Vanguard’s chief economist says artificial intelligence, automation stymieing market forecasters
Published: Sept 8, 2017 12:19 p.m. ET
Joe Davis, Vanguard’s global chief economist, argues that a burgeoning technological revolution is a big risk factor for the market
For years, the U.S. stock market has been moving on issues that would be familiar to investors of nearly any era: questions over monetary policy, uncertainty stemming from the political and economic environments, concerns over valuation and growth, jobs reports and corporate earnings. However those catalysts, both good and bad, aren’t addressing the most important factor facing the economy, one that will fundamentally change the way all people and businesses go about their businesses.
That view is according to Joe Davis, Vanguard’s global chief economist, who argued that the factors moving markets on a day-to-day basis, or even a month-to-month, or year-to-year basis, were minor compared with a trend that may be less visible when looking at data or balance sheets.
“The forces that trump all others are human capital and technology, man and machine, and we’re at an inflection point with them, in the midst of a structural change that will impact the labor market globally,” he said. “The trend that will define our lives is how work changes for all of us. I can’t think of a more seminal issue that will face the economy in the years ahead.”
Changes in technology, primarily trends involving automation, robotics, and artificial intelligence, were behind what Davis dubbed “the three big paradoxes” in the market that have been stymieing economists. The paradoxes are an environment marked by low inflation, but nearly full employment; low growth, but high valuations; and high political uncertainty, both in the U.S. and abroad, but also “an eerie calm over the market” in the form of low volatility.
Davis suggested these arose from an economic variation of Moore’s Law, which argues that computing power will double in its ability every couple of years. “Technology lowers the cost to produce everything, and tech itself gets cheaper every year,” he said. “The more we use technology, the harder it will be to have 2% inflation.” He estimated technology subtracted 50 basis points a year from inflation, in addition to having an impact on economic growth, wages, and standards of living.
“It would be a mistake for any of us to be complacent about any of this,” he said. “Tech is always moving faster and getting smarter. Artificial intelligence and computer programs are doing more things and, more importantly, they’re learning. Work has always been an arms race between education and technology, but the pace of change, and the scale of the change, are different this time.”
–– ADVERTISEMENT ––
Many have sounded the alarm about automation and the growing ability of machine to perform the jobs that had previously been manual labor. Robert Shiller, the Nobel Prize-winning economist, said this factor was the one that most concerned him about the economy, calling artificial intelligence “a deeply challenging thing,” adding, “people who are growing up today don’t know if they’re preparing for the right career, and challenges associated with that could lead to secular stagnation.”
More detail: Thjs is what scares Nobel Prize-winning economist Robert Shiller about the economy
The impact of automation and AI could be massive, particularly as new technologies—such as driverless cars and “bot” programs that can mimic human speech patterns—mature and disrupt ever-larger segments of the economy. According to recent research, every industrial robot takes up to six jobs, with some 6 million jobs at risk of being lost to automation over the coming decade. Last year, the White House’s annual economic report of the president (under President Barack Obama) forecast an 83% chance that automation will take a job with an hourly wage below $20, a 31% chance automation will take a job with an hourly wage between $20 and $40, and just a 4% chance automation will take a job with an hourly wage above $40.
Vanguard’s Davis cited work from Carl Benedikt Frey and Michael Osborne, professors at Oxford University who suggested that as early as 2022, 47% of U.S. jobs—which would translate to about 70 million people—could be automated. Among other major economies, 69% of jobs in India and 77% of jobs in China could face the same fate. While Davis said he didn’t think the impact would be that severe, he used it as an example of how negative forecasts had become. “Imagine a world where three-fourths of Chinese works are out of a job,” he said. “If that’s the future, who cares who the next Fed chair is? I don’t know who will be buying ETFs in that world. Maybe the rich machines.”
He forecast four future “paradoxes,” all related to technology revolutions: the economy will see more automation, but higher labor shortages at the same time; those labor shortages will exist alongside low inflation; that low inflation will persist even as real interest rates rise (he said this was the forecast he was least confident about); and that even with higher real interest rates, equity market returns will be lower. “No doubt there is higher risk in the equity market right now than the bond market,” he said.
Currently, Vanguard’s outlook for future returns is at its lowest level in 10 years, since the financial crisis was accelerating. (The outlook isn’t bearish, nor does it call for losses; it simply anticipates returns that are below historical norms.)
Despite the scorched-earth nature of these forecasts, Davis suggested there was room for optimism that humans wouldn’t become obsolete, even as the work they do changes dramatically.
This view comes from distinguishing between the job that a person does, and all the individual tasks that make up that job. He listed three types of task: basic, such as recording information or moving objects; repetitive, including assembling items and processing information; and advanced tasks, which include problem solving, strategizing, and developing teams. Based on an analysis of labor market data, Davis estimated that over 1,000 occupations, jobs require an average of 41 tasks.
“Jobs don’t get automated away, tasks do,” he said. “The jobs most at risk [of being fully automated] are the ones focused on a few repetitive tasks.” In comparison, jobs focused on advanced tasks -- what he called “uniquely human tasks”--“can’t be automated away, even by a really smart computer.”
Most workers will have experienced these changes already, as what was formerly busywork has gotten automated by word processing, email, or other technologies that have been incorporated into one’s daily life. Davis estimated that there has been “a 50% change in what we do day-to-day since 2000, and it would be wrong to think that some of these changes haven’t complemented workers, in addition to disrupting them.” While every job title has seen changes, there aren’t easy correlations between the degree to which one has changed, and how high-paying that job is. As an example, Davis noted that doctors from a century ago were limited in the medicines they could prescribe or the treatments they could use. Now, their job is very different, but there is still demand for it.
Future job skillsSo, what skills will future job seekers need?
What automation will change, in Davis’s view, is the number of advanced tasks that jobs require, relative to basic or repetitive ones. “For the first time ever, we spend 50% of our time on uniquely human work, compared with just 30% in 2000,” he said, estimating that this could reach 80% in the future. Jobs that are weighted toward basic and repetitive tasks could be lost—Davis noted there are 400 occupation titles recognized by the U.S. government that no longer exist—but “the job losses will be dwarfed by demand for new jobs in uniquely human tasks.” The U.S., he added, “is in the best possible seat for this,” noting that while China and India had massive workforces, their economies were more heavily weighted toward basic and repetitive tasks.
“I don’t care what my kids major in, but I want them to learn three skills: creative intelligence, technological acumen, and emotional intelligence,” he said. “Thinking creatively is the top requirement for a ton of jobs today, everything from chefs to engineers. Social skills -- things involving perception and empathy—will arguably be the most important skill in the future, and it can’t be automated away.”
“A great teacher,” he added, “doesn’t just have a collection of facts they can recite; they can inspire. That’s something a computer will never be able to do.”
More from MarketWatch