The semiconductor super cycle of booming demand and undersupply over the past two years is coming to an end, forcing chip companies to announce cost savings and reassess their long-term planned investment campaigns.
The average lead time for all semiconductors — the difference between when a chip is ordered and delivered — fell from 26.3 days in September to 25.5 days in October, according to data from Susquehanna International Group (SIG). This is a marked difference from a prolonged period of increasing average chip lead times, which rose from 13.6 days in May 2020 to 27.1 days in May 2022.
Christopher Rolland, a semiconductor analyst at SIG, says that the lead times have come down due to a combination of decreased demand and, to a lesser extent, increased supply.
Rising lead times was tracked by massive investments announced in the semiconductor sector, as chipmakers rushed to build factories to serve expected long-term demand and make use of Western government incentives to boost local chip production capacity, such as the US’s $52.7bn Chips Act. A record $83bn of capital was pledged in 2021 to foreign direct investment (FDI) projects in the semiconductor sector, according to fDi Markets figures.
In the first three quarters of 2022, more than $60bn has been pledged to FDI projects, a figure which increases to $85bn when including US inter-state projects — or domestic investments made by a company based in another US state. One prominent example of such a project is California-based Intel’s plans to invest at least $20bn in a new chipmaking complex outside of Columbus, Ohio.
Mr Rolland says that these planned capital spending plans have had “limited” impact on semiconductor lead times, given that much of this capacity is due to come online in 2023 and beyond. But there are worries that there could be near-term overcapacity in the market over the next 12 to 18 months, as new factories come online when demand is subdued.
Several of the largest chip companies have announced cutbacks and cost savings measures in recent months. US-based Intel, the largest investor in the semiconductor sector by capital expenditure pledges so far in 2022 according to fDi Markets, said it would reduce costs by as much as $10bn by 2025 and layoff an unspecific number of people.
“We are planning for the economic uncertainty to persist into 2023,” Intel’s CEO Pat Gelsinger said on an earnings call on October 27. “It’s just hard to see any points of good news on the horizon.”
While the chip industry is renowned for being cyclical, the most recent boom and bust is unique due to the huge pandemic-induced spike in demand for personal computers, as the world shifted towards to working and learning from home.
“We went from a period of supply shortages to demand declines. It’s kind of an unprecedented change over a short period of time,” said Akash Palkhiwala, the chief financial officer of Qualcomm, a major US mobile phone chipmaker, in a call with analysts on November 2. Qualcomm has announced spending cuts in some areas and hiring freezes amid the gloomy outlook.
Another factor bringing lead times down is the significant inventory building by companies in response to persistent shortages of chips over the past few years. Michael Yang, a senior director of semiconductors at consultancy firm Omdia, notes that there is a nuanced picture across different segments of the industry, with “severe oversupply” in areas such as consumer electronics, while demand remains high for the chips used in electric vehicles and industrial applications.
Despite the expected “dry spell” over the next 12 to 18 months, the outlook for the long-term chip demand remains strong, said Mr Yang. The overall semiconductor industry is expect to almost double in size by 2030 and surpass global sales of $1tn, according to Semi, an industry association.