Semiconductor Stocks as a Defensive Investment

The economic commentariat opines that if there is a severe downturn in the stock market—and even President Trump seems resigned to a recession this year— technology stocks will be hit the hardest, on the general theory that the bigger they are, the harder they fall. Tech has led the market up, so it just makes sense that it will lead the market down.

We think that’s far from obvious; in fact, if we had to bet (which we do!!!!), we think that in our favorite corner of the tech market—semiconductor stocks—company fundamentals will be relatively less hard hit by a recession. As for stock prices, in a panic, tech stocks including semiconductors may lead the way down for several months, but the chips could also be the first to stabilize and rise again.

We say that for two reasons.

1. The leading semiconductor stocks are not nearly as overvalued as the commentariat claims—and nearly are all less overvalued than they were a month ago.

It’s true that by many standards the value of, say, Nvidia (NVDA), looks ridiculous with its PE tickling 40 and a price to sales ratio exceeding 20. How can a large-cap stock merit that?

Fair enough, but how would you price a stock with a three-year compound annual revenue growth of 69%, three-year compound earnings growth of 93%, operating margins above 60%, a Return on Capital of 87%, and a Return on Equity of 129%. Oh, and with $43 billion in cash hanging around.

Strange as it may seem, NVDA is undervalued. Investors pay for growth and rightly so. In NVDA’s case, they are underpaying. The stock sports a price-earnings growth (PEG) ratio of 0.7, with 1.0 considered neutral or fairly priced, and lower being better.

What’s true of NVDA is generally true, to a less extreme degree, for the semiconductor sector. We’ll have lots more detail on that in the upcoming issue of Gilder's Technology Report as we systematically review our portfolio. For now, note that NVDA is the only one of the Magnificent Seven with a PEG below 1.0 and the only one primarily engaged in semiconductor manufacturing and sales.

2. Growth in semiconductor sales will accelerate.

Warren Buffett famously proclaimed, “in the short run the marker is a voting machine; in the long run it is a weighing machine.” That's just another way of saying that over time earnings govern prices. How much time? Generally, too much for our tastes, but this time round, if there is a crash, we think the resilience of the semiconductor sector will become obvious early on. At a minimum, it will present a massive buying opportunity.

The AI boom is just beginning. We have taken only the first few baby steps toward the future we have called “the animated world.” That’s the world we get when the vast majority of manmade objects incorporate three semiconductor-based technologies:

• “Supersensors” far superior to human senses in gathering data from their environment.

• Local or “edge” artificial intelligence capable of interpreting that data in microseconds.

• Actuators to respond to the AI’s conclusion by executing some physical change.

It will be a world in which machines act “as if” they were alive. (They won’t be, any more than an AI can “think,” but they will be fun to watch.)

As we pointed out here, developments such a DeepSeek, which make AI more affordable, will not slow the AI boom but accelerate it. After all, Nvidia triggered the boom precisely by making AI affordable and practical. DeepSeek (and competitors nipping at its heels) will boost sales for Nvidia and boost the entire global economy by making AI a universal tool.

The animated world is closer than most of us imagine. Tesla, for instance, is finally being revealed as what it always was, not a car company but an intelligence company. It’s self-driving AI will be more important to its future than its batteries. Tesla’s AI is emerging as the enabler of Tesla’s humanoid robots, which within this decade will be cheaper than a budget model Tesla car. In the 2030s they will be a part of every household.

In short, NVDA’s sales will go up, not down, and the sales of the entire AI ecosystem will ascend as well. When a product changes the world, recessions are a hiccup. No sector will prove as resilient to recession as the AI industry.

As for the rest of the so-called Magnificent Seven, AAPL seems the least likely to benefit from the AI boom—and hasn’t so far—AMZN could either rise to even greater things or discover AI has made every retailer a competitor. Its data center business should grow; the same for META, GOOG, and MSFT, which already benefit from its consumer-level AI software. TSLA is an AI company.

Most likely to join the Seven? Our Gilder’s Technology Report newsletter subscribers will get our answer this week.