[Week 45 of 2026] Elections and Markets

Welcome back to Price and Prejudice with a few musings from Week 45 of 2026. Today we talk about elections, and how to think about what they do to stock, bond, and other asset prices.

Democrats won both chambers of Congress on Tuesday, by much more than the polls suggested. Markets did not seem to know what to make of it at first. Stock futures fell on Tuesday night as the results came in, and then on Wednesday stocks rallied, Treasury yields fell (i.e. prices went up), the dollar fell, and the VIX, a measure of how much investors pay to protect themselves against big moves in the stock market, dropped sharply. So it is fair to ask what, if anything, the election did to asset prices, and how one should think about it.

As always, a useful place to start is that the price of a stock or a bond reflects the cash flows investors expect to receive, discounted for time and for risk. An election is such a large event and it can change each of these. For example, it can change expected cash flows, through taxes, regulation, and spending. It can also change interest rates, through the path of deficits and inflation. And it can change how much investors demand to be paid for bearing risk.

In this thought process, it is also important to remember that prices move only on the part of the result that was not already expected. A party winning an election that everyone expected it to win should not move prices much, however much its policies matter.

The cash-flow channel is the most intuitive. This election ends unified Republican control, and from January Democrats will control Congress while the president keeps his veto. That makes some policies less likely to pass and others more likely to stall, and different companies care about different policies. The clearest example this week was data centers (see my colleague Stijn van Nieuwerburgh's quick recap). Several newly elected Democratic governors had campaigned on slowing data-center construction in their states, and AI infrastructure stocks fell on Wednesday even as the broader market rose.

A different but related question is whether stocks simply do better when one party is in power. Here there is a famous and somewhat surprising fact: since the 1920s, US stocks have earned much higher returns under Democratic presidents than under Republican ones, by about 9 pp per year in the original study by Pedro Santa-Clara and Rossen Valkanov. This is the opposite of what many investors would guess. It is also less robust than it sounds, since there have only been a couple of dozen presidential terms, and a few episodes, such as the Depression and the 2008 financial crisis, do a lot of the work. Later work using different methods finds a smaller and less stable gap.

The most interesting explanation, by Lubos Pastor and Pietro Veronesi, argues that the gap need not have anything to do with what Democratic presidents actually do. When times are bad and people become more afraid of risk, two things happen at once. Voters want more protection from the government, so they are more likely to elect Democrats, and investors demand higher returns for holding stocks, so stock prices fall and future expected returns rise. The high returns that follow are then compensation for the fear that got the Democrat elected in the first place. For example, Roosevelt in 1932 and Obama in 2008 both took office at moments like this.

The last channel, and probably the most relevant for this week, is uncertainty itself. Before an election, investors do not know which policies they will get, and many of them pay to protect themselves against the bad outcomes. Options that expire just after an election are more expensive than options that expire just before it, which is exactly what one would expect if political risk carries a price. Once the result is known, that protection becomes less valuable, the premium investors demand for bearing political risk falls, and prices can rise even if the winner is not the one investors would have chosen. In some sense, this is one way to read Wednesday, when the VIX dropped and stocks rose despite a result many investors had not expected. It does not mean every election leads to a rally, however, since a large enough policy surprise can outweigh the relief of knowing the result.

So elections clearly move asset prices, through expected policy, through interest rates, and through the price of risk. Different asset classes can move in opposite directions depending on which of these dominates. Unfortunately, the partisan gap in returns rests on a small number of presidential terms and could partly reflect when parties win rather than what they do, and the well-known pattern of strong stock returns after midterms has the same small-sample problem.

One way around this is to look more closely at the elections we do have. On election night, prediction markets update the odds of each outcome minute by minute while stocks, bonds, and currencies keep trading, so one can see how much each asset moves for a given change in the odds. Another is to add elections in other countries, which is how researchers learned that options covering elections are more expensive. Simulated voters, which political scientists have started to use in place of surveys, cannot add new elections in the same way, since a simulation can only tell us what we already assumed about how markets work. But perhaps that might change soon as well.

Other Stuff

• Hollywood studios are telling employees not to bet on prediction markets about casting, release dates, or earnings.
• Big banks are cracking down on people posting their research charts on X, partly out of fear that AI tools will copy them. Bank research was always hard to keep private, and AI makes copying it almost free (but you still need the data!)
• Tesla is winding down Cybertruck production, which shows that a product can get enormous attention without many people actually buying it.