Seagate has set an example for everyone, and there are still 4 heavyweight earnings reports coming tonight.

US stocks pulled back last night, and there are four important earnings reports worth watching tonight. Let's reiterate one point: at this stage, whether the three major indices fall a little more or a little less is actually less significant. So far, 138 companies in the S&P 500 have reported their Q1 earnings, accounting for nearly 28%. Based on the companies that have already reported, total revenue grew 23.1% year-over-year, profit grew 9.6% year-over-year, and the proportion of EPS beats and revenue beats are both around 76.8%. This means the earnings season actually started off not bad, with no significant collapse in corporate profits.

The problem, however, is that the market is now most critical not of ordinary companies, but of tech giants and the AI supply chain. Last night's decline was mainly due to collective pressure on AI-related stocks. Oracle fell over 4%, CoreWeave fell over 5%, and NVIDIA, AMD, and Broadcom also saw varying degrees of pullback. The underlying reason is not a sudden disappearance of AI demand, but the market beginning to re-examine the commercial closed loop of the OpenAI-related ecosystem.

In the past, AI trading was smooth: model companies expanded, cloud providers increased capital expenditure, chip and computing power companies rose, and the market assumed future demand would definitely absorb these investments. But now that stock prices have reached this level, investors will no longer just listen to the phrase "the future potential is huge." Money has been spent, but will there be revenue to support it in the end? Are the orders sustainable? Will profit margins be eaten up by capital expenditure? These questions need to be answered through earnings reports. So last night's decline was not essentially a rejection of the AI theme, but rather the market starting to price the AI supply chain in tiers.

Another one worth watching is Seagate's earnings report, which gave a completely different signal.

Seagate's Q3 revenue was $3.11 billion, higher than the market expectation of $2.95 billion, a year-over-year increase of 44.1%; adjusted EPS was $4.10, also significantly higher than the expected $3.50; adjusted operating margin was 37.5%, exceeding the market expectation of 34.8%. More crucially, the company's revenue guidance for the next quarter is $3.45 billion, higher than the market expectation of $3.16 billion. After the report, Seagate surged in after-hours trading, and Western Digital, Micron, and SanDisk were also lifted.

The value of this report is not just that Seagate beat expectations, but that it reminds the market: AI infrastructure is not just about GPUs.

Training and inference need chips, but data also needs to be stored. Enterprise AI applications need to enter data centers, and video, images, logs, model parameters, and training data will all bring greater storage demand. In the past, the market understood the AI supply chain too narrowly, mainly focusing on core computing power stocks like NVIDIA, Broadcom, and CoreWeave. But Seagate's report shows that AI construction is spreading to deeper and broader infrastructure segments. This is the most incremental part of last night: expectation-based assets were cut, while realization-based assets were rewarded.

Within the same AI supply chain, companies relying on stories to support their valuations will be questioned by the market about their business models, while companies that can deliver revenue and guidance will be repriced instead. Funds are not refusing to buy AI now, but they no longer want to pay the same premium for all AI concepts.

Tonight is the real main event. Microsoft, Google, Meta, and Amazon are all reporting earnings, with Apple reporting later. The combined size of these companies is too large; they won't just affect their own stock prices but will also directly impact the valuation anchor of the entire Nasdaq and S&P 500.

For Microsoft, focus on whether Azure's cloud business is re-accelerating. In the previous few quarters, Microsoft kept saying AI demand was strong, just limited by capacity constraints. If Azure's growth rate improves tonight, it means demand is not the issue, but rather supply can't keep up. However, if the growth rate remains lackluster, the market will start to suspect that the distance between AI capital expenditure and revenue realization is farther than imagined.

For Google, focus on cloud and profit margin. Its advantage is that search advertising still generates cash flow, making the pressure from AI investment relatively easier to absorb. But what the market wants to see is whether Google Cloud can continue to deliver strong enough growth, and whether AI investment is significantly eroding profits. If the cloud business is strong and margins are stable, Google will be one of the companies most likely to stabilize tech stock sentiment.

Meta's focus is on advertising and CapEx. Its current problem is not whether it's profitable, but whether the market is willing to accept its continued massive investment in AI. As long as the advertising side remains resilient and AI recommendations continue to improve monetization efficiency, funds can tolerate its high spending. But if capital expenditure continues to be raised while revenue doesn't significantly beat expectations, stock price pressure is likely.

Amazon is still about AWS. All the market's anxiety about AI will ultimately fall on cloud providers, because AI computing power ultimately needs to be monetized through cloud revenue. If AWS re-accelerates, it will directly alleviate market concerns about the AI payback cycle. If AWS performance is flat, then last night's doubts about the OpenAI supply chain may continue to spread to cloud and computing power companies.

So in the coming days, earnings reports are more important than the market action itself. The market action is just the outcome; earnings reports are the answer. Seagate has set an example for the market: as long as demand can land on the financial statements, there is still room in the AI infrastructure supply chain. But last night's pullback in AI stocks also shows that where there are only stories and no numbers, volatility will increase significantly in the future.

The earnings reports from several tech giants tonight are the new watershed for this round of the AI rally. After seeing these reports, the market will have a clearer picture of whether AI is continuing to spread or entering a more pronounced phase of internal differentiation.

The above content represents personal views only and does not constitute investment advice. Give Vivian a follow, and I wish everyone that what you buy goes up and what you sell goes down~