Advertisement

Analysis: How the semiconductor shortage could last into 2024

Experts differ on when the microchip shortage will end as companies work to redesign products and rethink procurement strategies.

Among all the supply shocks sparked by the pandemic, the shortage of semiconductor chips is among the most widespread and the hardest to resolve.

Factory shutdowns and leaps in demand have driven up prices and lead times for silicon semiconductor chips, which are the digital brains of everything from automobiles to laptops and healthcare devices. It's a phenomenon that's affecting businesses far beyond the traditional tech sector.

"The common perception is it's just about electronics, laptops, and computers," said Mohit Sharma, ACMA, CGMA, a procurement and supply chain expert based in India who advises Fortune 500 companies.

But he added: "This shortage is a result of the overall digital acceleration that was on the way before COVID hit. You cannot even imagine — it has been across industries."

The most acute problems materialised in 2020. Since then, governments have started to pour billions of dollars into solutions — but shortages may still persist late into 2023 or 2024. It's an example of the long-ranging supply effects of the pandemic, and it's one that will have a direct effect on nearly every business.

Sharma laid out the causes, effects, and solutions to this problem, along with another expert, Lincoln Clark, the US-based head of KPMG's global semiconductor practice, in separate interviews with FM magazine.

Here's what finance needs to know.

The causes of the supply chain shock

Sharma and Clark agreed that the initial silicon supply shock was a result of whiplash changes in supply and demand amid the pandemic, along with the complex and slow-moving nature of the semiconductor industry.

"It became very visible to everybody at some point in 2020 just how important semiconductors are to many different things," Clark said. "The pandemic was clearly the shock to the system to make that clear."

As with many industries, the initial lockdowns drove production and shipping to a standstill. At the same time, consumers were buying televisions, computers, phones, and other digital devices that rely on chips.

"There was a phenomenal boost on the [consumer device] side, but nobody realised the scale," Sharma said. The "veil was lifted" and the true scope of the problem was revealed later, he said, when demand for another product began returning in late 2020: automobiles.

Automobile makers in recent years have come to use huge numbers of microchips, which are used for everything from camera systems to power steering, Sharma said. But as automakers moved to restore higher production levels and meet rising demand after the initial lockdowns, they found fewer supplies of semiconductors were available.

"The capacity at the foundries," Clark said, "had been absorbed by the other COVID demand driver — work from home."

In other words, automotive manufacturers found themselves in competition for chips with the tech industry. The tech sector needs chips for both consumer devices and in large numbers for data centres.

The effects of the shortage — and how companies are responding

Two years on, higher prices and long lead times persist. The cost of semiconductors imported into the US increased 8% in July year-over-year, according to the US Bureau of Labor Statistics. Russia's invasion of Ukraine may also add to costs because Ukraine is a major supplier of neon and Russia provides palladium, both crucial materials for semiconductors, although industry experts remain unsure whether the effect will be significant.

Any increases to costs are likely to filter downstream, resulting in continued growth of prices for devices, vehicles, and even the cloud services that run on these chips — although the picture is blurred by the threat of a recession and a drop in demand.

Companies along the semiconductor supply chain have tried a variety of strategies to navigate the ongoing shortages.

Some sellers of goods that require chips are trying to reduce their reliance by redesigning their products. "What kind of re-engineering can you perform?" Sharma asked. For example, he said, automakers are trying to leave certain chip-reliant features as add-ons, rather than making them the default.

Others are recycling older chips from consumer devices and other sources. "You are relying on retrofitting, reshoring, [and] reworking," Sharma said.

Companies that use chips also are trying to create more robust supply chains, Clark said.

"Procurement is focused on, 'How do we avoid ourselves being in the situation that we may feel we've been in for a couple of years?'" Clark said. That includes signing multi-year procurement deals, prepaying for supply, and creating geographic diversity in suppliers.

And that same kind of strategic thinking also applies for companies that purchase, sell, or rely upon silicon-based devices. Sharma said company leaders should look further up their supply chains, ensuring that they understand how rising prices and long waits might affect their business.

"Invest into demand forecasting, supply chain analytics, design of the supply chain," Sharma said. "No one is invincible."

Even companies that simply use consumer-grade devices like laptops might take a closer look at their budgets and plan IT investments more carefully, asking which upgrades are the most necessary and embracing savings strategies, such as bulk buying programmes. Sharma urged executives to draw a clear line between "wants" and "needs".

"Try to look at it more from the cost modelling standpoint," he said. "How much do these kinds of components contribute to your overall costs?"

What will solve the shortage?

Supply disruptions have been persistent across industries, but semiconductors are an especially tricky puzzle to solve, thanks to the technical challenge of producing a high-end chip.

A typical semiconductor production line can involve 700 manufacturing steps across 14 weeks, Sharma said. The lithographic machines that are used to project circuits onto the chip can cost from $25 million to $200 million each, he added.

Given the importance and scale of the problem, national governments have initiated multibillion-dollar responses. At the end of July, the US Senate passed a bill to boost domestic semiconductor production with industry subsidies of $52 billion.

In France, according to Reuters, a $5.7 billion chip factory is expected to receive substantial government support — while Intel has planned even greater investment of €80 billion across the continent over the next decade. India has a $10 billion plan to attract chip makers — a huge amount for the nation, Sharma said.

And in China, the government has invested $150 billion into chips and has a $1.4 trillion plan for tech investment over five years, according to the US Congressional Research Service. That kind of global investment might help ease the crunch, at least in some markets, as nations build their chip supplies.

"So, we may be at this transition point where the supply constraints are still there, somewhat, and demand in some end markets is beginning to realign to reflect post-COVID and other macroeconomic factors. However, there is now funding and long-term plans to address future needs," Clark said.

But it could take years for the market to fully rebalance, and it's hard to say when the situation will improve for buyers. In a recent KPMG survey of semiconductor industry executives, around 65% expected the shortage to "ease" by mid-2023.

But Sharma is more pessimistic. "It's not coming to an end in 2023, if you ask me, honestly," he said. "Because the entire gestation period, the cycle time [for manufacturing] is wrong. There is a phenomenal amount of complexity."

— Andrew Kenney is a freelance writer based in the US. To comment on this article or to suggest an idea for another article, contact Oliver Rowe at Oliver.Rowe@aicpa-cima.com.