Bitget App
Trade smarter
Buy cryptoMarketsTradeFuturesStocksEarnInstitutionAI & More
Weekly Preview: Micron (MU.US) earnings test the quality of AI infrastructure, OpenAI and White House AI meeting resonance, PCE and Nonfarm Payrolls set the tone for October interest rates

Weekly Preview: Micron (MU.US) earnings test the quality of AI infrastructure, OpenAI and White House AI meeting resonance, PCE and Nonfarm Payrolls set the tone for October interest rates

智通财经智通财经2026/09/28 00:21
Show original
By:智通财经

This week, the market's focus will shift from politics and product launches to financial reports and macroeconomic data.

According to Zhitong Finance APP, last week the market was dominated by two major headlines. The first was Meta (META.US) launching the Muse AI agent and accompanying AI hardware Charm at the Connect conference; these products quickly topped both Apple (AAPL.US) App Store and Google (GOOGL.US) Play Store, making them the most noteworthy new variables in the AI application layer. The second was the state visit of the President of China to the United States, with investors closely assessing its potential impact on the AI trading landscape and the relationship between the two largest economies, the US and China. Despite the flurry of headlines and mixed market forces, US stocks performed steadily overall last week, with the main indices not experiencing dramatic swings from these events.

This week, the market focus will shift from politics and product launches to earnings reports and macroeconomic data. After the US market close on Wednesday Eastern Time, memory chip giant Micron Technology (MU.US) will release its fourth-quarter results—a key point for investors to test whether AI infrastructure plays can continue. Over the past two weeks, AI infrastructure trades have shown volatility, and memory chips have emerged as one of the most prominent bottlenecks in AI computing power expansion. Investors will focus on evidence of sustained strong demand for memory.

In addition, OpenAI is expected to preview GPT-6 Cyber and launch hosted agent products at DevDay in San Francisco on Tuesday; US President Trump, House Speaker Mike Johnson, and technology CEOs are scheduled to meet on September 29 local time to discuss AI. Together, these events form a three-pronged validation for AI trading: from model development, to regulation, to profitability.

On the macro front, this week is a typical “employment and inflation week.” The Federal Reserve’s preferred inflation indicator, PCE, will be released on Wednesday, and the September nonfarm payrolls report will be published on Friday. These will jointly determine market pricing for the Fed's October meeting and even influence policy trajectory through 2026. Additional data like JOLTS job openings, ISM Manufacturing Index, Euro Area CPI, and the Reserve Bank of Australia’s decision will also vie for market attention.

Meta’s “Muse Moment”

Last week, the Meta Connect conference became a core focus for investors. Meta CEO Mark Zuckerberg outlined the company's future vision, with Muse at the core. The social giant's new AI agent quickly took first place on both the Apple App Store and Google Play Store, and Zuckerberg stated that millions already use it. During the conference, Meta’s Chief AI Officer Alexandr Wang demonstrated multiple use cases for Muse, ranging from scheduling meetings and other daily tasks to negotiating better rates for cable services and completing purchases across multiple services—more complex operations.

Meta is also integrating Muse into its next-generation Meta VR glasses, allowing users to converse with the agent’s digital avatar. At the same time, it launched Muse Charm—a handheld dedicated device—so users can interact with Muse without the need for a smartphone or smart glasses. The company stated it will monetize by taking small commissions from transactions completed by the Muse agent.

This week, the market will watch whether Meta can maintain this momentum. For Meta to win over users beyond early adopters and AI enthusiasts, it must convince the public that it takes privacy and security seriously. Reports indicate that Muse has been found to have security vulnerabilities, potentially affecting user data and virtual environment safety. In response, Meta is preparing to enhance risk notifications and implement more prominent warning prompts.

William Blair analyst Ralph Schackart noted that as consumers increasingly trust AI systems to handle sensitive tasks like financial information, communications, calendars, and purchasing decisions, trust could become an important factor for competitive differentiation.

Micron, OpenAI, and the Trump AI Meeting

Micron Technology will release its Q4 earnings after the close on Wednesday. As a leading US manufacturer of DRAM and NAND storage, Micron is considered one of the direct beneficiaries of surging AI memory demand. Investors will look for signs that the AI-driven memory boom can further boost revenue, margins, and long-term demand visibility.

The company previously disclosed that third-quarter revenue reached a record $41.46 billion, a year-on-year increase of 346%; non-GAAP gross margin expanded from 74.9% last quarter to 84.9%. HBM4 has entered massive production for major customer platforms, and the company has also delivered qualification samples to several clients. For FQ4, management guides revenue at $50 billion ±$1 billion, a non-GAAP gross margin of about 86%, and non-GAAP EPS at $31 ±$1, signaling that demand for high-value memory continues to be driven by AI infrastructure needs.

Micron also announced management changes recently: Manish Bhatia was promoted to President and COO and Scott DeBoer to President and CTO, strengthening leadership in capacity expansion and product development stages. Micron stated it has signed 16 strategic customer agreements, with 14 covering an estimated $100 billion minimum cumulative revenue during their contract periods. These agreements span DRAM, HBM, and NAND, covering data centers, consumer, automotive, and industrial markets. Currently, Wall Street maintains a “Strong Buy” consensus rating on the stock.

Also debuting this week alongside Micron’s earnings is the OpenAI developer conference (DevDay 2026). The AI giant is reportedly set to preview a cybersecurity-focused model—GPT-6 Cyber—on Tuesday, as well as launch new tools that help customers deploy the product more securely and automatically. According to multiple sources, GPT-6 Cyber is the fourth cybersecurity model released by OpenAI this year and could officially debut at DevDay, alongside at least a dozen other products. Under the application-based cybersecurity project Daybreak Red, select customers have begun alpha testing GPT-6 Cyber.

The launch comes at a time when the AI sector is facing an extremely complex landscape. Earlier this month, OpenAI CEO Sam Altman joined rivals including the Anthropic CEO in calling for a slowdown in AI development and increased safety measures. Cybersecurity is rapidly becoming one of the fastest-growing AI markets, with technology companies developing tools to counter threats and rein in uncontrolled AI agents. OpenAI had warned that its flagship GPT-6 model Astra occasionally tries to circumvent human oversight; last Thursday, Australian authorities also reported that an OpenAI agent breached a government health data portal in June. These events heighten attention on OpenAI's new security deployment tools.

On the policy front, US President Trump, Republican House Speaker Mike Johnson, and tech CEOs will hold an AI meeting on September 29. Global AI leaders have previously warned about risks posed by AI to humanity and urged governments to cooperate on managing this increasingly powerful technology, but regulatory development struggles to keep up with the rapid evolution of AI. As the two main AI powers, the US and China have taken different regulatory paths: Trump argues the US does not need new AI-specific regulation, while China imposes rules on AI algorithms and training data—including clear labeling of AI-generated content. Trump stated during the UN General Assembly debate that he “rejects any attempt to build a globalist scheme to control AI,” calling the technology amazing but asserting the US will remain cautious.

PCE and Nonfarm Payrolls: Dual Test for Inflation and Employment

On the macroeconomic side, this week’s focus is the PCE inflation print and the September nonfarm payrolls report.

The Fed’s preferred inflation gauge—August’s PCE Price Index—will be released on September 30. Compiled data shows the market expects August PCE to rise by 3.8% year-on-year, accelerating from the previous 3.7%; core PCE (excluding food and energy) is expected to rise from 3.3% to 3.4%. Inflation-adjusted real personal consumption is forecast to increase 0.5% month-on-month, the largest single-month gain in over a year. Should consumption remain robust while inflation reaccelerates, the Fed would find itself in a trickier spot, with a significant challenge in exiting its tightening stance.

On employment, the September nonfarm report is due October 2. In August, US nonfarm payrolls increased by 162,000—about three times the market's expectation of 53,000—setting a high bar for September’s data. Currently, expectations are clearly divided: the consensus is around 90,000 to 100,000, BNP Paribas forecasts 90,000, while Bank of America is more cautious at just 60,000. Bank of America economist Aditya Bhave noted that if September’s number comes in below expectations, it should not be surprising since abnormally favorable seasonal factors in August could lead to a pullback. The unemployment rate is expected to stay at 4.1%, the lowest in a year, indicating that while hiring momentum is slowing, the labor market has not deteriorated rapidly.

There are also optimistic views. BNP Paribas and senior US economist Andrew Husby believe that as the Fed’s tightening cycle advances, momentum in the US economy could drive unemployment further down by early 2027, possibly resulting in the market under-pricing the risk of falling unemployment.

But there is a disconnect between consumer sentiment and employment data. The University of Michigan’s September consumer confidence index fell from 51.7 in August to 48.1, slightly above the market's expectation of 47.5, but consumers’ expectations for their personal finances dropped by about 10%. The survey’s director Joanne Hsu noted that high fuel prices and renewed trade disputes sharply weakened short-term business outlooks.

From interest rate futures pricing, following the recent rate hike, the market sees a roughly 70% chance the Fed hikes by another 25 basis points at its October meeting. If PCE and nonfarm data land in line with expectations—moderate inflation rebound, job growth slows but does not collapse—the case for continued tightening will be reinforced. Other US data this week include the September 29 JOLTS job openings and October 1 ISM Manufacturing Index, both expected to confirm US economic expansion and inflation pressures; some observers warn this could lead to heightened volatility in global financial markets.

Internationally, Eurozone September CPI will be released on October 2 and is expected to rise to 3.7% year-on-year—the highest in three years—with energy prices the main driver. This has raised market speculation about a third European Central Bank rate hike this year. The Reserve Bank of Australia will hold a monetary policy meeting on September 29, with all four major Australian banks expecting a 25bp hike to 4.6%, the fourth this year. Although Australia’s unemployment rate rose to 4.6% in August, indicating some cooling in the labor market, the global rate environment and inflation pressures still support continued tightening by RBA. Some analysts expect another hike in November, which would bring the benchmark rate to its highest since late 2008 if five hikes materialize this year. Indonesia and South Korea will release CPI on October 1 and 2, respectively. Ongoing Middle East conflict boosting oil prices and El Niño driving up food prices will be key inflation variables in Asia.

0
0

Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.

Understand the market, then trade.
Bitget offers one-stop trading for cryptocurrencies, stocks, and gold.
Trade now!

You may also like

Small models "steal" demand for cloud computing power, trillion-dollar data center capital expenditure faces a test

A Jefferies report points out that improvements in small language model performance and declining costs may lead enterprises to shift AI deployment from the cloud to localized solutions, impacting investments in hyperscale data centers. With falling server utilization rates and a mismatch between the massive capital expenditures of tech giants and actual computing power demand, some data centers risk becoming "stranded assets." Although AI demand persists, the scale and returns of centralized computing power are facing reassessment.

华尔街见闻•2026/09/28 09:51