The global microchip shortage has become headline news, as repercussions
have spread from core technology products to other chip-heavy consumer
goods — from autos to appliances and beyond. But while the unique
circumstances of the pandemic brought the supply/demand dynamics of
the chip industry to the brink, it can be argued that the path to the crisis
started decades ago. And as the pandemic hit, an unlikely but powerful
set of factors combined to create an unprecedented shortage of one of
the basic building blocks of our increasingly digital society.
Columbia Threadneedle analysts and investment leaders with a long history
of researching and investing in technology came together to discuss the
story behind the semiconductor shortage, how it happened and what we
can expect next.
Question: It’s been said that the chip shortage was along time coming. What were some of the trends in
supply and demand leading up to the current situation?
Paul Wick: The dynamics we’re seeing play out today have been building for a while. Digital semiconductor companies used to have their own production
facilities. In the 1990s that started to change as the industry increasingly
moved to outsourcing manufacturing to fabricators in Asia, mostly in Taiwan.
So, instead of 40 large chip companies with their own facilities running at
70% or 80% utilization, you have a handful of dedicated fabricators that tend
to run their factories at 90% plus utilization, which is much more effi cient. As
a consequence, there’s a lot less slack in semiconductor capacity.
Sanjay Devgan: The way semiconductor companies hold inventory has also
changed. Before the tech bubble in 2000, most original equipment
manufacturers (OEMs) carried chip inventory on their balance sheets. One
of the most infamous inventory write-downs in history was a $2.7 billion
write-down by one of the big OEMs when the bubble burst. They no longer
wanted to take that risk. So, over the last 20 years inventory on the balance sheets — and in the warehouses — of the OEMs has come down and the
inventory on the balance sheets of semi suppliers has gone up.
David Egan: Concurrently, we’ve been witnessing a dramatic growth in tech intensity. Personal
computers in the 1990s. Phones and then smartphones in the 2000s. And the soon-to-be
gigantic companies that were building businesses around mobile internet — the Amazons and
the Googles. At the same time, businesses in many industries were undergoing digital
transformations. All of this added up to a steady increase in demand for semiconductors.
Question: How did the COVID-19 pandemic impact the current shortage?
Rahul Narang: There was a basic miscalculation of how COVID-19 was going to impact the world
economies. Most assumed companies would cut back production as times were getting diffi cult, but demand proved to be quite resilient — even growing for some categories. People were stuck
at home and buying and using more technology, whether it was new or additional PCs, bigger
displays, 4K TVs, game consoles, remote meetings, etc. A sort of perfect storm formed that
combined the demand miscalculation with lean inventories and a surprising consumer appetite
for products full of microchips.
Charles Mann: Automakers in particular were hard hit, as they were unable to reinstate canceled
chip orders when they realized consumers looking for safe transportation in the pandemic
wanted more cars, not fewer. As cars have gotten more complex, data-driven and automated, the
auto industry has become one of the major consumers of microchips. Chips have begun to take
on and evolve into a new role in the automotive supply chain, becoming central to it – not only
because of the “smart” content, but also the data sharing capabilities. In some cases, they’ve
become so central that the lack of one specific, unremarkable chip can bring production to a
virtual standstill.
David Egan: On the supply side, COVID restrictions on factory workers and international trade
slowed production and delivery. On top of the pandemic disruptions, there was a fire in a major
substrate facility (a key part of the chip packaging) in Asia. This affects things like gaming
processors and networking equipment. We’ve seen reports that this has pushed product back
from some companies for more than a year. And in February, the polar vortex in Texas shut down
a number of big facilities, creating even more chinks in the semiconductor supply chain.
David Egan: On the supply side, COVID restrictions on factory workers and international trade
slowed production and delivery. On top of the pandemic disruptions, there was a fire in a major
substrate facility (a key part of the chip packaging) in Asia. This affects things like gaming
processors and networking equipment. We’ve seen reports that this has pushed product back
from some companies for more than a year. And in February, the polar vortex in Texas shut down
a number of big facilities, creating even more chinks in the semiconductor supply chain.
Question: How will this situation be resolved? What is the path
forward for the industry in the next couple of years?
Rahul Narang: It’s important to remember that some powerful longer term trends were also a factor in the resilience of demand during this crisis. The move to 5G, digital payments, digital
enterprise transformation, cloud computing, cybersecurity, AI, autonomous driving, electric
vehicles, battery technology and more. These are all things we’re watching closely, and when we
step back and look at these themes, we feel that this pandemic has kind of hit what I’d call a
giant fast forward button to the future for many of them.
Sanjay Devgan: From an investment standpoint, visibility for the semiconductor industry is probably at an all time high. Historically with semis, when you place an order, the order is only
good for that quarter. And then you have to build the backlog for the next quarter. One major firm
is leading a change in that paradigm, offering what they call a “preferred supplier program.” If you
give them one year’s worth of non-cancelable orders, then they will move you to the front of the
supply queue. Basically, this gives the company revenue visibility out for a year. And we’re seeing
this type of thing happening across the entire industry. We’re also seeing that pricing from the
foundries is going up, as well as materials prices, and the semi companies are able to pass that
onto their customers on a cost-plus basis.
Sanjay Devgan: From an investment standpoint, visibility for the semiconductor industry is probably at an all time high. Historically with semis, when you place an order, the order is only
good for that quarter. And then you have to build the backlog for the next quarter. One major firm
is leading a change in that paradigm, offering what they call a “preferred supplier program.” If you
give them one year’s worth of non-cancelable orders, then they will move you to the front of the
supply queue. Basically, this gives the company revenue visibility out for a year. And we’re seeing
this type of thing happening across the entire industry. We’re also seeing that pricing from the
foundries is going up, as well as materials prices, and the semi companies are able to pass that
onto their customers on a cost-plus basis.
The bottom line is that from the current vantage point, industry analysts and
semiconductor companies are predicting chip shortages until the end of the year, and
in some situations, into 2022. Despite this setback, we see the key demand drivers for
technology continuing relatively unabated — some even accelerated by the realities and
lessons of the pandemic.
As the pandemic economy fades, we expect to see a slow, and hopefully orderly, shakeout of the chip shortage over the next 9-to-12 months. Over that time, our team
sees strong fundamentals, and opportunities for value in the semiconductor sector, including equipment suppliers, OEMS and manufacturers, against continued broad and
strong demand. Compelling growth trends in advancing and emerging technologies also
have the potential to provide support for expanded capacity and chip supply over the
longer term.
