Paul
Hello and welcome to Macro Bytes the economics and politics podcast from abrdn. My name is Paul Diggle and along with my co-host Stephanie Kelly, we're guiding you through the macro economic and political themes which are driving global markets. And today we're talking about the global semiconductor shortage, its causes, consequences and market implications. Semiconductor chips are in short supply and high demand at the moment. Semis are technologically hard to make, production is highly concentrated in the likes of Taiwan and Korea. Yet these chips are in an ever increasing share of the goods that global consumers buy. So a semiconductor shortage is akin to a modern day oil shock with widespread consequences for economic activity, inflation, and even geopolitics. So I'm really excited to discuss this topic with Robert Gilhooly and Sree Kochugovindan, two of our Senior Economists in the abrdn Research Institute, Bob and Sree, welcome back to Macro Bytes.
Robert Gilhooly
Thanks for having us.
Sree Kochugovindan
Thank you.
Paul
So Bob, let's start with you. Why is there a global semiconductor shortage?
Bob
Okay, well, first of all, I guess there's been a few kind of idiosyncratic issues affecting supply in some industries. For example, a fire a Japanese factory that specialise in semiconductors for autos, we had earlier the US regulatory action against Huawei in China, that was rumoured to have led to a kind of some emergency stockpiling, but other firms within China, but probably say more generally, it just represents combination of very long lead times for semiconductors to ramp up, production was very capital intensive, very specialised, especially at the top end, and then it's just made it very hard to deal with what's turned out to be just much stronger than expected global demand.
Paul
Great. So as you say, a lot of supply side issues which have held back supply. And Bob, why has demand been so strong, because we've been living through a massive negative global economic shock for much of the past year, which you would expect to push down on demand. But actually, demand for semis has been extremely strong. Perhaps you could explain why that's been the case.
Bob
Yeah, that's right. And this has been one of the big surprises, I think, the COVID shock, compared to the global financial crisis, goods demand did fall, clearly. But the nature of the COVID shock is definitely mattered here. Higher demand for work from home, school from home and entertainment from home. As you know, we've just not been able to go out and enjoy the services that we would do so normally We don't have a kind of comparable data for all countries. But I've been really struck, that the US share of goods consumption has jumped by about four percentage points. And that's really quite remarkable. And, also we've seen global trade, which you might have expected to take a big hit, given the nature of the trade, given the huge shock to global economy. But global trade, especially that originating from Greater China, linked to semis. It's been running at levels far above your sort of pre-COVID norms, you would say many economies still having room to catch up.
Paul
Yeah, strikes me that a lot of semiconductor producers expected this big negative demand shock. And partly that occurred early on when a lot of autos orders of semis were cancelled, pared back production. And then it turned out that actually, as you explained, Bob, the demand for semis was actually very strong due to the nature of the COVID shock. And the other thing, this underlines for me is just how widespread semiconductor chips are in our economy. Now they are in everything and in increasing amounts as well,. The amount of semis in electric vehicle is much more than a traditional petrol car, the new PlayStation, latest iPhone, all these things are extremely chip intensive. I think that that's an important change in the structure of our economy. This is really highlighted. Let's talk about how long will this might last then, at what point are global semiconductor producers, the foundries that make semiconductors going to be able to meet this rise in demand and this shortage be alleviated?
Bob
Yeah, so I guess this goes to two word factors here, you know, a one, would you kind of think some of the supply constraints that we've seen, should be unwinding perhaps towards the end of this year. So we might see some easing on that front, it'll be the end kind of Q3, early Q4. And that could kind of ease some pressures and kind of availability, if you will. But I guess the other part is going to be related to really this kind of normalisation in the rest of the economy. So to the extent to which goods demand has been this uncharacteristically strong. Will it stay that way? Or will it kind of ease off as services kind of become available, restrictions are eased, the populations are vaccinated, and they may be going to just come back to services rather than kind of buying goods. So there's a limit to how much computer equipment you want, or that's certainly something I always I always tell my kids.
Paul
Brilliant. Yeah, so supply should eventually respond and demand itself, it's going to change away from these chip intensive goods. But I suppose when you require such large factories of such large amount of investments, such big leads time, there is a limit to the price elasticity of supply presented by that. Sree, let's talk then about the implications this has had, we want to talk about inflation. But let's first talk about the production implications. Because as we were saying, semiconductors are in a lot of things now, and they've impacted the production of a lot of goods. And the real, the biggest impact has been in in autos, right, car production.
Sree
Absolutely, yes, I'm glad you say that. I mean, this comes through in a lot of household goods, such as washing machines, pelletons, PlayStations have been hits as well, consoles, some suppliers being restricted. But autos has been the key focus and really bore the brunt of the supply chain disruption. So countries like Japan and Germany, we've seen a big impact there in terms of their industrial production auto sectors there, partly because they've really been running on lean, just in time production. So there was very little room for manoeuvre there. There was, as Bob has mentioned, there was a fire let's take Japan as an example, as a fire at Renesas, which is a chipmaker. And accounts for around 8% of Japan semiconductor production for autos, about 6% of the global auto semis supply. And this led to quite a sharp decline in Japan's auto production, industrial production as well. And there are also reports of prices from TSMC, which is one of the main producers in Taiwan, one of the largest contract producers of chips. Now they're raising the prices for auto semiconductors. And that really compounds the problem broader supply constraints and shipping costs. And it's it's a real headache for the industry at the moment. But we do think that this is likely to start easing in the second half of this year. And there have been some, the industry has started to adapt, for example, there are different types of semiconductors used for different parts of the car. So the high tech, leading edge tech is used in assisted driving systems. And some cars are being produced without that and just being using the lower end, older tech chips. There are some changes that producers can can adapt to.
Paul
Great, but nonetheless, some some really big impacts, which are concentrated in certain economies as well. Auto intensive economies.
Sree
Yes and also longer term. Structural demand is there as well. So this, we're just talking at the moment about the near term issues, but we also have the longer term structural demand for it for chips from 5G, AI, healthcare, electric vehicles. There's a lot of changes to come in terms of the demand over the coming years.
Paul
Yeah. And it seems to be going in one in one direction. So what about the inflation implications, then? Because you've been looking at this very closely, Sree in some of our work at abrdn? And can we detect a direct influence of chip prices in things like headline inflation indices. Are chips yet a big enough part of the inflation baskets are they actually influence headline inflation at this point?
Sree
It's very difficult to see how chip prices are feeding through First of all, what is the chip price there are different types of chips and they all have different prices. And they will have different movements. So some high end could be less volatile than the lower end chip. So it's a very diverse set of prices. So use various proxies to try and test how it goes through into headline inflation. But really, we're seeing this pushing through in some subcomponents of PPI or subcomponents of import prices, it doesn't directly feed into into CPI. It's very dispersed, as you said, it goes through different sectors, different goods. And some of these would go back to the example of autos, and these costs are quite small relative to the production of the car. So you can see a lot of this is going to be absorbed in margins. And in some case, there could be some pipeline pressures, again, in the broader context of other supply chain disruptions. It has been a bit of a headache, but from the tests that we've done, we've seen these kind of price increases before, we've seen them in 2016, 2017. They don't necessarily feed into headline inflation. It's not enough to really trigger a shift in inflationary pressures. That will be one thing. The other thing is to say that generally, semiconductors or tech can be quite disinflationary. We have quality improvements, and these can push down technology prices, as processing power increases as a hedonic adjustment that's made. So this tends to be quite disinflationary rather than than an inflationary process.
Paul
Yeah, I think that's a really important point, technological changes fundamentally disinflationary and might not push down on the observed price of the good. But if the quality of that good is getting constantly better, our computers are constantly faster, that is actually disinflationary once you account for the quality adjustment. But you mentioned there Sree broader supply chain disruptions because this chip shortage comes at a time when global supply chains are disrupted in all sorts of ways post pandemic, the Suez Canal blockage was obviously a sort of a really high profile, obvious example of that. But there's been a lot of other examples at the same time, as demand has been rebounding strongly. And is this combination of supply chain disruptions and strong demand causing price pressures elsewhere in the economy? Is this a sustained rise in prices that we're seeing?
Sree
Well, we've been talking about this a lot. It is much more general, it's not just confined to semiconductors and autos. This is a broader, broader issue. But we're seeing that these are pandemic related distortions that are likely to be transitory, we are going to see this speeding through into core inflation later on in the year, there are some price pressures and volatility and inflation going into the end of the year. As we move into next year, then we'll see a cleaner picture in terms of how inflation responding to the main fundamental drivers such as the output gap, inflation expectations, Central Bank policy, and so those things will become much more evident as we move into next year. But at the moment, for now, we have a lot of distortions in play, whether it's measurement issues or supply constraints.
Paul
And Bob are, so at the top, I made a comparison between semiconductors now and oil in the past as sort of a key ingredient almost a base commodity of the global economy. And that might be a reason why the semis shortage is so important that really grabbed market attention. Do you think that comparison is fair? Should we think about these as being akin to each other? Or is it the case that you semis are just not like oil? You can produce them elsewhere? Or how do you feel about that comparison?
Bob
So I think this comparison has got got some some merits for so for example, you could think of kind of currently, semiconductor production is very, very much based in Asia, in particular, you know, there's some concerns, but risks are, a geopolitical risks around Taiwan. And, of course, Samsung is the other major contributor, major producer of high end, semiconductors and it's got its neighbour, the north of North Korea. So you do have roughly 80%, if not 100%, when you just look at the top end of semiconductors, in what you might consider kind of challenging or kind of risky regions. And it's not just geopolitics, there's this kind of natural risk as well. You know,Taiwan sitting on a fault line, sort of risk of earthquakes there. And Taiwan's also been affected by kind of water shortages, most recently, they've had to divert water away from foreign land to kind of keep the TSMC foundries running there. But I think, fundamentally, is it like oil? Well, no, not really, you know, you're kind of limited into where oil is geographically in the world. You know, things change a little bit with shale. But you still got to basically dig it out wherever it is clearly over the kind of long run, you know, you can move semiconductor manufacturing, you can increase supply, by more, you're not constrained in the same way as oil. So I think in most respects, that kind of analogy starts breaking down. It breaks down quite quickly, which gets stuck into that.
Paul
So yeah, brings us nicely to sort of, the final topic we want to deal with here, which is the geostrategic aspects of semiconductor production and the shortage because, as we've said, these chips are mainly produced in Taiwan and Korea, potentially politically volatile parts of the world, or at least potentially facing the sort of political risks, at least entail scenarios. One of those is proximity to China. And maybe they're sort of extreme tail scenarios involving China, Taiwan and China. But more broadly, chip production has become an important aspect of global geostrategic competition. And that's because China wants, I think, to in-house, more of this production, the EU, and the US wants to in-house more of this production and be less reliant on these global supply chains, which, as we've seen are subject to disruption. Bob, what is the prospect of production moving from the likes of Taiwan and Korea, to China and into the US and the EU?
Bob
Yeah, I think over the kind of medium term, very high. So kind of, can it be done? Well, yes. But you know, governments are going to have to throw a lot of money at it through subsidies. And I think there's a caveat here, even then, you know, you might not end up with the most cutting edge technology. So at the moment in the US we put the chips at, secures about, well, between about 30 to $37 billion in federal funding from manufacturing and R&D. And then we've got local subsidies, like at the state level, which will probably go on top of this, or TSMC, has said it will build a five nanometer fab, in Arizona, you know, US could spend about 12 billion US dollars. It does, of course, in a way kind of, I guess, pales in comparison to the kind of fabs that are actually in Taiwan at the moment, you know, those costs more or less in the region of $20 billion, or factory of fab. And you're when the US fabs are completed, they are actually going to be somewhat behind the latest factories in Taiwan. One illustration, I think, showing you know how hard it is to catch up. In terms of the EU, you know, the EU wants a quote unquote, digital digital decade. And it's talking about maybe, subsidising production to the tune of about 20 billion euros. There's a question for both the US and the EU though, you know, it's not just about spending this once off, you've got to basically kind of commit to spending this sort of money almost year after year, if you want to kind of stay close to the top end, where your national security concerns are playing into that kind of dynamics you've discussed there for China, the EU and the US. And we think China is willing to keep on providing money on this front. And US-China, tensions are clearly eating into this as a priority for China, given the fear that China may potentially be cut out leading leading-edge technology that might hinder its development going forward.
Paul
Brilliant, and certainly no doubt that it is an increasing focus of policymakers both in China and the US and the EU. And we hear an increasing amount of communication from the Biden administration on this as a national security considerations. I'm sure, we're going to hear more about this very important industry. Bob and Sree, thank you both very much for joining the podcast. That's about all we have time for this week. Thank you for listening to Macro Bytes. Remember that we have a mailbox macrobytes@abrdn.com and we'd love to hear from you with feedback on the podcast. Don't forget to like and subscribe on our podcast on Apple Spotify, or wherever you listen to your favorite podcasts. But until next time, goodbye and good luck out there.
Voice-over
Please note that email is not a secure form of communication, so don't send any personal or sensitive information. This podcast is provided for general information only and assumes a certain level of knowledge of financial markets. It is provided for information purposes only, and should not be considered as an offer, investment recommendation or solicitation to deal in any of the investments or products mentioned herein and does not constitute investment research. The views in this podcast are those of the contributors at the time of publication, and do not necessarily reflect those of abrdn. The value of investments and the income from them can go down as well as up and investors may get back less than the amount invested. Past performance is not a guide to future returns, return projections are estimates and provide no guarantee of future results.