Even in industry known for their roller-coaster cycles, chipmakers are bracing for particularly major shifts in the coming months as a record surge in sales threatens to give way to the worst decline in a decade or more.
During the pandemic, the semiconductor market has seen a huge surge in orders, sending sales and stock prices to new highs and sparking a global scramble to find enough supply. There was hope in some quarters that the boom could continue for a few more years without a painful pullback, but now chipmakers are facing a familiar problem: growing inventories and shrinking demand.
This dilemma is as old as the age of computing. It takes years to build a chip factory, and they don’t always come online when you need them most. In the last few years, the problem has been a lack of supply. As recently as this quarter, auto makers and some other customers complained that they still couldn’t get enough electronic components.
But the fortunes of the largest chip makers quickly turned. Companies such as Nvidia report an annual decline of more than 40% in their core business, while Micron Technology warns that demand in many areas is rapidly declining.
The betrayal of the semiconductor cycle was brought to the fore when US President Joe Biden signed the US$52 billion Chip and Science Act to subsidize domestic manufacturing — the same day that Micron, the largest US memory chip maker, told investors that demand fades.
“It’s kind of dark humor,” said Sanford C Bernstein analyst Stacey Rasgon. “Policymakers are going to find out how quickly the deficit can be closed as industry turns around.”
PC makers, some of the biggest buyers of chips, were the harbingers of the dark times. Desktop processor shipments fell to their lowest level in nearly three decades in the second quarter, according to Mercury Research. Total processor shipments experienced their biggest year-over-year decline since roughly 1984.
Excruciating hangover
It’s a painful hangover after the pandemic lockdown, when the work-from-home trend has fueled demand for PCs and other devices. Chipmakers were scrambling to keep up with the flow of orders, and disruptions in the supply chain made customers even more desperate. Electronics manufacturers were willing to buy the chips at any price they could afford.
Now consumers are cutting back on bulk purchases, and chip shoppers are following suit. This has created what the industry calls an “inventory correction.” The last such downturn was in 2019, and it usually doesn’t last long.
But it is expected to be especially pronounced due to the weakening of the global economy. If the inventory correction occurs at the same time as the economy slips into recession, the industry will not experience the rapid recovery seen after the last downturn.
“It’s going to be a bad downturn,” said Gus Richard, an analyst at Northland Securities.
Christopher Daney, an analyst at Citigroup, expects the industry’s decline to be the worst in a decade, possibly two. Every company and every chip category is likely to suffer, he said.
One unusual factor this time around is the widespread push by governments to subsidize new plants and equipment, from the US and Europe to China and Japan. Companies such as Intel lobbied for the chip legislation, arguing that the US should be more competitive with Asian manufacturers. Now they are ready to start adding new capacity during periods of fluctuating demand.
According to Semi, the chip industry association, 24 new construction projects for large factories, known as fabs, will begin in 2022. That’s well above the average of 20 that Semi has tracked since 2014. Total hardware spending will reach $117.5 billion in 2022, up 15% from the previous industry record set in 2021. Next year, those costs will increase to $120.8. – billion, predicts Semi.
“It used to be a competition between companies,” Richard said. “Now it’s a competition between countries because of strategic importance. There is a race between China and the United States.”
The chip business is becoming increasingly volatile due to huge upfront costs. The plants, costing up to $20 billion, must run 24 hours a day to be profitable for several years before they become obsolete. The scale required for this kind of investment has reduced the number of leading-edge technology companies to less than five. And only three, Samsung Electronics, TSMC and Intel, account for the majority of production.
Read: Intel to raise chip prices – in some cases by more than 20%.
These companies built their dominance by understanding the economics of the industry better than their competitors. They added production lines at the right time and made their supply chains as efficient as possible.
But the drive to ramp up chip production in the US and Europe, which is an alternative to Asian production, could hamper this drive for greater efficiency.
The industry is “effectively creating duplicate supply chains in the U.S. and Europe,” Fitch Ratings analyst Jason Pompei said. “This transition will result in short-lived periods of increased earnings and cash flow volatility, especially compared to the industry’s increased efficiency over recent decades.”
Read: TSMC sales up 44% – and that’s before demand for iPhone 14 chips
In the near term, the risk is “overinvestment in production capacity, leading to an economic downturn,” he said.
Chipmakers remain optimistic about long-term demand. Executives still expect total industry revenue to reach $1 trillion by the end of the decade. This means that their massive factory construction may be worth it.
And in the end, nobody really knows what’s going to happen, Bernstein’s Rasgon said. This is the story of the chip industry. “Everyone is very bad at forecasting demand,” he said. “They’re too bullish, then too bearish.” — (c) 2022 Bloomberg LP
