Microsoft's first quarter financial report for fiscal year 2025: Revenue was $65.585 billion, a 16% year-on-year increase, also a 16% year-on-year increase excluding the impact of exchange rate changes; Net income was $24.667 billion, an 11% year-on-year increase, also a 10% year-on-year increase excluding the impact of exchange rate changes; Diluted earnings per share were $3.30, a 10% year-on-year increase, also a 10% year-on-year increase excluding the impact of exchange rate changes (Note: Microsoft's fiscal year is not aligned with the calendar year).
After the financial report was released, Microsoft CEO Satya Nadella, Executive Vice President and CFO Amy Hood, Chief Accountant Alice Jolla, and Deputy General Counsel Keith Dolliver and other company executives attended the subsequent earnings conference call, interpreting the key points of the financial report and answering analysts' questions.
The following is a transcript of the conference call:
Morgan Stanley analyst Keith Weiss: The expansion of generative artificial intelligence-related functions, the speed of innovation, and significant opportunities are the most exciting situations I have seen in the software industry in my 25 years of experience. From the conversations of the management during this conference call, it seems that you also have the same feeling. However, in my conversations with investors, I heard two concerns related to factors constraining the future development of generative artificial intelligence technology.
The first question is, what internal constraints or obstacles does Microsoft face when making relevant investments, especially in funding for the next generations of basic models, as we estimate the scale of related funds could be as high as hundreds of billions or even over a trillion dollars. Another question is, what external constraints has Microsoft encountered in expanding capacity to meet demand and seize opportunities, especially in sustainable power supply for building new data centers.
Satya Nadella: Regarding the first question, when we consider capital expenditures for tasks like artificial intelligence training – this should be the approximate meaning you mentioned – expenses in this area may be limited by the revenue generated from the inference business. As in the past, we will allocate capital based on signals of market demand to build our cloud services, then we use this to forecast demand and make additional constructions accordingly. The situation is the same for the training business. As we build next-generation, more powerful models, our products themselves will drive more inference demand, so even though investments in some periods are very substantial, overall, the investment pace will return to normal.
I believe the best way to think about this issue is the effectiveness of Moore's Law in chip and system development, so it's not just about computing, but also about computing efficiency, data, and algorithms. Everyone hopes to stay on that curve, updating devices each year according to Moore's Law and then effectively depreciating them over their lifecycle. Ultimately, inference demand will determine how much we invest in training because I think fundamentally, the ultimate limiting factor must be demand.
Regarding the external factors you mentioned, we have obviously encountered many external constraints because market demand has surged very rapidly. Consider that the most popular artificial intelligence products of this generation all use our cloud services, whether it's ChatGPT, Copilot, GitHub Copilot, or Tax Copilot, these common products in the market are all within our ecosystem or in close proximity. Therefore, we face a series of constraints, data centers cannot be built overnight, so there are issues related to data centers, including power supply, etc., but these are short-term limitations. For example, in the second quarter, some issues related to demand or capacity to meet customer needs are actually caused by external, third-party factors, and we are gradually addressing these issues.
But looking at the long term, we do need an effective electrical supply, we need datacenters, some of these issues will take longer to resolve, the good news is that in the second half of this fiscal year, some supply-demand will achieve a perfect match.
Jeffrey analyst Brent Thill: Amy, it's great to hear that the Azure business will accelerate growth again in the second half of the year. I guess many people have this question, the growth rate of 34% in the first quarter may drop to just over 30% in the second quarter, I know this may be related to changes in the growth base, but apart from the high growth base in the second quarter of last year, in the outlook for the second quarter, has the company also considered other factors that may cause the growth to slow down?
Amy Hood: Let me reiterate some of the points I made before, and then integrate them slightly to answer your question. In the first quarter, we achieved a growth rate of 34%, while we expected 33%, the excess was mainly due to some revenue recognition reasons, I will look at this issue from the perspective of pure consumer and AI. We expect a drop of one or two points, with a significant reason being the supply delays mentioned by me and Satya, in terms of basic consumer growth, it is actually stable from the first quarter to the second quarter.
Regarding the details you mentioned, there will definitely be some fluctuations, but we are confident because in the second half of this fiscal year our supply will increase significantly, especially in terms of AI, better matching of supply and demand will be achieved, and the accelerated growth we talked about in the second half of the year will also occur. I also take this opportunity to mention that when it comes to the use of AI workloads, there is always a tendency to think that graphic processing units (GPUs) are the key issue, but the collaboration between GPUs and central processing units (CPUs) is also very important, so this is also a factor for the acceleration in the second half of the year.
Bernstein analyst Mark Moerdler: Investors are obviously very concerned about the growth of capital expenditures and the direction of expenditures. My understanding is that half of capital expenditures are long-term, so I would like to ask management to talk about their views on the growth of capital expenditures? Will Microsoft's capital expenditures return to the traditional model, where capital expenditure growth is basically synchronous or slightly slower than cloud business revenue? If so, can you provide a rough timeline, for example, by next year at some point will we have enough facilities investment in use?
Amy Hood: Looking back over the past decade on the company's pursuit of cloud business transformation, it may be quite useful for understanding the issues you mentioned. In the early stages, you see the work we have done, including what we are currently doing, which is constantly building facilities to meet market demand. Unlike the cloud business transformation, due to the nature of demand, we are currently expanding globally simultaneously, rather than sequentially. And as long as we continue to see demand growth, the growth of capital expenditures will slow down, but revenue growth will accelerate.
As you said, over time, the growth rates of these two will get closer and closer, and the level of growth actually depends entirely on the speed of adoption of new technologies. As Satya mentioned, some of the spending will be used to build the next training infrastructure, but will not be included in costs, and will be included in operating expenses. Overall, this is a balanced and reasonable way of capital expenditure, as in the company's previous development cycle, the two will get closer and closer.
UBS Group analyst Karl Keirstead: I would like to ask Satya and Amy a question about OpenAI. Since Microsoft announced its investment in the company three months ago, investors have been inundated with a lot of media reports about OpenAI and Microsoft. I would like the management to elaborate on the relationship between the two companies. We have all received some signals from various channels that perhaps Microsoft wants to diversify to some extent at the model level and provide more choices for customers. Also, in terms of financial numbers, Amy, how is Microsoft helping OpenAI realize its expansion plans and the need for capital expenditures? And how is the company addressing the impact on other revenue items that you just mentioned?
Satya Nadella: The partnership between OpenAI and Microsoft has been very beneficial for both parties. After all, our investment in the company four to five years ago was a true bet on innovation, providing support for one of the most valuable private companies in today's market. This investment has brought great achievements to Microsoft and tremendous success to OpenAI.
Based on this foundation, Microsoft has provided world-class infrastructure for OpenAI. Utilizing this infrastructure, OpenAI continues to innovate in terms of models. We are also doing some training-related work at the model level, as well as innovating in building small models and all product innovations. Products like GitHub Copilot (an AI programming assistant), DAX Copilot (a medical AI assistant), or M365 Copilot (data and AI assistant tools) have strengthened my belief in OpenAI and the work they are doing, and there will be more outstanding innovations in the future.
As investors, we are very pleased with Microsoft's involvement in OpenAI's development, especially because in such a partnership, both parties have always maintained dialogue to ensure mutual success. This means that we need to drive each other to do more and seize this rare opportunity. This is our collaboration plan, and we intend to continue developing the relationship between both parties based on this.
(Updating...)
