Stifel upgraded Microsoft Group from Hold to Buy and raised its price target to $575.00 from $530.00, expressing growing confidence in the company's ability to maintain its mid-to-high-percentage-second revenue growth.
Analysts led by Brad Ryback said Microsoft had clearly passed the tipping point following its June quarter results, citing continued momentum in Azure, a slowdown in investment in big language model research, and an accelerating contribution from OpenAI as reasons for optimism in the second half of the year.
Azure benefited from a significant acceleration in the June quarter, driven by capacity gains resulting from efficiency improvements, along with a larger contribution from OpenAI. Stifel anticipates further growth in Azure of 200 to 300 basis points, as Microsoft achieves a more sustainable pace of efficiency across its entire chip, model, and software ecosystem, noting that management has softened its tone regarding capacity constraints.
Analysts cited comments from Chief Financial Officer Amy Hood, who said that Microsoft had reduced setup-to-operation times — the time between powering up devices and getting them ready to generate revenue — by more than 50% over the past year, accelerating the conversion of capacity into revenue.
Copilot reached approximately 30 million seats in the fourth quarter, an increase of 10 million seats compared to the previous quarter, while GitHub's shift to a consumption-based pricing model provided further impetus to growth. Analysts noted that these positive factors should offset the slowdown in seat growth as Microsoft 365 approaches the 500 million seat mark, keeping growth at mid-2020 levels for at least several years.
Stifel acknowledged that her previous concerns about margin pressures were overly negative, citing Azure's efficiency gains, the cancellation of revenue share payments to OpenAI following the April contract review, and the extension of the asset's productive life to 25 years instead of 15.
Analysts also noted that Microsoft's need for heavy investment in developing its own large language models is diminishing as open-weighted models deliver similar results at a lower cost. Capital expenditure is expected to remain high due to AI infrastructure supply constraints, but the company anticipates achieving leading levels of operational efficiency and positive free cash flow in fiscal year 2027, reflecting data center space limitations rather than a spending cap.
