US election: are investors underestimating tail risks?

Not only are markets not accounting enough for the threat of a messy, contested election, they are also assigning too low a probability to the possibility of a Democratic landslide

Nicholas Spiro

Published: 10:00pm, 29 Oct 2020

Why you can trust SCMP

When it comes to elections, political analysts follow opinion polls. Investors, for their part, pay more attention to prediction markets, where bettors trade futures tied to different outcomes. Just days before the United States holds its momentous presidential election, an average of national polls compiled by RealClearPolitics shows that the race has tightened over the past few weeks.

RealClearPolitics’ average of seven polls indicates that Democratic candidate Joe Biden’s lead over president Donald Trump has narrowed from 10.3 points on October 11 to 7.5 on October 28. This is still a comfortable margin, and one that has proved more stable than Hillary Clinton’s shrinking lead in the final stretch of the 2016 campaign.

Yet, the tightening in the race has contributed to a significant shift in betting markets, notably with regard to the probability of a “blue wave” whereby the Democrats clinch the presidency and both chambers of Congress. According to Predictit, a political betting site, the Democrats’ chances of a clean sweep have fallen from 62 per cent on October 7 to 54 per cent.

Investors, who as recently as a fortnight ago were pricing in a blue wave, are also revising their assumptions on the outcome of contests in some of the key battleground states, where Biden’s margins are smaller. Predictit bettors now see Trump winning Florida – the swing state with the most electoral votes – by a hefty margin, and taking North Carolina and Ohio, two other key states.

04:54

The US Electoral College: how does it work and why does it exist?

In the space of a little over a month, prediction markets have gone from betting on a disputed election result to anticipating a Democratic sweep, only to start pricing in a more modest Biden win. The uncertainty is exacerbated by markets’ failure to predict Trump’s upset victory four years ago, and the lingering feeling among investors that next week’s election could throw up a surprise.

The nervousness among investment strategists is palpable. On October 27, JPMorgan published a report tellingly entitled “What if Trump wins?” Yet, in assigning a higher probability to an outcome that is less extreme than the two other scenarios investors were betting on over the past several weeks, markets are again in danger of underestimating the tail risks.

How Trump could end his presidency with a wild transition

Although many investors believe a contested election result is far-fetched, fears of a less than clear-cut outcome were showing up in futures contracts last month, and have not entirely dissipated. Trump still has paths to victory and, just as importantly, has done enough to sow distrust of the legitimacy of the electoral process, particularly given the surge in Covid-19-induced mail-in ballots.

The much tighter races in swing states, coupled with the controversy over postal votes (which are favoured more by Democrats than Republicans, and will take longer to count in several key states since they cannot be processed until election day), increase the scope for a disputed result.

A voter puts his ballot into a ballot drop box on October 28 in Seattle. Washington state is one of five states, along with Colorado, Hawaii, Oregon and Utah, that is conducting elections entirely by mail-in voting. Photo: AP

For markets, the question is not just whether support for Trump in some of the battleground states proves stronger than anticipated, but also whether the political and legal thresholds for a contested outcome are lower than investors anticipate. To put it another way, are a sufficient number of key swing states susceptible to a prolonged and contentious vote count to tip the balance in favour of a disputed election?

Just as investors are underestimating the threat of a messy, contested election, so they are assigning too low a probability to the other tail risk: a Democratic landslide. The dramatic increase in early voting – both in person and by mail – points to a record turnout which could benefit Democrats, given that many of the party’s traditional voters, in particular African-Americans, failed to turn up on polling day in 2016.

Worst place, worst time: Trump faces virus spike in Midwest

According to the US Elections Project, more than 75 million Americans had voted as of October 29, 17 million more early votes than were cast in all of 2016 and 54.8 per cent of the total votes cast four years ago. Moreover, in states with party registration data, Democrats accounted for 47 per cent of early votes, compared with 29 per cent for Republicans.

The uncertainty in this year’s election cuts both ways. While investors have been concerned about polling data that understates the support for Trump, they have been less sensitive to the risk that polls are underplaying support for Democrats. This is because markets have yet to fully recover from the shock of 2016. Investors are hypersensitive to any sign that Trump is gaining on Biden, and fear polling errors that benefit the president far more than those that favour Democrats.

In the final days of the campaign, prediction markets are playing it safe, betting on an outcome that occupies the middle ground between a contested result and a Democratic sweep. These wagers may well pay off. Yet, for an election that remains highly unpredictable, markets are in danger of playing it too safe.

Nicholas Spiro is a partner at Lauressa Advisory

This article appeared in the South China Morning Post print edition as: Markets in danger of playing it too safe in an uncertain election

Post