ReutersMSFT.OQ TMTSoftware Valuation & Risks Putting margins and returns under themicroscope: Azure Edition Brad ZelnickResearch Analyst+1-212-250-8563 Demand for AI infrastructure has drivena real inflectionin public cloud revenueand backlog growth, enabling long-standing leaders to reaccelerate off anincreasingly large base and creating an opening for new Neoclouds to enter whathad been largely an oligopoly market at significant scale. At the same time,however, margins across the industry have come under increased pressure fromrapidmix shift to what today is generally lower margin,less IP-rich AIinfrastructure services and the upfront costs required to massively scale newcapacity in an effort to meet demand. This is an issue we see management teamsacross the space consistently talking about and working through, and a difficultone for the investment community to wrestle with as we try to dissect theeconomics of these increasingly large AI businesses within the broader publiccloud market. To put it bluntly, despite all the important questions we have aboutthe long-term competitive dynamics and structure of the industry,we often stillfind ourselves in conversations about whether contracts already generatingrevenue and/or booked in backlog actually have positive economic returns as abasic starting point. Bhavin Shah, CFAResearch Analyst+1-212-250-6775 Daniel KnauffResearch Associate+1-415-2622041 Yash KejriwalResearch Associate+1-212-250-1203 We are planting our flag and affirming that current AI businessesacross the cloudproviders we coverareindeed value creative and based on current economics inthe market constitute a sustainable commercial foundation. Our view is based ona detailed analysis of public filings, financial statements, presentations andcommentary from companies across the industry that lead us to believe that thosewith more bearish views may be under-appreciating the impact of upfront coststo scale multi-year revenue generating cloud capacity, the significant OpExleverage in these businesses, and the natural risk mitigationembedded in manylong-term contracts. Following our prior reports onCoreWeave&Oracle, this third in our series focuseson Microsoft, and its Azure business in particular, where we dive into margins, mixand returns.Our conclusion is that while Azure gross margins appear to havecontracted sharply in recent years largely on mix dynamics, a majority of thisheadwind has been offset by OpEx leverage within the Intelligent Cloud (IC)segment. We see some of these pressures moderating ahead, enabling Azuregross margins to stabilize over FY27/28 and IC segment operating margins tomove gradually higher. Furthermore, we find company-level ROIC (and ROIIC),albeit somewhat below historical levels, has remained healthy amid the heavy AI-driven investment cycle in recent years. This deep dive gives us incremental confidence in Microsoft and its ability to sustain and expand company-leveloperating margins and continue compounding robust, value creative growth inthe years ahead. We maintain our Buy rating and $550 target price. The Setup & Key Takes Microsoft has been no exception to investor scrutiny around profitability of its AIcloud infrastructure business amid surging CapEx. Admittedly, this has beenrelatively less of a concern given the company's high margin core businesses andmassive embedded operating leverage, but the narrative here follows a similarpattern-AI driven growth pressuring margins that are further obscured near-term by upfront (pre-revenue) costs associated with an increasing scale of newcapacityadds and fading depreciation benefits from prior-year useful lifeaccounting changes. It's clear to us that Microsoft's management remains keenlyfocused on meeting unprecedented demand in the current capacity-constrainedenvironment, upon whichit isconfident in then capturing efficiencies overtime. Inthis report,we:(1)deconstruct the Intelligent Cloud (IC) segment within whichAzure sits to estimate current Azure gross margins;(2)frame our view on thecurrent state of Azure AI Services gross margins and the path to stabilization inoverall Azure margins;(3)isolate the impact of disciplined OpEx management todemonstrate the durability of the IC segment's operating model; and(4)look atreturns as the business mix becomes increasingly concentrated in Azure & AI. Ourkey takeaways are: •Azure gross margins could trough in the low-50s in FY27/28after havingtrended down from what we estimate with a high degree of confidencewere near 60% in FY24. This contraction has been driven by mix shifttoward lower-margin AI revenue, drag from pre-revenue scaling costsand the y/y impact of prior useful life accounting changes, each of whichwe think should ease as a headwind ahead and along with ongoing like-for-like engineering efficiency gains help put in a floor on margins nearcurrent levels. Off this, we see potential upside / opportunity for marginsto inflect back higher into the mid-50s if