Palantir Technologies (PLTR -9.04%) and Super Micro Computer (SMCI -3.43%) are two of the most popular artificial intelligence (AI) stocks on the market. Since January 2023, Palantir has returned 1,900%, and Supermicro has returned 480%. But certain Wall Street analysts think the stocks are wildly overvalued.

  • Rishi Jaluria at RBC Capital recently set a target price of $40 per share for Palantir. That implies 69% downside from its current share price of $130.
  • Mike Ng at Goldman Sachs recently set a target price of $24 per share for Supermicro. That implies 50% downside from its current share price of $48.

Here's what investors should know about Palantir and Supermicro.

A downward-trending red arrow overlaid on a $100 bill.

Image source: Getty Images.

Palantir Technologies: 69% implied downside

Palantir develops data analytics and artificial intelligence (AI) software for customers in the commercial and government sectors. International Data Corporation has ranked the company as the market leader in decision-intelligence platforms, and Forrester Research has recognized its technology leadership in AI platforms, a market forecast to grow at 40% annually through 2028.

Palantir reported solid first-quarter financial results. Revenue jumped 39% to $884 million, the seventh consecutive acceleration, driven by particularly strong sales growth in the U.S. commercial and government segments. Non-GAAP (non-generally accepted accounting principles) net income increased by 62% to $0.13 per diluted share. Management also raised its full-year guidance, with sales now expected to increase by 36% in 2025.

In short, Palantir has a strong presence in a quickly growing industry, and the company is evidently executing on that opportunity. The problem is valuation. Wall Street expects the company's adjusted earnings to increase by 31% annually through 2026. That makes the current valuation of 280 times adjusted earnings look absurdly rich.

Alternatively, Palantir currently trades at 105 times sales. No other company in the S&P 500 has a price-to-sales multiple above 35. That means Palantir's share price could drop 66% and it would still be the most expensive stock in the S&P 500. I doubt the stock will fall 69% (as Rishi Jaluria anticipates), but I think shareholders should maintain very small positions. The slightest bit of bad news could cause this stock to crash.

Super Micro Computer: 50% implied downside

Super Micro Computer develops data center servers and storage solutions, including rack-scale systems optimized for artificial intelligence and other high-performance computing applications. The company says its internal design capabilities and modular approach to product development enable it to bring new technologies to market very quickly, often months ahead of its competitors.

Supermicro accounted for 6.5% of global server sales during the fourth quarter of 2024, second only to Dell Technologies, which had a 7.2% revenue share. More importantly, Supermicro has emerged as the early leader in AI servers, a market forecast to grow by 37% annually through 2030, due to its rapid time-to-market capabilities.

Some analysts think Supermicro will lose market share as competition increases because there is nothing about its business model that cannot be replicated. Supermicro is not responsible for the innovation; it simply assembles chips built by other companies, such as Nvidia, into servers. Also, KeyBanc analysts argue that Supermicro has "structurally lower margins" because it owns fewer patents than its peers.

Supermicro reported disappointing financial results for the third quarter of fiscal 2025, which ended in March. Revenue increased by 19% to $5.6 billion, gross margin contracted by 2 percentage points, and non-GAAP net income fell 53% to $0.31 per diluted share. The company also cut its full-year guidance for the second time in as many quarters. Revenue is now expected to grow 49% in fiscal 2025, down from the initial forecast that called for 87% growth.

Wall Street expects Supermicro's adjusted earnings to grow by 18% annually through fiscal 2026, which ends in June 2026. That consensus makes the current valuation of 21 times adjusted earnings look reasonable, but analysts may be overestimating the company. Supermicro has missed the consensus earnings estimate by an average of 17% in the last four quarters.

Personally, I doubt the stock will drop 50% (as Mike Ng anticipates), though history suggests that sort of volatility is possible. But investors should limit stock purchases to companies with a durable competitive moat, and I'm not sure Supermicro satisfies that criterion.