Photo from Reuters
Nvidia is facing a situation where it has a high demand for its artificial intelligence chips, but unfortunately, there is no other company in the chipmaking industry that can match its level of success. This issue is becoming evident at the beginning of the second-quarter earnings season. Taiwan Semiconductor Manufacturing, also known as TSMC, reported a nearly 10% decrease in second-quarter revenue compared to the same period last year. This decline marks TSMC's first revenue drop in 16 quarters, which was the most extended period of continuous growth since at least 2000, based on data compiled by S&P Global Market Intelligence.
The decline in revenue was not entirely unexpected, as analysts had already foreseen it due to the ongoing slowdown in the chip industry. This slowdown is mainly attributed to decreased demand for products like personal computers and smartphones.
According to the Semiconductor Industry Association, global chip sales for 2023 up to May have experienced a significant 21% decrease compared to the same period in the previous year. Unfortunately, there doesn't seem to be any indication of a swift recovery on the horizon. TSMC has adjusted its full-year 2023 revenue outlook, now expecting a decline of approximately 10% in U.S. dollar terms from the previous year. This projection is worse than the company's estimate from three months ago, which anticipated a decline in the low- to mid-single-digit range.
TSMC has become the top semiconductor company globally in terms of annual revenue. It operates as a contract manufacturer, producing chips designed internally by tech giants like Apple, Amazon, Microsoft, and Alphabet's Google, as well as other major chip designers such as Nvidia and Advanced Micro Devices. These companies rely on TSMC's manufacturing capabilities since they don't have their own chip fabrication facilities.
As a result, TSMC's financial performance is a reliable indicator of the overall semiconductor industry's well-being, making its latest financial results significant for the entire sector. The impact of TSMC's results was substantial, evident in the PHLX Semiconductor Index, which fell by 3.6% on Thursday. Every company listed on the index experienced losses for the day.
The decline in chip-manufacturing equipment manufacturers was notably significant, mainly because TSMC announced that its capital expenditures for the year would be towards the lower end of the previously estimated range of $32 billion to $36 billion. This could result in a potential decrease of up to 12% compared to the previous year.
On Thursday, the stocks of chip-equipment manufacturers ASML, Applied Materials, KLA, and Lam Research collectively experienced an average decline of approximately 5%.
Chip stocks have performed remarkably well this year. Some of the gains can be attributed to traders speculating on the eventual recovery of the cyclical industry. However, a significant portion of the surge in chip stocks is driven by the hype surrounding generative AI, the technology that fuels chatbots like ChatGPT and relies on high-performance computing and graphics processors.
Microsoft, Google, and Amazon, major players in cloud computing, are in a competitive race to develop and enhance GenAI tools and services. This increasing focus on generative AI is leading to a surge in demand for the required components. Nvidia's impressive revenue forecast during its last quarterly results in May significantly boosted the company's stock value, propelling its market worth above $1 trillion. Between Nvidia's report and TSMC's recent financial results, the PHLX index surged by 22%, which is more than twice the performance of the S&P 500 during that period.
Nvidia's remarkable success does not have a universally positive impact. TSMC stated on Thursday that AI represents only 6% of its current revenue, whereas smartphones account for 33%. The situation for smartphones may continue to be difficult, as Wall Street predicts a 4% decline in Apple's iPhone unit sales for the fiscal year ending in September. This projected drop would mark the first decrease in four years, as per consensus estimates from Visible Alpha.
TSMC anticipates robust expansion in its AI sector, expecting a yearly growth rate of 50% over the next five years. Nonetheless, this growth is hindered by production limitations, particularly in the packaging process where chips are assembled with other components. According to Brian Chin from Stifel, increasing production starts for Nvidia's AI chips fourfold won't yield immediate output gains as the bottleneck is in the restricted capacity for advanced packaging. In essence, even Nvidia cannot completely overcome the broader limitations of the chip industry.