This is a special 2020 U.S. Elections focused guest blog from Tiago Quevedo Teodoro, a Quant at Refinitiv partner, MarketPsych.
Markets don’t like uncertainty. That is one of the many adages in the financial world, but one of the few that proves valid. After all, when investors are unsure of the probability of returns, they demand a higher premium.
To check for the relation between uncertainty and market returns around U.S. presidential elections, we looked at the S&P 500 level 90 days before and after the past 23 elections (thus going back as far as 1928).
The median level of the main index is shown below as the dashed blue line in the upper sub-plot (right axis).
As it appears, the market keeps moving sideways for most of the 90 days leading to the election.
A rally then starts about 10 days before the election (yielding ~8% in the following 100 days).
And it happens that this cycle is indeed correlated with the level of uncertainty in the US, as measured by the Uncertainty Refinitiv MarketPsych Index (RMI).
This series measures the amount of references in the news and social media to uncertainty and confusion. The data goes back as far as 1998.
The blue line in the upper plot indicates the average level of uncertainty in the last 5 U.S. presidential elections.
The individual values for each election are displayed in the lower sub-plot.
Overall, uncertainty rises considerably for most of the period leading to the elections, and it peaks about 2 weeks before the voting day.
It is very simplistic to infer causality, but if there is some, one could expect that the “uncertainty crash” in the current election will have a more substantial effect than usual.
As the blue line in the lower sub-plot shows, this is the presidential election with the highest level of uncertainty since the RMI coverage starts.
The 30-days average is about two standard deviations from its historical mean.
Hence, all else being equal (and there are a lot of “else’s”), a rally following the fall in uncertainty would not come as a surprise.
The Economy and Trump’s Odds
When it comes to re-elections, the odds of the sitting president being re-elected are affected by the current economic conditions.
Even more so when the president focuses his campaign message on his economic performance.
Besides sentiment topics such as uncertainty, the RMI feeds also provide data regarding factual themes.
One example of such themes is Business Expansion, which measures references of business expansions net of contractions in the news and social media.
This index provides a real-time proxy for business conditions within a country.
Interestingly, as shown in the plot above, the level of this index during 2020 has been tightly correlated with the odds of a Trump victory (tests even showed causality, but let’s not get too bold).
The odds were compiled from betting websites.
The chart shows that Trump’s re-election chances suffered two main hits, first during the beginning of the COVID-19 crisis and second during the U.S. social unrest in June 2020.
In both cases, the business expansion RMI fell along with the progression of those events.
In July, the media started reporting a less gloomy condition for business in America, and the Business Expansion index started rising.
By August, bettors were getting more optimistic on Trump’s re-election odds.
Based on this indicator, the U.S.’s recent spike in COVID-19 cases and the failure of fiscal stimulus negotiations portend worsening odds for Trump’s re-election.
For a comprehensive overview on our election coverage, including updates on the Global Markets Forum, webinars, data insights and more follow our U.S. Election page.
MarketPsych has comprehensive finance-specific sentiment data, covering all major countries, currencies, equity sectors, and individual US and non-US equities.
