BofA Securities has adjusted its price target for Arm Holdings (NASDAQ: ARM) upward to $140 from $135 while retaining a Neutral rating on the stock. That target remains below an analyst high of $201 but sits above the prevailing share price of $123.78. According to the article's data, that market price is judged by a Fair Value framework to be overvalued, placing the company on a Most Overvalued list.
Central to BofA’s revision is a reassessment of Arm’s potential to capture a larger share of the merchant CPU market. The firm now projects Arm could attain roughly 20-25% or more CPU share by calendar year 2030, an increase from a prior expectation of 15-20%.
Those share gains hinge on a strategic shift: Arm is expected to introduce an in-house designed merchant CPU, which departs from its long-standing model based around intellectual-property licensing and recurring royalty income. BofA’s analysis highlights the dramatic difference in addressable market and margin profiles between Arm’s existing IP-licensing business and a move into full silicon.
With full chiplets in play, BofA estimates Arm’s serviceable addressable market could expand by a factor of 30. The firm contrasts per-chip economics at roughly $50-100 for IP licensing versus approximately $1,500-2,000 per chip for full silicon. It also projects a material difference in operating margin potential - noting addressable operating margin could be about 40-45% for IP licensing versus roughly 25-30% for chips, while the company’s current IP-licensing model already posts an elevated gross profit margin of 97.5%.
Those margin and market dynamics appear alongside valuation considerations: the stock trades at a price-to-earnings ratio of 163 under the metrics cited.
BofA’s scenario modeling includes a more modest near-term outcome as well. The firm suggests that even a 3-5% share of the CPU market could lift Arm’s calendar year 2028-2030 earnings per share by an estimated 10-20%.
BofA also flags a countervailing effect tied to current SoftBank licensing revenue. The potential discontinuation of roughly $800 million per year in SoftBank-related licensing sales would erase a portion of the near-term financial upside from chip initiatives, given that chip research and development may represent a one-time cost element versus recurring IP royalties. In the modeling described, achieving the chip-related revenue profile would require approximately $1.2 billion per year in chip sales.
Arm’s recent operating performance is noted as strong: the company reported robust fiscal third-quarter results, beating expectations on both royalties and licensing revenue.
Broker reactions elsewhere in the market have varied. KeyBanc trimmed its price target on Arm to $170 from $200, citing concerns about memory costs. RBC Capital reduced its target to $130 from $140 while maintaining an Outperform rating, pointing to a slower outlook for fourth-quarter royalty growth. Conversely, Susquehanna upgraded Arm to a Positive rating, emphasizing the company’s artificial-intelligence initiatives despite continued headwinds in Mobile and PC markets.
Investor ownership movements in other filings were also highlighted. SoftBank Group reported that Arm and OpenAI now account for 65% of the company’s net asset value in its December quarter results; SoftBank’s net asset value showed a modest decline of 1% between the end of September and February 11. A separate disclosure in NVIDIA’s latest 13F filing showed NVIDIA eliminated its stake in Applied Digital; the filing also indicated NVIDIA no longer holds stakes in Arm Holdings and WeRide Inc.
For readers seeking more extensive coverage, Arm is listed among the more than 1,400 U.S. equities covered by the referenced Pro Research Reports.
Analysis
The BofA update underscores a fundamental trade-off for Arm: its high-margin, recurring-IP business is exceptionally profitable today, but expanding into full silicon could multiply addressable revenue by orders of magnitude even as operating margins migrate toward chip-industry norms. BofA’s higher share assumptions and target reflect that trade-off while acknowledging transitional risks tied to one-time R&D and the potential loss of existing licensing streams.