Schwab Sector Views: Semiconductors—Boom or Bust?
The semiconductor shortage and its impact on everything from autos to smartphone production has been much in the news. The shortage has been a boon for semiconductor stock prices. But it likely will resolve itself in the coming months—or years, depending on whom you talk to—raising the specter of a bust. How likely is it? Could it be different this time? And how should investors respond?
Chip shortage: How did we get here?
The saga began with the emergence of the COVID-19 virus in early 2020. The chip industry expected a major cyclical downturn in demand, as has happened many times during past recessions. Initially, that seemed to be the case, with automakers canceling orders as they prepared for what they thought could be a 20% to 40% decline in car sales.
However, amid stay-at-home mandates, consumers quickly shifted their spending from dining out and traveling to home-entertainment electronics and home-office equipment. Businesses scrambled to snap up PCs, monitors, and internet routers to help workers telecommute. Cable companies had to expand their infrastructure to meet the demand, and cloud data centers ramped up capacity. As all these things require semiconductors, chip demand rose 10.8% in 2020 and is estimated to have grown another 17.3% in 2021.1
Manufacturing orders for computers and electronics have risen 15% from pre-pandemic levels
Semiconductor supply did not keep up with demand. First, there were the COVID-related closures of chip fabrication facilities. Then a fire in Japan, freezing weather in the southern United States, and drought in Taiwan—where the vast of majority of chips are made—temporarily halted manufacturing. With many industries operating with just-in-time inventory (keeping enough on hand for a few days), their slim stockpiles of chips were quickly drawn down. This impacted all sorts of products from agriculture and heavy machinery to gaming consoles and smartphones. The auto industry’s problems are well-known, with automakers forced to halt production of many models and even cut back on electronic features like digital speedometers and infotainment screens.
Another problem has been who gets the chips that are available. For example, autos, kitchen appliances, and basic electronics tend to use cheaper, earlier-generation “commodity” chips that have become less attractive for the major semiconductor fabrication (“fab”) companies to make. Not surprisingly, semiconductor fabs—like Taiwan Semiconductor Manufacturing Company (TSMC), Samsung, and GlobalFoundries—have prioritized their biggest and most profitable customers, like Apple, Qualcomm, Nvidia, and Advanced Micro Devices, “fabless” companies that create chip designs to be built by the fabs. In recent years, the fabs have built more capacity for the sophisticated designs sought by fabless clients for their newest products (more on this below), leaving reduced production capacity for older semiconductors.
The fabs are trying to catch up with demand, increasing production of the commodity semiconductors by 60%, in part by recommissioning retired equipment. But this will take time. Completing a chip requires nearly 700 steps, which can take as long as three months. At the end of October, wait times for chips and related components were between 22 and 38 weeks.2
Chip delivery times have gotten longer
But there are some signs that we’re closer to the end of the severe shortage than the beginning. Intel CEO Pat Gelsinger has predicted the backlog will ease in 2022 or possibly 2023.3 International Data Corporation (IDC) expects the chip market to rebalance in 2022, with the potential for oversupply in 2023 that could lead to a bust.4
The dreaded boom-to-bust cycle
To be sure, the semiconductor industry has been one of the biggest beneficiaries of the COVID-19 crisis, with the S&P 500® Semiconductor & Semi Equipment index—a subset of the S&P 500 Information Technology sector index—outperforming all other industries since the onset of the crisis. The biggest fabricators are running near 100% capacity, chip equipment orders have surged, and chip sales are up nearly 35% from 2019 through October 2021.5 The world’s largest chip fabricator, TSMC, is planning to raise prices for sophisticated chips by 10% and commodity chips by 20%.6 Intel has so far resisted hiking prices, but many of the fabless companies are showing that they have the pricing power to pass along higher chip costs.
Semiconductor stock prices have risen along with semiconductor sales
However, the industry has a long history of booms and busts, due to several factors. First, it is very cyclical in nature—it tends to do well when the economy and chip demand are strong, but falters during recessions. Our economic outlook is for continued but slower growth in 2022, so there may continue to be a cyclical tailwind for demand.
Second, semiconductors tend to become commoditized as innovation is quickly copied and prices decline. Former General Electric CEO Jack Welch once said that the semiconductor industry was always on the edge of “commodity hell,” when companies’ return on investment falls well short of expectations.
Finally, a surge in demand for chips—whether from an economic upturn or new technologies—tends to spur a lot of capital spending to expand production capacity. But it can take several years to bring a new semiconductor foundry online. And just when supply capacity is built up, demand often slows along with the economic cycle, or simply too much supply driving prices lower.
There is a risk that overinvestment is happening now, with massive spending planned around the world: An estimated $146 billion was plowed into semiconductor capital expenditures in 2021, up 50% from pre-pandemic 2019, according to technology market researcher Gartner Inc.7 And between Intel and TSMC, more than $200 billion in spending is planned—on top of potentially $52 billion by the U.S. government, $190 billion by the E.U., and $450 billion by South Korea.8
There are several things at play that’s spurring new investment besides just cyclical demand and the chip shortage. The shift from a just-in-time inventory approach toward a just-in-case system, means demand could be sustained as extra inventories are built up. Resilience and national security are also part of the equation. The chip-foundry side of the business is extremely concentrated: Taiwan’s TSMC accounts for 54% of total global chip production and 90% of the advanced chips, with South Korea’s Samsung producing the remaining 10% share. As the cliché goes, “semiconductors are the new oil.” There are obvious geopolitical risks associated with such a concentration in one region, to say nothing of natural-disaster risks. To mitigate these risks, many U.S. companies are planning to build their own foundries and/or buy more of their chips from foundries located outside of Asia—including in the United States.
Will it be different this time?
Is a bust coming? The reasons outlined above suggest coming excess supply. Once the foundries are built, it is very difficult to idle capacity. Huge capital investment requires cash flow to pay for it, which means production often continues—even at a loss.
What could prevent, or at least delay, a bust? The answer may lie with the ongoing technological evolution and the chips that will be required to make it possible. Onshoring of manufacturing and increasing labor costs will require automation to offset higher expenses to remain competitive. The rapid expansion of the “internet of things” (in which physical objects are increasingly embedded with sensors, processing ability, software, etc. to connect and exchange data with other devices over the internet) is necessitating faster and higher-capacity data transfer channels via broadband and 5G cellular. Data storage worldwide is growing at an estimated 17.8% per year—doubling about every 4 years—with data and cloud storage centers likely absorbing 60% of this growth.9 The aggressive adoption of clean-energy technologies, artificial intelligence, cyber-security hardware, and advanced smartphone and wearable technologies adds to the long list.
Perhaps some of the most consequential innovations and transformations are coming from the auto industry—ironically, the industry that has been crippled by the shortage of older-generation semiconductors. While automotive electronics account for just over 10% of global semiconductor revenue—dwarfed by data processing and communications electronics, which comprise nearly 65% of sales—auto-related technologies are expected to grow at an accelerating pace in the coming years. A standard combustion-engine car today has up to 1,000 chips, whereas electronic vehicles (EVs) have twice that number. It’s estimated that EV production will exceed that of gas vehicles by 2026, and the major car manufacturers will be fully electric by 2035, according to McKinsey & Company.10 And a whole host of semiconductor-dependent components come along with EVs: advanced batteries, charging stations, upgraded electrical grids, autonomous driving capabilities, and internet-connected management systems.
The transition into this new era is not possible without cutting-edge semiconductors. Yet an estimated $1 out of every $6 is still being invested in “mature” semiconductors.11 With over $1 trillion in planned investment, that’s a lot of capacity being built, so a bust in the short term is certainly possible. However, the vast majority of investment is going into the newest semiconductors, which are much more complex and difficult to commoditize. Given the rise in technologies that require these chips, the strong trend in industry revenue growth could be sustained well into the future.
Takeaway for investors
Investors who want to participate in this new era in technology need to understand the multifaceted dynamics of the semiconductor industry. Investing in it can be complicated. First, there are many different types of semiconductor companies—chip equipment makers, designers (fabless), foundries (fab), and hybrids (both design and fabricate).
Then within each category, companies specialize in different types of chips. While it’s common to use “semiconductors” generically, there are actually numerous classifications: microprocessor, memory, logic, analog, graphic processors, integrated, sensory, mixed circuit, and the standard “commodity” chip. To give you an idea of how far designs have evolved, consider this: “mature” chips measure about 40-45 nanometers (a nanometer is one billionth of a meter) and can contain a few hundred million transistors. However, newer 5- to 7-nanometer chip designs contain upwards of 55 billion transistors. The newest flash memory chips have over 1 trillion transistors and store more than 8 terabytes of data. That’s a lot! In some of the latest memory storage research, data is stored on DNA (instead of the traditional semiconductor chip), of which a single gram can theoretically store 215,000 terabytes of data!
Adding to the complexity, the line has blurred on what defines a chip business, as companies like Tesla, Apple, Amazon, and many of the major automakers are stepping into designing and/or fabricating their own chips—a more vertical supply chain structure. This contrasts with currently more prevalent horizonal structures where companies buy chips from third-party suppliers.
An index fund can be a way to gain exposure specifically to the semiconductor industry. There are some slightly more diversified index funds that also capture exposure to some of the more vertical companies—such as many of those mentioned above, as well as some in the software and cloud computing segments, for example.
Of course, investing directly in specific companies is another option, but this can require in-depth research on the industry as a whole and the fundamentals of each stock. There also are active funds that are more selective in their exposure to various aspects of the semiconductor industry. We provide resources on Schwab.com that can be a good starting point to research investment options.
Semiconductor valuations are elevated
Before investing, however, there are many risks that should be considered. While there are some reasons to believe the industry is transforming, it has a long history of being very volatile, prone to busts, and susceptible to high valuations—which may be the case now, as the chart above illustrates. It’s also important to take a diversified approach both within the industry and across all the equity sectors.
1 “Semiconductor Market to Grow By 17.3% in 2021 and Reach Potential Overcapacity by 2023, IDC Reports,” International Data Corporation (IDC), September 19, 2021.
2 Yang, Stephanie, and Sohn, Jiyoung, “Global Chip Shortage ‘Is Far From Over’ as Wait Times Get Longer,” The Wall Street Journal, October 29, 2021.
3 Gryta, Thomas, “Supply-Chain Problems Will Last Through 2023, Intel Chief Says,” The Wall Street Journal, December 7, 2021.
4 “Semiconductor Market to Grow By 17.3% in 2021 and Reach Potential Overcapacity by 2023, IDC Reports,” International Data Corporation (IDC), September 19, 2021.
5 Source: Bloomberg.
6 Yang, Jie, Yang, Stephanie, and Kubota, Yoko, “World’s Largest Chip Maker to Raise Prices, Threatening Costlier Electronics,” The Wall Street Journal, August 26, 2021.
7 Sohn, Jiyoung, “Semiconductor Industry Isn’t Spending Big on Scarce Old-Tech Chips,” The Wall Street Journal, November 9, 2021.
8 Kim, Sohee and Kim, Sam, “Korea Unveils $450 Billion Push for Global Chipmaking Crown,” Bloomberg May 13, 2021; Yang, Stephanie, and Jie, Yang, “TSMC to Invest $100 Billion to Increase Semiconductor Output,” The Wall Street Journal, April 1, 2021; “EU Eyes 'Chips Act' To Cope With Semiconductor Shortage,” Agency France Press, Barron’s, September 15, 2021; Fitch, Asa and Boston, William, “Intel to Invest Up to $95 Billion in European Chip-Making Amid U.S. Expansion,” The Wall Street Journal, September 7, 202; Leswing, Kif, “Intel is spending $20 billion to build two new chip plants in Arizona,” CNBC, March 23, 2021. Sohn, Jiyoung, “Semiconductor Industry Isn’t Spending Big on Scarce Old-Tech Chips,” The Wall Street Journal, November 9, 2021.
9 “The tech research firm thinks core data centres (on-premises and public cloud) will hold 60 per cent of the world’s data in 2024,” Block & Files.com quoting International Data Corporation (IDC), May 14, 2020
10 “Why the automotive future is electric,” McKinsey & Company, September 7, 2021.
11 Sohn, Jiyoung, “Semiconductor Industry Isn’t Spending Big on Scarce Old-Tech Chips,” The Wall Street Journal, November 9, 2021.
What do the ratings mean?
The sectors we analyze are from the widely recognized Global Industry Classification Standard (GICS®) groupings. After a review of risks and opportunities, we give each stock sector one of the following ratings:
- Outperform: likely to perform better than the broader stock market*
- Underperform: likely to perform worse than the broader stock market*
- Neutral: no current view on likely relative performance
* As represented by the S&P 500 index
Want to learn more about a specific sector? Click on a link below for more information or visit Schwab Sector Views to see how they compare. Schwab clients can log in to see our top-rated stocks in each sector.
How should I use Schwab Sector Views?
Investors should generally be well-diversified across all stock market sectors. You can use the Standard & Poor's 500® Index allocations to each sector, listed in the chart above, as a guideline.
Investors who want to make tactical shifts in their portfolios can use Schwab Sector Views' outperform, underperform and neutral ratings as a resource. These ratings can be helpful in evaluating and monitoring the domestic equity portion of your portfolio.
Schwab Sector Views can also be useful in identifying stocks by sector for potential purchase or sale. Clients can use the Portfolio Checkup tool to help ascertain and manage sector allocations. When it's time to make adjustments, Schwab clients can use the Stock Screener or Mutual Fund Screener to help identify buy or sell candidates in particular sectors. Schwab Equity Ratings also can provide a fact-based and powerful approach for helping you select and monitor stocks.
Schwab Sector Views do not represent a personalized recommendation of a particular investment strategy to you. You should not buy or sell an investment without first considering whether it is appropriate for you and your portfolio. Additionally, you should review and consider any recent market news.
All expressions of opinion are subject to change without notice in reaction to shifting market conditions. Data contained herein from third-party providers is obtained from what are considered reliable sources. However, its accuracy, completeness or reliability cannot be guaranteed. Supporting documentation for any claims or statistical information is available upon request.
Examples provided are for illustrative purposes only and not intended to be reflective of results you can expect to achieve.
Diversification and asset allocation do not ensure a profit and do not protect against losses in declining markets.
Forecasts contained herein are for illustrative purposes only, may be based upon proprietary research and are developed through analysis of historical public data.
Indexes are unmanaged, do not incur management fees, costs and expenses and cannot be invested in directly. For more information on indexes please see www.schwab.com/indexdefinitions.
Past performance is no guarantee of future results and the opinions presented cannot be viewed as an indicator of future performance.
Investing involves risk including loss of principal.
All corporate names are for illustrative purposes only and are not a recommendation, offer to sell, or a solicitation of an offer to buy any security.
The Schwab Center for Financial Research (SCFR) is a division of Charles Schwab & Co., Inc.