US – Midterm update: Pre-election tail risk scenarios

With two weeks until the US midterm elections on 3 November, early voting has already opened in more than 40 states. Prediction markets are broadly unchanged since we published our Midterm update, 13 October. The prediction market Polymarket continues to see Democrats winning both chambers as the most likely outcome, with a 62.5% probability.

This has created a growing sense of urgency, with President Trump and the administration appearing ready to 1) implement policies before Republicans potentially lose their majorities in both chambers of Congress and 2) pursue policies aimed at swaying voters towards Republican candidates.

1) became clear when the White House formed a committee to probe Fed Governor Cook, whom Trump has repeatedly tried to remove from the Fed Board of Governors. Trump appears pressured to remove Cook before the new Senate takes office in early January, as a Democratic majority could block his Fed Board nominations even if he removed Cook later. We do not expect him to succeed before the new Congress takes office. The prediction market Kalshi put the probability of Cook leaving or being removed before January at below 10%, and only a 25% chance of it happening before 2029.

The first point could also become clear if Republicans seek to pass the full-year government funding package when the current stopgap bill expires on 11 December. We doubt any spending increases would materially affect the US macro-outlook. The early framework passed by the House in July included USD 95 billion in new spending, or 0.3% of GDP, and was already criticised by Republican fiscal hardliners. For details, see RtM USD, 20 October. 

2) The first example is the suggestion for a payment to every adult American if Republicans win the midterms. Furthermore, 2) became clear when Trump signalled willingness to strike a diesel deal favourable to Russia, under which Russia would release diesel to world markets to help ease US diesel prices ahead of the midterms. So far, the decline in US spot and retail diesel prices has been limited, with most of it attributable to seasonal patterns. The probability that Trump would impose a problematic diesel export is now only at 5% at Kalshi, which we have flagged several times.

Trump could try to interfere directly with voting on Election Day. Kalshi assigns a 13% probability to Trump declaring a national emergency before Election Day and a 29% probability before the end of 2026. However, such a declaration should not give the president power over elections, as these are managed at state level.

An even more severe measure would be for Trump to invoke the Insurrection Act or declare martial law. The former could be used to deploy the military at polling stations, which Kalshi assigns a 17% probability before the end of 2026. Martial law would mean military authority replacing civilian government and would be even more extreme. A low volume Kalshi markets estimates 24% probability of this happening in 2026. Trump should not legally be able to use these powers to take control of elections, but he could still try.

Download The Full Research US