US Presidential Election Preview
FOREWORD
In this research report, we outline everything you need to understand about trading the US presidential election on election day in G10 FX markets.
The election has been described as pivotal in shaping the direction of markets. While as traders, we are naturally contrarians and hence bound to react to such hype with healthy skepticism, this is one of the rare occasions when the event will most likely be very volatile, with opportunities abounding.
Given the US presidential election occurs only once every four years, we decided to reveal some content from our historical databases leveraging data logs of trading the past presidential election events. In particular, we will discuss how the 2016 election was initially mispriced.
Of course, nothing in the reports constitutes investment advice and you must come up with your own conclusions on how the pieces of the election puzzle come together. Although the report primarily aims to target institutions, even a less experienced trader can walk away with ideas on how to trade, hedge or manage positions during the election.
CONTENTS
What is priced into the market?
Understand how the vote counting affects the markets
How will the probabilities shift in practice? A historical look
The US Elections in brief
The US will hold a Presidential Election on the 5th of November. On top of the presidential election, there will also be an election for the House of Representatives and the United States Senate.
In the unique US presidential election system, 270 electors are required for victory. Most of the electors, e.g. New York for Harris, Texas for Trump, are what we can call ’certain’ electors. In total, there are 226 certain electors for Harris and 219 for Trump.
There are seven key swing states, Arizona (AZ), Georgia (GA), Michigan (MI), Nevada (NV), North Carolina (NC) and Pennsylvania (PA). These states determine who wins the election.
Out of the swing states, AZ, GA and NC are leaning more towards Trump – 11 + 16 + 16 = 43 more for Trump (total 262).
NV, WI, MI and PA are very close and could go either way – 6 + 10 + 15 + 19 = 50 electors up for grabs.
So, Harris needs to win 44 electors to win, and Trump needs 8 (assuming the Trump-leaning swing states vote for Trump).
The most likely path to Harris's victory is grabbing all the WI, MI and PA. The most likely path for Trump's victory is winning any one of WI, MI or PA which seals the deal.
Therefore, no wonder betting market odds for Trump are looking high!
Candidate’s policies
There are two key elements to assessing each candidates’ policies, fiscal policy and trade. Both candidates are seen fiscally “looser” than Biden, reflecting stimulatory proposals (Harris less tax hikes than Biden, Trump outright extension of tax cuts) leading to a larger deficit and more government debt.
Trump is clearly seen as more stimulatory (and inflationary) than Harris. For this reason, expectations are that the Federal Reserve might not be able cut rates as much with Trump in office as with Harris, which is one of the reasons why Trump presidency is seen as a positive for the USD. Moreover, more fiscal stimulus comes with accelerated short-term growth, which is another positive.1
Finally, Trump is likely to implement tariffs on trade, including a 10% tariff on all imports and 60% tariff on Chinese imports. With Harris one can simply expect continuity of Biden’s policies, though moderate tariffs for Chinese goods are expected.
In terms of roughly assessing probabilities what policies will get implemented, there’s a rule of thumb that
1. A sweep (the President’s party controls both houses of Congress) will enable the President to pass any legislation more easily. With a divided Congress the enactment of FIPO (taxation & budget related) is harder and compromises must be made. Hence the house election is also particularly in focus.
2. Tariffs can be implemented via executive orders meaning the balance of power in Congress is less important
What is priced into the market?
At the time of writing, the market is pricing in roughly 65% chance of Trump victory depending on source and roughly 40-50% chance of Republican Sweep2. The market reaction on Election Day is initially based on how these probabilities swing, until potentially secondary logics step in.
The probabilities are contingent on
1) Initial “certain states” showing any surprises in terms of victory margins
2) Any of the Trump leaning states providing “surprises” for Harris
3) The outcome of the ever-important swing states i.e. WI, NI and PA. In base case, these electors decide the election and Pennsylvania is the most likely Kingmaker.
Note on prediction markets: A Bloomberg analysis of Polymarket data has shown that a significant portion of the rise in Trump’s odds can be attributed to a few “whales.”, i.e. big orders placed on Trump. This has sparked speculation of manipulation created a sense of momentum for the Trump campaign and has made some question the credibility of betting odds as an indicator of the real election sentiment. 3
However, we have seen Trump’s odds simultaneously rise in other betting markets apart from Polymarket, and this has also coincided with a strong USD across the board.
Therefore, our view is that whether the betting odds were manipulated or not, and whether the intentions of the “whales” were benign or malign, doesn’t really matter to us as traders. The bottom line is this: a Trump presidency is more priced in than a Harris presidency.
Market reaction scenarios
Combining the market pricing and our knowledge of the candidates’ policies we can come up with rough scenarios on how the markets will price the election. Since this report is published on Friday, we highly recommend everyone to check the implied volatility data on Monday to determine the likelihood of exact market movements!
Buy the rumour, sell fact?
Could there be a “buy the rumour, sell fact” market reaction? That is the market reverting the initial euphoria/panic completely. In one of the most famous market reactions in history, in the 2016 election, stocks went initially down and the dollar sold off 3% (below is a chart of USDJPY). However, subsequently, the markets retraced and both stocks and USD trended significantly higher.
Why did the market revert? Well, it sure wasn’t because Trump’s presidential speech was “presidential” or “conciliatory”, although that was the news outlets' first opinion! 4 Most likely explanation was that despite all the scaremongering pre-election, Trump came with a lot fiscal stimulus which was a great thing for economic growth regardless of his brash demeanour.
Could there be a similar trade here? We highly doubt it as even if we think the mechanism could still exist in certain circumstances, the massive Trump trade in 2016 was based on the market having not priced in the fiscal stimulus at all.
On a smaller scale “buy the rumour, sell the fact” trades, we think the Trump victory is so well in the price that unless Republicans sweep the house, the market might revert a lot of it on the same day, especially if there are signs of no republican sweep.
Equally, we are not sure a Harris victory will have much long-term impact on FX markets, as that is a status quo and the dollar direction will depend on the data moving forward. On an intraday basis though, we believe the USD will sell off.
Understand how the vote counting affects the markets
The votes trickle in a way that can be confusing to understand because voting via mail is more popular among Democrat voters than Republican. For the first time in American history, the most common way of voting in 2020 was to receive a ballot by mail.5 However, law changes enacted since 2020 should speed up the counting of mail-in ballots, which have tended to favour Democrats but have been embraced more widely this year by Republicans.6
Regardless, if history is any guide these factors imply that initial results are skewed towards Trump in states where counting mail-in votes in advance is not allowed and vice versa.
Market will draw its conclusions much earlier than the election is called, therefore if one wants to bet probabilistically, you have to come up with conclusion way earlier than the vote is called.7
Certain states; Many of the “certain electors” are called in earlier than the swing states. For example, if Harris does well in “red ruby” states of South Carolina and West Virginia the probabilities could go up for her victory. Depending on victory margins for each candidate this could have an impact on probabilities, though we would not draw conclusions yet.
Arizona – Initial results favor Harris (pre-counting of mail-in votes is allowed, and mail-in voting is very popular). Trump win odds at the time of writing (1.11.) stand at 66% based on probabilities from FiveThirtyEight.
Georgia – No clear favourite for initial results (pre-counting votes is allowed, can vote in advance both in person and via mail). Trump odds at 63%. GA is one of the earlier swing states and Harris lead here could flip the odds.
Michigan – Initial results probably favor Harris (large jurisdictions are allowed to count mail-in votes in advance). Harris odds 58%. If initial results favour Trump, this is meaningful, and election could be priced in quickly.
Nevada – Initial results slightly Favor Trump (the state accepts mail-in votes up to 4 days after the election as long as they are postmarked by Nov. 5.). Odds are tied.
North-Carolina – Initial results slightly favor Harris (both mail-in and in person votes are counted in advance). Trump odds 60%.
Pennsylvania – Initial results favor Trump (the counting of mail-in votes isn’t allowed until election day). Trump odds 54%.
Wisconsin – Initial results favor Trump (similar rules to Pennsylvania). Harris odds 56%. If Harris leads Pennsylvania and Wisconsin, this is meaningful.
When do we know the winner?
It depends on how down to the wire the vote and counting is. Many news providers focus on the fact the election will not be “called” until a couple of days. This does not mean the market will not price the winner fully much earlier.
During the last election, for example, in Pennsylvania, the National vote was only called on Saturday, but the market had priced in a Biden victory 90% ten hours after the polling stations closed. Trump's victory in 2016 was priced in even sooner, roughly five hours after the polls closed.
Overall reviewing our sources and internal material from last two elections, we think the election will be priced in roughly six-eight hours after the closing of polls in a base case scenario. However, in case Trump wins, especially in any of the contested swing states (PA, MI, WI), the election might be priced in a couple of hours after the polls close.
How will the probabilities shift in practice? A historical look
To showcase in practice how market moves based on probabilities, below is an excerpt of the headlines from Trump Victory in 2016 (from our internal library) and five-minute screenshots from EURUSD. Time zone is Eastern European Time (EET), i.e. six hours ahead ET.
As a background Trump odds were roughly 15% right before the polls closed, having stood at 20% a week earlier. At the time the consensus market reaction expectation was that Trump would be terrible for dollar and stocks, and the view that his policies might actually be great for the economy was a fringe opinion.
Notice how at 3.45 EET (9.45 pm ET) Trump was leading Florida by 1.1 percentage points with 91% of votes reported. (red). At the time Florida was the key swing state to watch.
Source: Ransquawk, Thales Research
Notice how Trump probabilities had arisen at 4.35 to 50%, from initial 15%. Even though at 3.45 based on Florida you could just Trump odds must be much higher than 15%, market had barely budged! Our theory is that many participants hadn’t studied the election system, leading to inefficiencies!
Now by 5.41 the odds had shifted to 90% for Trump. By that time the election had been in the price. On the above chart market EURUSD topped 1 hr after.
Concluding remark
Given markets understand the importance of betting odds much better than in previous election (we have been slightly disappointed as mainstream media has begun to pay attention!), most likely similar mispricings won’t exist this time as in 2016 election.
Most likely the trades will be very different, than in 2016 and 2020 election, but there are important mechanisms one can learn from this report
1) That the market moves initially probabilistically to the simple logic whatever is concerned by the consensus
2) Afterward there’s a possibility of “buy the rumor, sell the fact”, depending on the extent of participant being priced in and their policies.
We hope this report has been of good use!
Good luck,
Thales Research team
ps. Stay tuned for more insights on our trading and proprietary research.
Based on aggregate of betting odds from Polymarket, Predictit, Fivethirtyeight and Betfair.
https://www.bloomberg.com/news/features/2024-10-25/polymarket-whale-traders-how-1-of-bettors-are-boosting-trump-s-odds
https://www.bloomberg.com/news/articles/2016-11-09/dollar-erases-loss-as-clinton-leads-in-florida-north-carolina
For potential timelines, see https://www.livenowfox.com/news/election-results-vote-count-timeline-swing-states-2024
This is amazing research! Thank you very much for sharing, let's see if we get some volatility again in these elections.