《Qiu Sheng Yi Chou》With the U.S. restarting its interest rate hikes and market expectations of a significant increase in short-term U.S. Treasury issuance, short-term U.S. bond yields have recently risen more sharply than long-term yields. The yield spread between two-year and ten-year U.S. Treasuries has narrowed from over 70 basis points at the beginning of the year to just 21 basis points now. The U.S. yield curve is expected to flatten further. In such a high-interest-rate environment, the valuation expansion potential for U.S. stocks will be constrained, and equity market funds are increasingly favoring high-growth stocks.
Meanwhile, after Meta launched its personal AI assistant Muse, it rose to the top of the U.S. free app download charts within less than two weeks, triggering a sharp rally in Meta (US.META) and various semiconductor stocks. Unlike previous AI chat applications primarily based on conversational Q&A, AI assistants (Agents) can think independently, search for information, and even handle tasks for users. OpenAI's Apps in ChatGPT, launched at the end of last year, represented an early form of AI assistants, enabling operation of third-party applications such as Spotify and Figma within ChatGPT via conversational interfaces. Later, xAI launched the AI assistant Grok Bot in August this year, which attracted over 400,000 weekly active users within a month. As for Meta's Muse, it achieved more than 83,000 downloads within two weeks. According to information from the X platform, users have shared experiences using Muse to cancel insurance policies—Muse signed the cancellation forms via the software DocuSign when applying for policy termination with the insurer. Other users reported using Muse to apply for residential fiber broadband services, during which Muse not only searched for cost-effective service packages but also negotiated with service providers, saving the user $1,920 on internet services over the next two years.
What surprises the investment market about Muse is not only the user experience but also its ability to connect Meta's AI ecosystem—including AI models, proprietary social platform data, in-house data center computing, and proprietary user application scenarios—thereby unlocking value. Although Meta's AI model Muse Spark 1.3 does not match the capabilities of other cutting-edge models, it still delivers practical use cases, marking a turning point for consumer-grade personal AI assistants. However, the author believes the real test for Meta lies in whether the initial market response can be converted into long-term revenue. Moreover, rumors suggest OpenAI is accelerating the launch of its personal AI assistant, while the startup Instinct's personal AI assistant has already attracted 100,000 users. Intensifying market competition is expected, benefiting hardware demand for devices targeting AI assistant applications, including AI computing chips, CPUs, and memory.
Although many tech stocks have rebounded sharply recently, both the Nasdaq and semiconductor indices remain at reasonable valuation levels. The Philadelphia Semiconductor Index's forecast P/E ratio is around 20 times, still far below its recent peak of 28 times. Referring to the S&P 500 index components, only 28% of stocks are currently trading above their 20-day moving average, indicating that market participation remains low. The author believes investors can hold onto quality tech and semiconductor stocks for now. 《Wah Shing Securities Senior Analyst Qiu Ziwei》
*The author does not hold the aforementioned stocks
*Articles published in 《Economic Times》, whether signed or unsigned, represent the personal opinions of the authors and do not reflect the stance of 《Economic Times》. 《Economic Times》 serves as a platform providing free speech.
Meanwhile, after Meta launched its personal AI assistant Muse, it rose to the top of the U.S. free app download charts within less than two weeks, triggering a sharp rally in Meta (US.META) and various semiconductor stocks. Unlike previous AI chat applications primarily based on conversational Q&A, AI assistants (Agents) can think independently, search for information, and even handle tasks for users. OpenAI's Apps in ChatGPT, launched at the end of last year, represented an early form of AI assistants, enabling operation of third-party applications such as Spotify and Figma within ChatGPT via conversational interfaces. Later, xAI launched the AI assistant Grok Bot in August this year, which attracted over 400,000 weekly active users within a month. As for Meta's Muse, it achieved more than 83,000 downloads within two weeks. According to information from the X platform, users have shared experiences using Muse to cancel insurance policies—Muse signed the cancellation forms via the software DocuSign when applying for policy termination with the insurer. Other users reported using Muse to apply for residential fiber broadband services, during which Muse not only searched for cost-effective service packages but also negotiated with service providers, saving the user $1,920 on internet services over the next two years.
What surprises the investment market about Muse is not only the user experience but also its ability to connect Meta's AI ecosystem—including AI models, proprietary social platform data, in-house data center computing, and proprietary user application scenarios—thereby unlocking value. Although Meta's AI model Muse Spark 1.3 does not match the capabilities of other cutting-edge models, it still delivers practical use cases, marking a turning point for consumer-grade personal AI assistants. However, the author believes the real test for Meta lies in whether the initial market response can be converted into long-term revenue. Moreover, rumors suggest OpenAI is accelerating the launch of its personal AI assistant, while the startup Instinct's personal AI assistant has already attracted 100,000 users. Intensifying market competition is expected, benefiting hardware demand for devices targeting AI assistant applications, including AI computing chips, CPUs, and memory.
Although many tech stocks have rebounded sharply recently, both the Nasdaq and semiconductor indices remain at reasonable valuation levels. The Philadelphia Semiconductor Index's forecast P/E ratio is around 20 times, still far below its recent peak of 28 times. Referring to the S&P 500 index components, only 28% of stocks are currently trading above their 20-day moving average, indicating that market participation remains low. The author believes investors can hold onto quality tech and semiconductor stocks for now. 《Wah Shing Securities Senior Analyst Qiu Ziwei》
*The author does not hold the aforementioned stocks
*Articles published in 《Economic Times》, whether signed or unsigned, represent the personal opinions of the authors and do not reflect the stance of 《Economic Times》. 《Economic Times》 serves as a platform providing free speech.
