A Brilliant Business, Just Not Cheap Enough

All figures in US dollars. Microsoft reports in USD with a fiscal year ending 30 June; the tear sheet shows earnings per share in cents (1,364 = $13.64).

Microsoft is one of the world's largest technology companies, operating across three reporting segments: Productivity & Business Processes (Microsoft 365, LinkedIn, Dynamics), Intelligent Cloud (Azure, server products, enterprise services), and More Personal Computing (Windows, gaming including Activision Blizzard, search and devices).

The business model is built on deep integration into corporate and personal computing — operating systems, productivity software, cloud infrastructure, and increasingly artificial intelligence tooling. This breadth is the core of the investment case: Microsoft isn't dependent on any single product or hardware cycle, which makes it unusually resilient compared with more concentrated technology peers.

Microsoft has seen reasonably stable turnover growth over the last several years, with the exception of FY 2023, which only saw a 6.9% revenue rise. FY 2024 saw 15.7% and FY 2025 saw 14.9%. The three years before that were strong too — FY 2020 at 13.6%, FY 2021 at 17.5% as COVID flowed through, and FY 2022 at 18.0%. So the company continues to see good levels of turnover growth.

Six-year financial history ($m unless stated)

As always, the company has amazingly good gross margins — as you'd expect of a software company. They hit 68.8% in FY 2025, which is pretty much where it has been for at least the last four or five years. Pre-tax profitability has been consistently and stably going up, leading reported earnings per share in FY 2025 to $13.64.

The shares in issue have stayed reasonably stable — they don't appear to be doing much in the way of buybacks recently, which is a slight contrast to some peers.

Pre-tax profit growth, year on year

The five-year pre-tax profit compound annual growth rate (FY 2020–FY 2025) is approximately 18.5%. I like the consistency of the turnover growth, although it is ever so slightly lacking. That sounds harsh on a company achieving nearly 14.9% revenue growth — but as a big tech company with so much going on, I might hope for even more. Perhaps at this scale, companies simply struggle to hit greater growth. Earnings per share growth for the last twelve months was 15.6%, which was good and slightly outstripped turnover growth.

Microsoft reports in three segments. The full-year FY 2025 breakdown:

Segment revenue

Intelligent Cloud — driven by Azure — is the growth engine, up 21% and now the second-largest segment. Productivity & Business Processes remains the largest at $120.8 billion but grew more slowly. More Personal Computing grew 20%, boosted by the full-year contribution from the Activision Blizzard acquisition. On profit, Intelligent Cloud and Productivity & Business Processes each generate segment operating margins comfortably above 45%, while Personal Computing runs materially lower (gaming and devices being lower-margin businesses).

FY 2025 capital expenditure of $32 billion was high — about 11.4% of revenue. But for FY 2026, Microsoft has guided towards roughly $190 billion of capital spending — a colossal step up, well above Wall Street estimates, driven partly by soaring memory prices. This is the area I find concerning: the AI capex burden is enormous, and the returns on it are still to be proven. But I do think Microsoft is likely to find opportunities towards its own profitability within these areas.

Capital expenditure

I was interested and surprised that cash and cash equivalents increased during FY 2025 from $18.3 billion to $30.2 billion, while long-term debt fell from $42.6 billion to $14.15 billion. That's a similar story to what we saw in the Apple figures — long-term debt reducing. I doubt that reduction will continue, given the scale of capex ahead.

A key part of my thinking on the probability weighting: I am not particularly expecting things to be outside the norm for Microsoft. They are not dependent on single pieces of hardware or technology. They are integrated into so many different areas of so many people's lives — unlike, say, Apple, which needs to ensure each model it releases is the leading premium device.

Microsoft, as more of a software and operating-system layer, doesn't have that same issue. I think they will be more adaptive to the AI shift: you can integrate almost any AI platform into Microsoft's services. In the shorter term, Apple is very likely to end up tied to a specific provider and focus on that, whereas Microsoft is likely to use whichever is best at any given moment. They are not locked in. I don't have particularly strong hopes for their own home-grown AI, but the lack of lock-in is an advantage.

Assuming a longer-term growth rate of 18% — higher than their existing earnings per share growth, but more in keeping with their two-year and five-year average growth, and reflecting the AI integration tailwinds — with a valuation multiple of 9x and a 5% valuation uplift, I arrive at a valuation of $337.71 per share.

VALUATION SNAPSHOT

This produces a valuation of $337.71 per share against a current price of $427.34 — a 21.0% negative value opportunity. The 26% risk factor is relatively low for a technology company, which reflects my view that Microsoft is unusually resilient and adaptable. The 50% probability reflects the uncertainty around the scale of AI capital expenditure and whether the returns will materialise, even though the underlying business is diversified, deeply embedded, and not reliant on any single product cycle.

• Capex burden: Roughly $190 billion of guided FY 2026 capital spending — a near-sixfold increase on FY 2025. If the AI returns don't materialise, this destroys value on a colossal scale.
• Valuation: At 31.2x trailing earnings, the stock prices in a lot of future growth. Any disappointment in the growth rate or margin would hit the share price hard.
• Growth deceleration: Revenue growth, while strong at 14.9%, is "ever so slightly lacking" for a company of this ambition. At this scale, sustaining high growth becomes progressively harder.
• Debt likely to rise: Long-term debt fell to $14.15 billion in FY 2025, but the scale of capex ahead means this reduction is unlikely to continue — borrowing will probably rise again.
• Home-grown AI uncertainty: Microsoft's own AI models are not a particular strength. The strategy relies on integrating third-party providers — which works while it has access, but is not fully within its control.

• Share price below $340: At or below my valuation, the stock would become genuinely interesting given the quality of the business.
• Evidence of capex payoff: Clear demonstration that the AI infrastructure spend is generating incremental, high-margin revenue would justify a higher multiple and a higher growth assumption.
• Re-acceleration of revenue growth above 18%: If Azure and AI tooling push group growth back above the high-teens, the current price would look more justified.
• Capex discipline: Any signal that the $190 billion guidance is being tempered toward returns rather than raw scale would reduce the key risk.
• Resumption of buybacks: Meaningful share count reduction would support earnings per share growth and signal management confidence in the cash position.