How Long Will the Semiconductor Shortages Last?
The shortage of semiconductor chips is straining global supply chains, causing a drag on the production of an array of products from smartphones to new cars. We spoke to Alexander Duval, a technology analyst in Global Investment Research at Goldman Sachs, after our virtual European Digital Economy Conference to discuss chip production and the impact of the Ukraine-Russia crisis on the sector. He highlighted how a number of European companies at the conference expect tightness in the supply and demand of semis to potentially persist in their particular product areas into early 2023, albeit with some gradual easing in certain types of chips in the second half of this year.
Based on your conversations at the recent technology conference, what did you hear about the backlog situation for European semiconductor producers and what is driving demand for their products?
Several companies highlighted very large backlogs for this year relative to what they can produce. The general view expressed was that new capacity would be needed for supply and demand to come into balance and that this could take some time.
As far as the demand picture for these European semis producers is concerned, there are obviously certain industries that their products go into which are rebounding or seeing strong growth — automotive, for example, particularly given the limitation on what was produced during the initial innings of Covid. But more broadly across semis there are interesting secular demand drivers, whether we are talking about electrification of the car, the proliferation of the cloud, right through to other areas where semis have become a digital enabler of things like automated driving. And then of course the transition to 5G, as Europe and other economies digitalise, requires a lot of wireless equipment that in turn drives volumes of advanced semis.
What does this all mean for the supply chain crunch?
And what does it mean for supply to come on stream?
This would mean these companies not only putting equipment into existing semiconductor fabrication facilities, but such producers, or their foundry partners who produce for them, needing to build new factories and ramping those up. And so our view is that we’ve had a number of good quarters for many of these semis players in terms of earnings prints in a very, very tight supply-and-demand situation.
But we do think there were decisions made a number of quarters ago to start ramping supply. So the relevant facilities will at some point be up and running with supply coming on stream. And what we have seen is some companies talking about supply gradually coming on stream in the second half for certain types of semis, but plenty of other companies talking about only really a step change in the broader supply situation potentially happening end of this year, maybe even early 2023.
If these companies are right in their view that it’s going to be more of a gradual easing, and only really in certain pockets, like power semiconductors, then this would mean there is tightness until next year.
You asked several executives about double ordering and what kind of inventory-correction risk there is. What was your takeaway from those conversations?
Broadly speaking companies did not highlight double ordering, with one talking about a high level of customer escalations to request more semis, implying genuine demand remains high. Further, several suggested inventories in the market are still very low. That said, several European semis companies will have made investment decisions multiple quarters ahead, and, given they don’t have perfect information, it is possible that there could be some kind of inventory correction in future, even if the implication of some companies’ statements was that the risk of this would be more likely in the second half and in particular towards the end of the year.
Does the crisis in Ukraine change your projections for semiconductor supply and demand?
It’s obviously a very fluid situation and we can’t say there’s no risk at all, but near term the risk seems relatively limited. If you think about the sort of end market exposure — Russia is very small. It’s less than 1% of global IT spending. For the European semis companies we cover, the exposure is typically below 2%. The direct exposure to Russia-Ukraine is pretty low.
Now, one thing we have seen press reports about is neon gas. A decent amount of the market for this is located physically in Ukraine. And you need that for lithography machines, which use light to print patterns onto silicon wafers, such as those provided by ASML, a semiconductor company in the Netherlands. Some of these lithography machines do need this neon gas. But actually around 50% of ASML's lithography business, in terms of the types of machines, is actually EUV, which does not rely on neon. The DUV machines, which are not quite as advanced, do require neon. But they’re only getting 20% from Ukraine, as they also source from areas outside Ukraine. So we think it will have only a limited impact for them.
In terms of the broader semis landscape, we’ve highlighted in our research that a certain number of the large semis manufacturers — who could also theoretically be impacted by changes in the availability of neon — have been saying that there isn’t very much to worry about, in the near term at least, as far as their ability to produce or cost inputs are concerned. What you’ve seen is that a number of these companies have diversified their sourcing of neon, meaning that they can get it from suppliers with whom they have contracts for a given price and given access to volumes outside Ukraine. We also believe it will be necessary to monitor the macroeconomic situation.