes past full of happy people, they tend to leap aboard without necessarily scrutinising the proposition carefully. So hype cycles drag in more and more investors and can be self-fulfilling for a while.
This is until the business disappoints, and everything falls in a heap. However, the hype cycle can lead to companies and markets being dissociated from fundamentals for a long time.
The Nasdaq bubble of 1998-2000 was the classic example. There have been many since. Think of Zoom Video Communications, whose price soared from $75 in January 2020 to $560 in October 2020, in the first phase of the pandemic. We were all using Zoom, and at its height, the company was worth $160bn. Then investors realised the hype had gone too far and began selling. Zoom is now trading around $68, and the firm is worth only $20bn.
How Magnificent are the Seven?
We have concerns about the Magnificent Seven, the seven huge technology stocks – Nvidia, Meta, Tesla, Amazon, Alphabet, Apple and Microsoft – that have been driving global markets this year. By late August, they were up by an average of 89% (!), compared to the rest of the S&P 500 on 3%. Well done to them.
And their valuations have soared. Tesla has a price-to-sales ratio of about 8.8, Microsoft 11.5, and Nvidia 32.9. Nvidia’s number is three times that of Sun Microsystems at its height, and is one of the highest for any large company in world financial history. The firm would need to grow spectacularly to justify this valuation.
Let’s assume that Nvidia’s investors expect a return of 10% per year. Its sales would have to grow by 30% per year over the next decade to bring its price-to-sales ratio down to a more reasonable 5 by 2033. That would mean sales surging from $33bn this year to $450 bn in 2033, which is higher than Meta, Microsoft and Tesla together. No huge company has ever grown sales that fast… not even Apple!
Sure, Nvidia is benefiting from the boom in artificial intelligence, which should boost world economic growth a little. But AI will have a limited impact on most companies and it may take many years for this to be achieved. Don’t expect a world productivity miracle anytime soon.
One of 7IM’s core investment axioms is to be cautious of fashionable assets. When we see investors piling into trendy companies, markets or products, we prefer to keep our distance, because they tend to become overvalued and deliver poor returns in the future.
We follow this discipline because we believe it works. We limited our exposure to the Magnificent Seven – missing out on some equity growth – through 2023 because we know that in the long run, avoiding market fashions and hype tends to pay off.
This is one of those occasions. The Magnificent Seven have been over-hyped. There will come a time when we will look back on investors in Tesla, Microsoft and Nvidia in late 2023 and ask, 'What were they thinking?'