While examining the difference between stock performance of firms in Republican and Democratic counties, they found that red-county stocks performed significantly better than blue-county stocks on important COVID-19 news days.
In the seminar above, Jinfei Sheng summarizes their findings and explores the relationship between COVID-19 and the stock market. He also briefly covers how Safegraph and Placekey empowered their team to complete this study.
The Partisan Return Gap: Republican-county stocks earn more than Democratic-county stocks on COVID-19 news days
Sheng, Sun, and Wang examine the influence of political polarization on the stock market during the COVID-19 pandemic. They found that on average, red-county stocks earned 18 basis points higher returns than blue-county stocks on days with important COVID-19 news; what they call the Partisan Return Gap.
Using location and foot traffic data provided by Placekey and Safegraph, they were able to determine people’s attitudes towards the pandemic; if people are going to non-essential places, they are therefore less concerned about COVID-19. From this, they were able to further demonstrate that stocks in counties where investors conduct less social distancing behavior have higher returns on COVID-19 news days.
There was no strong evidence to support the argument that the gap was correlated to local economic conditions or the firm’s fundamentals. They also found that while the gap exists for both positive and negative news days, the gap is larger for positive news days.
5 key takeaways from this seminar on the polarization of the stock market during COVID-19
In the seminar, Sheng walks us through the main findings and process of their study. We’ve extracted the five main takeaways below:
1. A behavioral analysis of the partisan return gap in the stock market
Time in seminar video: 3:30
This study was based on recent evidence that disagreements are particularly partisan in the United States. This carried over into reactions to COVID-19, with Republicans being generally less concerned about COVID-19 than Democrats, as evidenced by Republican states having stricter lockdown policies and looser requirements on social distancing practices.
For Sheng, Sun, and Wang’s study, they approached it with this information in mind. They then worked towards finding a behavioral explanation about how this impacts the partisan return gap on stocks.
2. Identifying important COVID-19 news days and linking them to stock market data
Time in seminar video: 12:19
Central to their study was understanding how companies react differently to COVID-19. To do this, they identified big COVID-19 shocks, as reflected by large shifts in the stock market that correspond to important news about COVID-19. They first identified big moves on the S&P500, which was any move larger than 2.5%.
They examined these big move days, and studied the big news cycles on those days, analyzing whether these stock market moves were being impacted by the COVID-19 news. Throughout their study, they identified 33 days with big moves, and determined that 28 were driven by major COVID-19 news.
3. Examining how negative and positive COVID-19 news stories impact the stock market
Time in seminar video: 14:12
With these big move days mapped on a graph that shows information about the pertinent news of that day, they could further analyze relationships in the data.
You can clearly see that negative news about COVID-19 (such as increased infection rates, overburdened healthcare system, significant policy changes, etc.) leads to drops in the stock market, and that positive news about COVID (such as plans to combat the virus, improved infrastructure, increased funding and support, etc.) leads to increases in the stock market.
4. Findings confirm red-county stocks perform better on average than blue county stocks
Time in seminar video: 16:12
Having established that important news regarding COVID-19 does impact the stock market, they wanted to further examine how companies in different political areas react differently. Specifically, they chose to analyze the difference between how investment firms in red and blue counties performed.
To do this, they used a combination of aggregate data and specific comparisons that were possible within their datasets. One example allowed them to compare two natural gas companies in neighbouring counties (one in a red county, and one in a blue county). Overall, the red stock earned a daily average of 1.37% and the blue stock earned a daily average of -1.00%, showing the significant difference.
Their analysis also proved that negative COVID news dragged the blue stock down further than the red stock, and that positive news led to higher gains for the red stock, showing how this gap grew as time went on. This clearly shows this divide steadily increasing.
5. People’s attitude towards COVID risk determines stock market behavior
Time in seminar video: 24:13
Their study factored in the impact of confirmation bias, the idea that the news people watch (more often than not) confirms their pre-existing beliefs and notions. In relation to COVID-19, confirmation bias will impact whether people think that COVID-19 does or does not pose a higher risk to them, which will in turn contribute to their decisions and behavior.
First, they needed to prove the relationship between partisanship and risk attitudes towards COVID-19 in red and blue states. As evidenced in the news, the majority of Republicans saw less risk with COVID-19 than Democrats. Foot traffic and location data from Safegraph and Placekey were able to prove this by showing that there was less social distancing behavior exhibited in red counties than blue.
Second, they needed to prove the relationship between this risk attitude and stock market performance. If risk attitude towards COVID explains the partisan return gap, then firms in counties with less social distancing behavior should earn higher risk-adjusted returns on COVID-19 news days (which they were able to prove).
Safegraph data that tracked non-essential travel was able to help them measure people’s risk attitudes. From this data, they found that blue counties engaged in more social distancing behaviors, further proving that red counties and lower concerns about COVID-19 led to higher firm earnings.
The authors’ backstories & research
Jinfei Sheng is an Assistant Professor of Finance at the Paul Merage School of Business at the University of California, Irvine. He obtained his Ph.D. in finance from the University of British Columbia. His research areas include empirical asset pricing, investments, and behavioral finances. He has experience with big data, machine learning, textual analysis, and frequently employs these to study financial markets. Sheng also reviews papers admitted to top finance journals and conferences.
Zheng Sun is an Associate Professor at the Paul Merage School of Business at the University of California, Irvine. She started in 2007 as an Assistant Professor shortly after receiving her Ph.D. from Stern School of Business at New York University. She currently teaches the MBA curriculum and has published papers on a variety of topics related to empirical asset pricing, investments, market microstructure, and banking.
Wanyi Wang is a Ph.D. student in Finance at the Paul Merage School of Business at the University of California, Irvine. Her primary research interests are empirical asset pricing, FinTech, behavioral finance, machine learning, and textual analysis.
How Placekey enabled research on risk attitude towards COVID in red and blue counties
An important part of this study relied on understanding the difference in risk attitudes towards COVID-19 in both Republican and Democratic counties. To do this, Sheng, Sun, and Wang leveraged Safegraph’s foot traffic and location data, powered in part by Placekey’s universal location identifier.
Safegraph’s extremely accurate point of interest and foot traffic data enabled their team to analyze device movement, helping them determine social distancing behavior during COVID-19 based on people’s visits to non-essential places. This allowed them to determine people's risk attitudes towards COVID, which they used to analyze the stock market impact.