The Semiconductor Industry is Heating up
Over-reliant and underinvested
Semiconductors are used in virtually everything electrical nowadays including phones, renewables, washing machines, medical equipment, cars, and any other digital device you can think of. These semiconductors are important to our military and power grid, and much like oil (which I wrote about last week ), they have become essential to daily life given our dependence on technology.
So, having said that, it is easy to see how we have found ourselves with a global shortage. Did you notice it? Anyone who wants the new PS5 will notice they are impossible to find. This is due, for the most part, to the semiconductor shortage.
So how did the global shortage occur and when will it be over?
COVID-19 was the obvious reason. Populations spent months bored at home during lockdowns and took the opportunity to upgrade their tech, play video games, and so on, all of which demand semiconductors as essential components. The global semiconductor industry generated $439bn in revenue during 2020, up 6% from the year before, and I believe this is only the start of a long secular bull market that could continue until the end of the decade.
Global Semiconductor market size by year (Billions of US Dollars) with predictions into 2022, Source: Bloomberg
We have recently observed many car manufacturers being forced to reduce production. General Motors (GM) said it could lose up to $2bn in earnings in 2021 due to semiconductor shortages. Vehicles are demanding more and more chips as they become more technology dependant – one example being Tesla’s recent software update for autonomous driving. It has been said that technology advanced by seven years due to the pandemic. This acceleration is set to continue with the rollout of 5G and the Internet of Things (IoT), which will inevitably require even more semiconductors.
On the supply side, there are also several issues. With many of these factories having to close for a period due to the pandemic, when the supply chain was interrupted. This was hindered further by the huge concentration of production, with Taiwan Semiconductors (TSM) and Samsung (SMSN) manufacturing 70% of the world’s semiconductor supply.
Semiconductor supply issues are here to stay
Global foundries is the third-largest semiconductor manufacturer in the world, supplying chips for Nvidia, Apple, and other global tech companies. The Vice President of its’ Singapore factory has said that supply-side constraints could last until 2023, but I believe he is being optimistic.
The process of chip manufacturing is protracted. It can take up to six months for a chip to be produced and installed into a device due to the extensive production process and rigorous testing methods. To get to that point, Global Foundries say orders need to be placed at least a year in advance. Therefore, many of the orders placed at the height of the global shortage are only now being processed.
“It takes between $10bn - $15bn to build a new semiconductor factory.”
Several semiconductor manufacturers have announced substantial spending plans to increase capacity to meet increasing global demand. Taiwan Semiconductors (TSM) is investing $100 billion in new plants, Samsung is spending $116 billion, and Intel $20 billion. These numbers are huge, but the capacity will take time to come onstream and so the chip famine is set to continue.
It can take up to 15 months to acquire new machines, making increasing supply extremely difficult in the shorter term – These machines need semiconductors also. Moreover, between $10bn - $15bn is needed in capital expenditure to build a new factory and it can take five to seven years to perfect manufacturing processes. Firstly, this huge investment provides a significant barrier to entry (or moat) for would-be competitors seeking to enter the market. Also, the complexity of construction means supply will not increase quickly.
Manufacturers are getting smarter
Once the commitment to building a new production facility is made, the company cannot turn supply on and off quickly due to the significant and prohibitive costs involved. This has been the problem with the semiconductor industry in previous cycles. Companies would build new factories and because of the increased product supply, once demand wavered, they oversupplied the market and lost pricing power (and therefore profits). This has caused it to become a cyclical industry.
Today is different. Management of semiconductor companies are being a lot more cautious about committing capital because they want to keep prices stable and high, ensuring that there is sustainable demand before increasing production.
Chip companies are now becoming more efficient too. 95% of chip manufacturing is automated so many are implementing control towers to monitor the output of every machine on the assembly line. This has led to a 20% increase in production alongside increasing margins as these control towers improve efficiency without added cost. We should continue to see this trend in many semiconductor companies in the next few years.
One risk to the industry is that Taiwan Semiconductors and Samsung are both located in geographically sensitive areas. China has already tested Taiwan by flying over their airspace, and if they were to attempt a takeover similar to that of Hong Kong, there would be a major threat to semiconductor supply. China would likely use the United State’s over-dependence on these manufacturing companies as a bargaining chip. This is an issue that we should mitigate by sensible position-sizing, as there is no easy way to hedge this risk.
Largest semiconductor companies by revenue (2020), Source: bizvibe.com
Growth at a reasonable price
There are several strong brands at the heart of semiconductor development that you can research further. AMD, NVIDIA, and European manufacturer ASML are all strong companies. Note that each company produces semiconductors for various different electrical products, so ensure you do your research and are happy with the type of exposure you are getting. Broader exposure to the industry could be gained through VanEck’s Semiconductor ETF (SMGB).
The semiconductor industry is trading at a lower valuation compared to the broader market. With the recent rise in yields, investors may be encourages to buy bonds, but I think this is a good time to add some exposure to growth companies, as investors seek superior returns to the risk-free rate, or the yield on the US 10-year bond, which currently stands at 1.46%. Higher inflation would benefit value stocks more, however, if I concentrate on companies with attractive valuations relative to their peers then I hope to outperform the market. No one can predict the future, but concentrating on companies with strong economic moats, a track record of high ROCE, and selling into markets with major growth prospects should increase the odds of doing so. The future trajectory for this industry is clear, with western democracies recognising the structural importance of semiconductor companies. The current debate over the takeover of ARM holdings demonstrates this.
An excellent opportunity I have found in this industry is Micron Technology (NYSE: MU). This company will be involved in the 5G transition through its semiconductors. Revenue is up 29% year over year, with 17% growth expected in 2022 and 2023. I would not be surprised if expectations were exceeded over the next few years, for the reasons outlined above. It has also produced $2.8bn in free cash flow, and management are using this to return value to shareholders. Dividends are coming, and a share repurchase program has begun. The company has already purchased $1.2bn worth of its own stock this year. Believe it or not, this all comes at a valuation of 4.3x 2022 EV/EBITDA, with EBITDA expected to grow 50% both next year and the year after.
I believe that the shares are trading so cheaply due to previous volatility in revenue and margins because the company committed too much capital and over-saturated the market with chips in previous cycles. That is now behind the company and pricing power is returning as the management team is more cautious this time. One pertinent risk I see is that the company obtains 72% of revenue from its DRAM segment, however 97% of the DRAM market is controlled by five companies, Micron being one. The estimated growth and valuation are too attractive for me ignore this company though.
Action: Adding 4.1% weighting in Micron Technology (MU) to my Portfolio. Price taken from the close on Friday at $70.97.
#Postbox
This is the first question from a reader. Thank you for this and anyone who has any queries please feel free to message me.
I am trying to move more into active investment from a passive strategy. How much time do you invest in tracking the market? I find myself getting caught up in the noise when reading multiple articles. Would love to hear your approach and the sources you utilise which help you identify market inefficiencies.
I track the markets daily. If you head over to The Spark Instagram, you will see in the highlights section a number of sources to keep up to date with markets. I watch the AmplifyMe market update every morning, and read the Financial Times every day. I also subscribe to Investor Chronicles and MoneyWeek. When analysing stocks, I use TIKR Terminal, but the most important resource when analysing any company is their quarterly and annual reports. These tell you almost everything you need to know.
It is very easy to get caught up in the noise when reading articles. Everyone has their own opinion on market direction, but you must remember that many articles are for information purposes and not specific to what is happening today (which is why I started The Spark). You have to remain objective when reading any article and have a process for how you decide market direction because, at the end of the day, I nor any other writer has control of your portfolio.
Being impartial comes with time and experience, and I am by no means an expert. But there is always someone predicting the end of the world, after all, it sells. Many articles such as these are good reminders of why you should always hold a diversified portfolio, and some insurance (which I will write about in a few weeks). I recommend sticking with the hard data. Numbers tell you everything and many are leading indicators of what is to come next. For example, the bond market (Yields), Inflation and GDP growth figures, ISM manufacturing data, and so on. For example, if you got defensive when the ISM numbers went contractionary (below 50) near the end of 2019, you would have been in an excellent position when the markets crashed in March 2020.
So, in summary, take everything on board and form your own views. Develop a process but continue to learn and improve it, especially after losses or mistakes. In financial markets, you can be right, or you can make money. Many financial columnists have been writing about the over-valued US market for years. They may be right but staying away would have lost you a lot of money.
Until next time,
Peter
Disclaimer
This communication is for informational and educational purposes only and should not be taken nor used as investment advice, as a personal recommendation, or solicitation to buy or sell any financial instrument. This material has been prepared without considering any particular recipient’s investment objectives or financial situation and has not been prepared in accordance with the legal and regulatory requirements to promote independent research. Any references to past or future performance of a financial instrument, index or structured product are not, and should not be taken as, a reliable indicator of future performance. I assume no liability as to the accuracy or completeness of the content of this publication.