Productivity holds the key to betting on AI

Investors deserve to enjoy the AI party. But it will not be too late to enjoy it after watching the long-awaited productivity scorecard first.

Kim Young-ikThe author is an adjunct professor of economics at Sogang University Graduate School of Business. The increased productivity from the information technology (IT) revolution in the 1990s helped the U.S. economy achieve high growth and low prices at the same time. Stock prices soared under the “new economy,” as shrewdly defined by some economists. The enhanced productivity from the artificial intelligence (AI) revolution — nearly on par with that from the earlier revolution — fuels the expectation that America will re-enact the new economy. Stock prices reflect such optimism preemptively. But investors had better rein in their overconfidence bias this time, too. If data fails to show heightened productivity from the AI revolution, a stock market bubble can burst anytime.

According to the National Bureau of Economic Research — a private and non-profit organization dedicated to conducting independent economic research — the U.S. economy was in the expansionary phase of the cycle for a whopping 120 consecutive months from March 1991 to March 2001. Before 1991, the expansionary phase lasted 35 months on average since 1857, when the non-partisan research body started to check U.S. business cycle expansions and contractions.

The reason for the longest-ever expansionary phase in the U.S. economy can be found in the IT revolution in the 1990s, starting with the computer technology development. Early computers — enormously costly and massive — were only available to a precious few. But the advent of personal computers enabled the multitudes to access them. The emergence of the internet, which prompted computer networking, further raised productivity in various fields of society.

Information technology lifted U.S. labor productivity remarkably. Labor productivity in the non-agricultural sector rose 2.2 percent between 1991 and 2000 after showing an average increase of 1.5 percent during the precious decade. In 1996-2000, in particular, the average labor productivity surged as much as 2.9 percent, nearly doubling from the 1980s.

Increased productivity means an ability to produce products at cheaper prices than before. Enhanced productivity across nearly all manufacturing industries escalated the U.S. growth rate while lowering the inflation rate. As a result, the average quarterly growth rate of the U.S. economy — which was 2.9 percent between 1980 and 1995 — soared to 4.3 percent for the following 15 years. And yet, the personal consumption expenditure price index (PCEPI) fell to 1.7 percent from 4.2 percent during the same period to signify the simultaneous achievement of high growth and low prices from the IT revolution. Some economists gladly called it a “new economy” or “Goldilocks economy.”

U.S. households' increased consumption from such optimism about the U.S. economy contributed to extending its expansionary phase. While the household savings rate at 6.9 percent in 1995 dropped to 4.2 percent in 2000, the share of household debt in disposable income shot up to 98.3 percent from 90.3 percent during the same period. In other words, U.S. citizens cut their savings to consume more even by borrowing money from banking institutions. Companies in the manufacturing and tertiary sectors, except for financial firms, also took out more loans for investment, as seen in the swelled corporate debt-to-GDP ratio from 80.4 percent in 1995 to 87.3 percent in 2000.

After the IT revolution in the U.S. worked the miracle of grabbing both high growth and low prices, share prices skyrocketed. The S&P 500 more than doubled to 1516.34 by the end of April in 2000 from 671.48 at the end of 1995. In particular, the Nasdaq — a barometer of the new economy — jumped more than fourfold from the end of 1995 to February 2000, pointing to a bubble in the U.S. stock market. An evaluation of the S&P 500 through the productivity index shows it was overvalued about 66 percent in the first quarter of 2000 alone.

In 2000, the bubble burst on the New York Stock Exchange. The Nasdaq, hovering at 4696.69 at the end of February 2000, plunged to 1172.06 by the end of September 2002. Since then, productivity also started to slow down. Between 2001 and 2007, the U.S. productivity increase rate in the nonagricultural sector declined to an average of 2.6 percent. That lowered the U.S. growth rate to 2.5 percent while pushing up inflation to 2.3 percent.

The productivity slowdown was worse after the 2008 global financial crisis. The average labor productivity increase rate regressed to the pre-1995 level at 1.5 percent. Lowered productivity drastically reduced the average growth rate of the same period to 1.9 percent, while escalating inflation to 2.1 percent.

The AI revolution covers all aspects of society, including the economy. The revolution refers to “enhancing efficiency and productivity by combining better computers with AI.” AI will change production methods in all fields and radically elevate productivity. The manufacturing sector already automated production processes and replaced workers with robots. The dazzling advancement of robotics by AI will accelerate the shift. The era of no humans in factories may arrive pretty soon.

AI-applicable areas will be further extended in the future. Just think of fully autonomous vehicles by connecting AI and the Internet of Things or AI-powered rob-advisors for big data-based investments, who will replace existing fund managers. AI will provide personalized healthcare services once it is connected to big data on medicine. At the Mobile World Congress in Barcelona, Spain, from Feb. 26 to 29, Google DeepMind CEO Demis Hassabis said, “We’ve just signed big deals with Big Pharma and on real drug programs. And I expect in the next couple of years, we’ll have AI-designed drugs in the clinic, in clinical testing.” He added that the average 10-year period needed to develop a new drug will be cut to a few months in the future. The intellectualization and sophistication of military weapons through AI also will drastically change future war patterns.

Above all, the increased productivity from the AI revolution will have a huge impact on the economy. The United States can catch the two rabbits — high growth and low prices — simultaneously, just like it did thanks to the IT revolution in the late 1990s. The NBER believes that the ongoing 48-month expansionary phase since April 2020 can be extended if productivity rises.

But sadly, data showing AI-prompted productivity hikes has yet to be released. Last year, U.S. labor productivity rose only 1.4 percent, lower than the past average. In the manufacturing sector, in particular, productivity rose just 0.5 percent, even lower than the 4.4 percent average between 1996 and 2000.

The U.S. economy needs to go the extra mile on the demand side. In 2022 and 2023, the household savings rate was 3.3 percent and 4.5 percent, respectively, lower than the 6.3 percent average in 1990-2021. This January and February, the savings rate was 3.9 percent, still lower than the average. Last year, the share of the household debt-to-disposable income was 100 percent, still not low. Real median household income, a key criterion of U.S. consumption, decreased between 2019 and 2022. The share of corporate loans and bonds in the U.S. GDP was 48.8 percent, still higher than 46.2 percent at the time of the IT bubble burst.

Stock prices have already entered the bubble sphere thanks to rosy expectations about the AI revolution. A regression analysis of the S&P 500 based on the labor productivity index shows that stock prices were 53 percent overvalued as of the end of 2023. At the height of the IT revolution, tech share prices were 81 percent inflated in the second quarter of 1998. Though they are not so much overvalued this time, they are certainly over-inflated.

If productivity rises as the stock market hopes, the bubble on the stocks can last longer. But if productivity data falls short of the market’s expectations, stock prices could plummet. Investors deserve to enjoy the AI party. But it will not be too late to enjoy it after watching the long-awaited productivity scorecard first. Translation by the Korea JoongAng Daily staff.