As it attempts to stem the coronavirus epidemic, the United States can expect to see its economy shrink by 9 percent in the second quarter of 2020, relative to the same quarter in 2019, according to research by Northwestern’s Scott R. Baker, Stanford’s Nicholas Bloom, Chicago Booth’s Steven J. Davis, and Stephen J. Terry of Boston University. The researchers predict a peak contraction of 11 percent in the fourth quarter of 2020, relative to a year earlier.
Trying to forecast the scope of the economic damage inflicted by the coronavirus presents economists with considerable challenges, as the COVID-19 crisis has few, if any, historical analogs, and the situation has evolved and escalated rapidly. To surmount these obstacles, Baker, Bloom, Davis, and Terry examined economic uncertainty in three ways: through newspaper coverage of the economy and equity markets, surveys that capture business leaders’ expectations about future sales growth, and stock market volatility. To estimate the economic impact of the COVID-19 crisis, they feed data on the stock market drop and the rise in its volatility into an empirical model of disaster effects developed by Baker, Bloom, and Terry.
To quantify uncertainty reflected in news about the economy and markets, the researchers used resources developed in previous research, much of it conducted by Baker, Bloom, and Davis, sometimes with other colleagues. One source is the daily and monthly US Economic Policy Uncertainty (EPU) indexes, which indicate the frequency of articles—drawn from about 2,000 US newspapers—containing one or more terms related to the concepts of economics, policy, and uncertainty. The daily index went from a monthly average value of around 100 (roughly normal uncertainty) in January to almost 400, a record high, in March. The monthly index showed a similar spike.
The EPU indexes, as well as another newspaper-based index, the Equity Market Volatility Tracker—developed to reflect the frequency of articles about the economy, markets, and market volatility—also provide a glimpse of how much uncertainty can be attributed to the pandemic.
In a dramatic shift from previous infectious disease outbreaks, which have historically had modest effects on financial markets, “COVID-19 developments receive attention in more than 90 percent of all newspaper discussions of market volatility and policy uncertainty,” according to other, recent research—by Baker, Bloom, and Davis, along with PhD candidates Kyle Kost at the University of Chicago, Marco Sammon at Northwestern, and Tasaneeya Viratyosin at the University of Pennsylvania.
Spiking uncertainty in executives’ sales expectations
Surveys capturing US business leaders’ initial reactions to the COVID-19 crisis find rising uncertainty in their outlook for their companies’ sales over the next year.
Altig et al., 2020; Survey of Business Uncertainty by Federal Reserve Bank of Atlanta, Stanford, and Chicago Booth
Surveys of executives provide another measure of economic uncertainty. The Survey of Business Uncertainty, developed by Davis and Bloom with colleagues at the Federal Reserve Bank of Atlanta, tracks US companies’ expectations for future sales growth.
The Decision Maker Panel, created by a team of researchers including Bloom, does the same for companies in the United Kingdom. Both surveys indicated spikes in sales-growth uncertainty in March 2020 that went “well above any previous peak in their (short) histories,” write Baker, Bloom, Davis, and Terry. Despite these historic highs, the researchers suggest the values derived from the surveys understate the true magnitude of business uncertainty, as uncertainty grew especially rapidly in the last half of the month.
Finally, US stock market volatility has been an especially visible reflection of uncertainty since the outbreak of the virus. The researchers note that the CBOE Volatility Index (known as the VIX), which measures investors’ expectations about future volatility, grew by roughly 500 percent from January 15 to March 31.
The researchers used that volatility in Baker, Bloom, and Terry’s model to predict the magnitude of the contraction that the US economy is likely to experience. The model looks at “first-moment” and “second-moment” effects of a disaster—that is, the initial economic shock that follows the disaster’s strike, as well as the sustained rise in uncertainty that follows. The researchers based the first-moment effect on the S&P 500’s 28 percent plunge between February 19 and March 31, and the second-moment effect on the rise of the VIX over the same period.
The findings forecast a peak GDP contraction in the fourth quarter of 2020. About half of this projected contraction is due to COVID-19-related uncertainty. What’s more, the researchers suggest,“There are reasons to think that our illustrative exercise understates the likely output effect of the COVID-19 pandemic”—for instance, the broad closure of schools and the shift to working from home for many workers may further hinder productivity, they write.