Should You Take Some Chips Off the Table with Nvidia?

For clients who didn’t already own it, we added Nvidia (NVDA) to suitable portfolios back in April, during the market sell-off triggered by tariff worries. We purchased the Canadian Depositary Receipt to avoid currency swings, but for context, U.S. shares traded roughly between $102 and $105 on April 23—the day of our purchase. Today, NVDA sits just above $210 per share.

Yesterday, Nvidia became the first company to achieve a $5 trillion valuation. To put that figure into perspective, it’s larger than both the Energy and Materials sectors combined. With gains achieved this fast, it’s natural to wonder if it’s time to take some chips off the table.

The answer depends on your portfolio, but for our clients, we’re holding. While the stock may seem expensive at first glance, Nvidia’s growth justifies its stock valuation.

In its latest quarter, Nvidia once again exceeded expectations. Even after losing access to China due to export restrictions, revenue and earnings jumped more than 50% year-over-year, comfortably beating analyst estimates.

Nvidia’s leadership in AI hardware is undeniable. Its Blackwell chip—the top-performing AI chip available—offers up to four times the training speed and thirty times the inference capability of its predecessor, the A100. In short, no competitor comes close to matching Blackwell’s performance.

Valuation concerns are understandable. Nvidia’s 30x price-to-earnings ratio appears high, but its PEG ratio—which is simply the price-to-earnings ratio relative to its expected earnings growth rate—is far more reasonable at 0.8x for fiscal 2026 and 0.7x for 2027. (As a rule of thumb, we’ve always considered a PEG ratio over 2.0x to be expensive, and 1.0x to be a bargain).

Nevertheless, like all investment ideas, there are risks to consider. The company depends on a relatively concentrated customer base, with its top two customers accounting for nearly 40% of total revenue. And, losing access to China’s $50 billion AI market limits potential.

AI spending is still in its early stages. Nvidia estimates $3–4 trillion will be invested globally in AI infrastructure by the end of the decade. It doesn’t need to capture all of this to keep growing; simply participating in a booming growth sector will be enough to drive continued expansion. According to the Jevons Paradox, as technology becomes more efficient, demand often increases, allowing all companies to benefit from greater adoption.

Nvidia’s valuation may look expensive on the surface, but its future growth justifies today’s price. Taking chips off the table might feel prudent, but we remain confident in Nvidia’s long-term leadership in AI computing.

-written by Jeff Pollock