Jeremy Siegel – The Market is Not in a Bubble
Membership required
Membership is now required to use this feature. To learn more:
View Membership BenefitsAmid fears of a recession, the S&P 500 is up more than 14% this year, and artificial intelligence (AI) stocks have risen many times more than that. But the market is not in a bubble, according to Jeremy Siegel.
Siegel is the Russell E. Palmer Emeritus Professor of Finance at the Wharton School of the University of Pennsylvania and a senior investment strategy advisor to Wisdom Tree Funds. He spoke yesterday via a webinar hosted by Wisdom Tree.
The focus of his talk was AI, but I’ll cover his remarks about the market first.
Valuations are “miles away” from those at the peak of the dot-com era, Siegel said. “These are real companies,” he said, and not the product-less dreams of dot-com entrepreneurs.
The price-earnings ratio of the Nasdaq is 30, versus 100 that it reached during the dot-com peak, according to Siegel. The P/E of the technology sector of the S&P 500 was 80 then, versus approximately 35 now. Interest rates were much higher then, with TIPS yields at 4%.
“We use the term bubble too quickly,” he said. That term should apply to valuations that are 200% to 400% above fair value, which might not even apply to Nvidia, he said. That stock has tripled since late last year when the AI boom started to gain popularity.
“This can go on a lot longer,” Siegel said. Momentum in the market is very strong and can persist for the long term. “But pullbacks can be very quick,” he said.
Momentum requires a story plus some concrete evidence, he said. In the latter case, it was the release of ChatGPT late last year.
As evidence of the power of momentum, Siegel cited the so-called meme stocks, which he called a “game” of valueless companies. But it has proved difficult to make money by shorting those stocks.
“Momentum, not valuation, can make the biggest difference,” Siegel said.
Is AI a good investment?
Capital investment in AI has soared. Approximately $14 billion was dedicated to AI 10 years ago, but in 2021 that number reached $276 billion. The U.S. is responsible for 54% of that spending, followed by China at 15%.
Siegel said he was puzzled by the adjective “artificial.” It would be more appropriate to call it machine intelligence (versus human intelligence). Calling it artificial makes it sound “fake,” he said.
AI has proven to be superior to humans in many ways, such as the ability to perform computations and play games like chess and Go.
Human intelligence has the advantage of being able to weigh judgment in decision making, although we are not always correct. But we learn the same way, he said. The difference is that AI has access to more experience that it case process faster.
Siegel warned against investing in AI-related stocks.
The “first movers” of the dot-com era are no longer with us: Netscape, Mosaic, Wordstar, WordPerfect, Lotus 123 and Harvard Graphics. Intel is now a poor second to Nvidia as a chip supplier.
“First movers don’t always survive,” he said.
He also cautioned against investing further down the value chain in so-called pick-and-shovel companies (a term that dates from the gold rush, when the gold miners were the ones who achieved disproportionate success). He questioned whether Nvidia will sustain its high valuation (a P/E of greater than 200). For Nvidia to grow and increase earnings, it will need to be part of products and services that gain a greater “share of wallet” from consumers. Siegel was skeptical that will happen.
In terms of individual stocks, Siegel said the world is a “roller coaster.” One of the first dot-com companies was Amazon, when it was just selling books. It soared in value and hit a high price in 2000. Then it fell 50%, at which point Barron’s magazine had the cover story, “Amazon Dot Bomb.” It then went down another 90% to 95%, only to subsequently rise and become what it is today. A similar story can be told of Apple, Siegel said.
“Investment themes are part of a cycle,” Siegel said, and you must be right at the right times.
AI and the macroeconomic outlook
Siegel acknowledged the power of AI but was skeptical as to whether it will represent breakthroughs on a par with the Industrial Revolution or the internet and technology revolution of the last 50 years.
The central question for AI is whether it will replace jobs and disrupt industries.
Siegel was skeptical.
Autonomous driving would be amazing, he said, and could replace truck drivers and make automobile transportation more efficient, with less congested highways.
But there will be legal obstacles. Human driver liability is protected by automobile insurance. But with autonomous driving, those in an accident will look for the “deepest pocket,” Siegel said. That will be Tesla and its chairperson, Elon Musk.
Siegel said AI will replace primarily low-paying jobs. Those could be truck or taxi drivers displaced by autonomous drivers, or healthcare workers replaced by robots. But there can also be displacement among highly paid workers. Surgeons, for example, could use AI to analyze complex medical conditions to determine possible courses of action faster than is now possible.
AI will be deflationary, as is the case with technological improvements in general.
Siegel also said that AI might not improve productivity, and productivity has declined slightly over the last 50 years. On this point, I disagree. Productivity is measured by output per unit of labor, but the conventional metric for output (GDP) is inaccurate.
As Woody Brock has explained, GDP does not capture the advances that vastly improved our standard of living. In my lifetime, we have improved automobile safety, saving millions of lives, and have developed statins, extending lifetimes by many years. Air conditioning is commonplace, making life more comfortable and reducing disease from mold.
In his talk, Siegel acknowledged some of those improvements. He said he does his writing on a word processor, instead of the typewriter he used when his academic career began. But the transformation and increased output from the typewriter to word processing is not captured in GDP data.
A final question remains as to what separates AI from human intelligence. What is the domain of thinking that is only possible by humans? AI has proven that it is possible for machines to learn, mostly through pattern recognition amplified by its ability to analyze vast amounts of data at a very high speed.
But creativity is different. Can AI compose a piece of music that is distinctly different from anything that preceded it? Or create a piece of art that represents a new school of drawing? True creativity is not an incremental improvement or variation. It is something distinctly new, and not calculable from experience. This is why I don't expect AI to succeed at active management. Beating a passive portfolio on a consistent basis requires more than the algorithms, high-speed processing and vast amounts of data that characterize AI. Even if AI has this advantage, it would soon be arbitraged away. It takes the extremely rare element of creativity to earn an advantage in the markets.
Robert Huebscher is the founder of Advisor Perspectives and a vice chairman of VettaFi.
Membership required
Membership is now required to use this feature. To learn more:
View Membership Benefits