In troubled times people take comfort in the familiar. Covid-19 has upended many things, but tech-stock prices have proved impressively invulnerable. The Nasdaq, a tech-heavy stock index, has leapt by 25% since the beginning of 2020, taking its total rise over the past decade to over 400%. Were it not for a handful of tech giants like Apple and Microsoft, the S&P 500, another share-price index, would be down so far this year. Not since the boom of the late 1990s have technology firms inspired such exuberant trading. For punters the comparison should be a sobering one; after a peak in March 2000 the Nasdaq crashed, eventually losing 73% of its value. But the economic differences between the two eras should be more unsettling than any market similarities.
The two booms do share features beyond their stock-price trajectories. Both were sustained by inflows of new money. In the late 1990s discount brokerages and online-trading platforms drew in amateur punters looking to profit off the seemingly one-way market. Today, an army of small-timers trade shares and derivatives on new platforms like Robinhood. In the 1990s raging bulls justified high prices by declaring the dawn of a new economy, built on more powerful computers, fancy software and the internet. Today’s optimists cite the potential of everything from cloud computing and artificial intelligence to electric vehicles and blockchain. At first glance the economic performance seems similar too. In the late 1990s the unemployment rate fell to 4% and pay soared. On the eve of the pandemic, America’s jobless rate stood at a half-century low and wage growth, after a dismal decade, had accelerated to its best pace since 2008. According to figures published by the Census Bureau on September 15th, real median household income grew by a very healthy 6.8% in 2019.
Yet in critical ways the two episodes look profoundly different. As the 1990s dawned economists were hunting in vain for the efficiency-enhancing effects of new technology. Robert Solow, a Nobel prize-winning economist, quipped in 1987 that “you can see the computer age everywhere but in the productivity statistics.” By mid-decade that was no longer the case. Output per hour worked in America rose by more than 3% a year in 1998-2000, a feat the economy had not pulled off since the early 1970s. Growth in total factor productivity (a measure of the efficiency with which capital and labour are used, often treated as a proxy for technological progress) rose by about 2% a year from 1995 to 2004, according to Robert Gordon of Northwestern University. That was a sharp pickup from the average pace of 0.5% in 1973-95, and nearly matched the rate achieved during the heady growth years of 1947-73.
Productivity in the 2010s, by contrast, looks pitiful. Annual growth in labour productivity has not risen above 2% since 2010. Growth in total factor productivity, according to data gathered by John Fernald of the Federal Reserve Bank of San Francisco, has been more dismal than ever: just 0.3% on average from 2004 to 2019. If you take the 2010s alone, the average falls to just 0.1%.
Strong labour productivity growth in the 1990s enabled wages to rise without squeezing corporate profits. While the dotcom boom is often remembered for the enormous valuations achieved by profitless upstarts with no clear path into the black, after-tax corporate profits during the decade rose from 4.7% of gdp in 1990 to 6.7% in 1997, before closing the decade at 5.6%. Corporate profits actually declined as a share of gdp during the 2010s, albeit from a much higher level than that prevailing in the 1990s: from 10.4% in 2010 to 9.0% in 2019. More telling, however, is the way in which firms responded to profit opportunities during the two decades. Investment in computer equipment, software and r&d leapt during the 1990s, by 1.5 percentage points of gdp over the decade. In the 2010s, despite the much higher level of profits, investment rose by just 0.7 percentage points of gdp.
The exuberance that powered soaring stock prices in the late 1990s, if in some cases irrational, occurred alongside tech-powered structural change. The uptick in productivity was at first driven by advances in computer-making. As prices tumbled and capabilities soared, other firms began investing in new equipment. Productivity gains began to spread across the economy, helping firms streamline manufacturing and transforming critical industries. These persisted, and even accelerated, for some years after the market crashed. Though many dotcom darlings disappeared, the digital infrastructure built during the boom remained. So did a number of firms that came in time to dominate the corporate landscape. In March 2001 The Economist grimly assessed the prospects of Amazon, a struggling retailer that had lost 90% of its market value in the crash, noting that “even if such companies survive, they are unlikely to resemble the businesses they once were.” (Holding Amazon through the crash proved a smart bet; its stock now trades at about $3,100, up a tad from under $10 in 2001.)
Only nineties kids will remember productivity
Some of today’s high-flyers will in time prove to be good investments. Optimism about the real economy requires a bit more faith. There are grounds for hope. Some economists reckon that hard-to-measure “intangible” investment—such as time spent re-engineering business processes—takes up a growing share of firms’ energies. If so, both investment figures and future economic prospects could be undersold.
Both output per hour and total factor productivity accelerated in 2019. Though it remained well short of the 1990s, this uptick might presage an economic transformation in the making. And the covid-19 pandemic has imposed constraints on business activity, which might in turn accelerate tech-driven restructuring. The possibility has probably contributed to the surge in tech stock prices since March. For now, technology valuations are based, to a far greater degree than in the 1990s, on what could be rather than what is. Invest accordingly.
September 19, 2020