Arm Holdings (ARM -3.34%) stock declined 3.3% on Thursday, following the leading central processing unit (CPU) chip designer's release on the prior afternoon of its report for the third quarter of its fiscal year 2025 (ended Dec. 31, 2024).

Let's look at the Q3 report and then the stock's valuation to answer this question: Is Arm stock a buy?

Why Arm stock dropped after the earnings release

Arm stock's modest post-earnings-release decline wasn't related to the quarter's results, as revenue and earnings both handily beat Wall Street's expectations. Rather, guidance was the culprit. Investors were not satisfied that Q4 guidance "only" met analyst consensus estimates for revenue and earnings. In 2025, Arm stock had surged nearly 41% through Wednesday. So many investors had very high expectations and were probably expecting Q4 guidance to exceed Wall Street's estimates.

Arm stock's big run-up in 2025 is in part due to investor optimism about the company being a key technology partner in Stargate, as announced on Jan. 21. Stargate is a private $500 billion artificial intelligence (AI) infrastructure project that's a joint venture (JV) between Oracle, SoftBank Group, and OpenAI, best known for its ChatGPT chatbot.

Arm's key quarterly numbers

Data source: Arm Holdings. GAAP = generally accepted accounting principles. Fiscal Q3 2025 ended Dec. 31, 2024.

Investors should focus on the adjusted numbers, which exclude one-time items. Wall Street was looking for adjusted EPS of $0.34 on revenue of $948 million, so Arm easily exceeded both expectations. It also surpassed its own guidance, which was for adjusted EPS of $0.32 to $0.36 on revenue of $920 million to $970 million.

The company generated $423 million in cash running its operations during the quarter, up 36% year over year. On an adjusted basis, free cash flow rose 39% to $349 million. Arm ended the quarter with cash, cash equivalents, and short-term investments of $2.67 billion, up 11% year over year.

Revenue breakdown

Data source: Arm Holdings. YOY = year over year.

Royalty revenue was a quarterly record, while license revenue was better than management had expected.

Licensing demand "remains strong as our partners make long-term commitments to more, and more powerful, energy-efficient Arm technology, including [for] AI training and inferencing," the company said in its shareholder letter. Investors should keep in mind that year-over-year licensing revenue growth will be "lumpy" from quarter to quarter due to the timing of larger licensing deals.

Royalty revenue growth was driven by continued adoption of the company's newest architecture, Armv9, which generally has a higher royalty rate than its predecessor; the ramp-up of its computer subsystems (CSS); improvements in the internet of things (IoT) devices market; and increased usage of Arm-based chips in AI-enabled data centers.

Guidance for Q4 and full-year fiscal 2025

Fiscal fourth-quarter guidance:

• Revenue of $1.175 billion to $1.275 billion (midpoint $1.225 billion), which equates to growth of 27% to 37% year over year.
• Adjusted EPS of $0.48 to $0.56 (midpoint $0.52), or 33% to 55% growth.

Going into the report, Wall Street had been modeling for Q4 revenue of $1.22 billion and adjusted EPS of $0.52, so Arm's outlook at both midpoints was in line with expectations.

For the full fiscal year, Arm narrowed its guidance ranges for both the top and bottom lines, which had the effect of slightly raising both midpoints.

Data source: Arm Holdings. YOY = year over year. *Calculations by author. Fiscal 2025 ends in late March.

In short, Arm turned in a very good Q3 report, including strong guidance for Q4.

So, is Arm stock a buy?

Arm stock remains highly valued. It is priced at 82.1 times Arm's projected (by Wall Street) earnings for fiscal 2026, which begins in April 2025. However, Arm consistently beats the analyst consensus earnings estimate, which means this forward price-to-earnings (P/E) ratio will probably prove to be inflated.

So, in my view, what investors should do in cases like this is tweak the forward P/E that's based on Wall Street's estimates. A decent assumption is that Wall Street's consensus estimate for Arm's fiscal 2026 earnings will prove to be too low by 16%, as has been the case, on average, for the last four quarters. In this scenario, Arm's forward P/E is 70.7.

Yes, 70.7 is still high, though it's more palatable than 82.1.

Arm is a high-quality stock, as the company has a strong balance sheet and a great business model, which generates a long-tailed royalty revenue stream from license agreements inked many years back. So, Arm stock will probably always be more highly valued based on commonly used valuation metrics than those of many of its peers.

For folks who have long-term investing horizons, it would seem reasonable to begin dollar-cost averaging your way into your full investment position. As an example, if you wanted to invest $2,000 in Arm stock, you could buy $500 worth of stock each quarter for one year. By dollar-cost averaging, you will avoid investing your entire sum at a peak price in the event of a significant decline.

For investors hesitant to start buying Arm stock, it's a good watch list candidate. The stock has significant long-term growth potential, as the company has several ways in which it is benefiting from the AI revolution.