The initial US stock market reaction to the COVID-19 crisis as it hit the United States was swift and brutal, and the recovery was nearly as fast. The S&P 500 went from record highs in early February 2020 to an almost 35 percent plunge by late March, then was back at record highs by mid-August.
But the reaction of individual stocks varied enormously, according to research by Chicago Booth’s Steven J. Davis, Imperial College London’s Stephen Hansen, and University of Chicago PhD student Cristhian Seminario-Amez. They find that risk exposures described in annual regulatory filings from before the pandemic helped predict how companies were affected by coronavirus news and how they might react in the longer term.
To quantify the discussions in the regulatory filings, the researchers used two text-analytic approaches: the dictionary method, which applies expert-curated term sets to search for specific content, and the supervised machine-learning approach, which uses an algorithm to select and tally relevant terms within a document, allowing for deeper interpretation of the information.
Davis, Hansen, and Seminario-Amez focused on 17 key trading days from February 24 to March 27, 2020, when the broad stock market rose or fell by at least 2.5 percent. The forces behind large one-day market moves are usually apparent, the researchers argue.
On days when bad news about COVID-19 drove the broader market, companies performed in line with the specific risk factors outlined in their 10-K filings, the research demonstrates. For example, bad COVID-19 news led to stock-price declines for companies in aircraft production, brick-and-mortar retail, energy supply, residential construction, and travel, and for REITs and restaurants. The same news powered higher stock prices for companies in basic foodstuffs, drug trials, e-commerce, health-care policy, metals and minerals, video games, and web-based services, as well as ones that influence the supply chain for cloud computing, electrical equipment, and semiconductors.
These stock-price reactions highlight the effects of downstream demand shocks on upstream suppliers, the researchers write. Aircraft production and energy supply were hurt by the fall in travel demand, while there was a positive effect for suppliers to businesses that were able to capitalize on social-distancing and work-from-home trends.
Company-level stock-price reactions to COVID-19 news in early 2020 helped to predict corporate earnings surprises later in the year. They also foreshadowed other broad economic shifts. For example, the traditional retail sector’s weak returns were followed by significant job losses related to the pandemic, while online shopping and delivery companies increased employment. Airlines were devastated, oil and gas companies went bankrupt, and print media ad revenue plunged. Meanwhile, cloud computing and digital services related to the work-from-home boom generated high demand.
This is only the beginning of the coronavirus fallout, and its economic and social effects will most likely continue for years to come, the researchers note. Interpreting company-level risk factors now may help us fathom at least some of what the postpandemic future may bring.